github
/
mpv
mirror of https://github.com/mpv-player/mpv
1
Fork 0
Code Releases Activity
mpv/mpvcore/input/input.c

2427 lines
73 KiB
C
Raw Normal View History

Add standard GPL license header. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@28146 b3059339-0415-0410-9bf9-f77b7e298cf2
2008-12-13 19:28:00 +01:00
/*
* This file is part of MPlayer.
*
* MPlayer is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* MPlayer is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License along
* with MPlayer; if not, write to the Free Software Foundation, Inc.,
* 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
*/
Unify include paths, -I.. is in CFLAGS. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@16861 b3059339-0415-0410-9bf9-f77b7e298cf2
2005-10-25 23:23:45 +02:00
#include "config.h"
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
#include <stdlib.h>
#include <string.h>
#include <stdio.h>
Merge svn changes up to r28951
2009-03-14 22:52:45 +01:00
#include <stdbool.h>
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
#include <unistd.h>
#include <errno.h>
#include <sys/types.h>
New feature to let mplayer read commands from a file. Mostly useful with a fifo. Turn it one with -input file=/path/. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@7841 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-10-22 16:32:17 +02:00
#include <sys/stat.h>
Tell the user when we fallback on default config. Added some default joystick bindings git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4526 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-04 15:48:32 +01:00
#include <sys/time.h>
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
#include <fcntl.h>
Add support for hex key code in input.conf git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@6395 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-06-12 12:39:04 +02:00
#include <ctype.h>
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
#include <assert.h>
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
cleanup: remove duplicated function, move escape parsing function
2013-02-16 22:51:10 +01:00
#include <libavutil/avstring.h>
#include <libavutil/common.h>
windows support: unicode filenames Windows uses a legacy codepage for char* / runtime functions accepting char *. Using UTF-8 as the codepage with setlocale() is explicitly forbidden. Work this around by overriding the MSVCRT functions with wrapper macros, that assume UTF-8 and use "proper" API calls like _wopen etc. to deal with unicode filenames. All code that uses standard functions that take or return filenames must now include osdep/io.h. stat() can't be overridden, because MinGW-w64 itself defines "stat" as a macro. Change code to use use mp_stat() instead. This is not perfectly clean, but still somewhat sane, and much better than littering the rest of the mplayer code with MinGW specific hacks. It's also a bit fragile, but that's actually little different from the previous situation. Also, MinGW is unlikely to ever include a nice way of dealing with this.
2012-02-03 08:05:11 +01:00
#include "osdep/io.h"
mp_msg, getch2: unix tty background support Now, when backgrounded, mpv plays and outputs messages to stdout, but statusline is not output. Background<->foreground transitions are detected by signals and polling the process groups.
2013-01-23 15:47:49 +01:00
#include "osdep/getch2.h"
windows support: unicode filenames Windows uses a legacy codepage for char* / runtime functions accepting char *. Using UTF-8 as the codepage with setlocale() is explicitly forbidden. Work this around by overriding the MSVCRT functions with wrapper macros, that assume UTF-8 and use "proper" API calls like _wopen etc. to deal with unicode filenames. All code that uses standard functions that take or return filenames must now include osdep/io.h. stat() can't be overridden, because MinGW-w64 itself defines "stat" as a macro. Change code to use use mp_stat() instead. This is not perfectly clean, but still somewhat sane, and much better than littering the rest of the mplayer code with MinGW specific hacks. It's also a bit fragile, but that's actually little different from the previous situation. Also, MinGW is unlikely to ever include a nice way of dealing with this.
2012-02-03 08:05:11 +01:00
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
#include "input.h"
input: move all key code lists to input/keycodes.h Move the definitions of all special key codes (those not passed by ASCII value) to input/keycodes.h. Before they were spread between osdep/keycodes.h, input/joystick.h, input/mouse.h and input/ar.h, plus some special values in input.h. This was especially inconvenient as the codes had to be coordinated to not conflict between the files. The change requires a bit of ugliness as appleir.c includes <linux/input.h> which contains various conflicting KEY_* definitions. Work around this by adding a special preprocessor variable which can be used to avoid defining these in keycodes.h.
2011-04-25 08:43:59 +02:00
#include "keycodes.h"
Unify include paths, -I.. is in CFLAGS. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@16861 b3059339-0415-0410-9bf9-f77b7e298cf2
2005-10-25 23:23:45 +02:00
#include "osdep/timer.h"
core: move contents to mpvcore (2/2) Followup commit. Fixes all the files references.
2013-08-06 22:41:30 +02:00
#include "mpvcore/mp_msg.h"
#include "mpvcore/m_config.h"
#include "mpvcore/m_option.h"
#include "mpvcore/path.h"
Create a context for input.c state Start moving static variables to a context struct. Only autorepeat state is moved to the struct in this commit. mp_input_check_interrupt now requires the context variable. Change stream functions to pass it. It's still stored in a static variable in stream/.
2008-04-30 06:15:52 +02:00
#include "talloc.h"
core: move contents to mpvcore (2/2) Followup commit. Fixes all the files references.
2013-08-06 22:41:30 +02:00
#include "mpvcore/options.h"
#include "mpvcore/bstr.h"
input: simplify input.conf parsing Now input.conf is loaded into memory at once, instead of streaming the file into the parser. The real reason for this change is that I want to be able to read the config file from memory. (Using fmemopen() would have been simpler, but that is available on sane platforms only.)
2012-08-01 21:50:24 +02:00
#include "stream/stream.h"
core: move contents to mpvcore (2/2) Followup commit. Fixes all the files references.
2013-08-06 22:41:30 +02:00
#include "mpvcore/mp_common.h"
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
#include "joystick.h"
Rename a bunch of miscellaneous preprocessor directives. Switch them from a HAVE_ to a CONFIG_ prefix. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@27423 b3059339-0415-0410-9bf9-f77b7e298cf2
2008-08-07 10:58:07 +02:00
#ifdef CONFIG_LIRC
Added lirc support in input git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4433 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-31 10:39:11 +01:00
#include "lirc.h"
#endif
Rename a bunch of miscellaneous preprocessor directives. Switch them from a HAVE_ to a CONFIG_ prefix. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@27423 b3059339-0415-0410-9bf9-f77b7e298cf2
2008-08-07 10:58:07 +02:00
#ifdef CONFIG_LIRCC
lirccd support by Fredrik Tolf <fredrik@dolda2000.cjb.net> git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@10216 b3059339-0415-0410-9bf9-f77b7e298cf2
2003-05-30 20:23:55 +02:00
#include <lirc/lircc.h>
#endif
osx: create macosx_events to deal with keyDown events On OSX with Cocoa enabled keyDown events are now handled with addLocalMonitorForEventsMatchingMask:handler:. This allows to respond to events even when there is no VO initialized but the GUI is focused.
2013-06-02 17:39:05 +02:00
#ifdef CONFIG_COCOA
#include "osdep/macosx_events.h"
#endif
input: make input queue thread safe If pthreads are enabled the input queue accesses are regulated by acquiring a mutex. This is useful for platforms like OS X, where the events are created in the cocoa thread and added to the queue to then be dequeued in the playloop thread.
2013-08-13 14:40:17 +02:00
#if HAVE_PTHREADS
#include <pthread.h>
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
#define input_lock(ictx) pthread_mutex_lock(&ictx->mutex)
#define input_unlock(ictx) pthread_mutex_unlock(&ictx->mutex)
input: make input queue thread safe If pthreads are enabled the input queue accesses are regulated by acquiring a mutex. This is useful for platforms like OS X, where the events are created in the cocoa thread and added to the queue to then be dequeued in the playloop thread.
2013-08-13 14:40:17 +02:00
#else
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
#define input_lock(ictx) 0
#define input_unlock(ictx) 0
input: make input queue thread safe If pthreads are enabled the input queue accesses are regulated by acquiring a mutex. This is useful for platforms like OS X, where the events are created in the cocoa thread and added to the queue to then be dequeued in the playloop thread.
2013-08-13 14:40:17 +02:00
#endif
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
#define MP_MAX_KEY_DOWN 4
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
struct cmd_bind {
input: store number of keys, instead of using 0-termination
2013-06-23 00:07:45 +02:00
int keys[MP_MAX_KEY_DOWN];
int num_keys;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
char *cmd;
input: track location info for input commands If parsing a command fails, its location is printed. The location is the path to the input.conf, and the line number of the key binding and the associated input command.
2012-10-13 21:09:42 +02:00
char *location; // filename/line number of definition
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
bool is_builtin;
input: track location info for input commands If parsing a command fails, its location is printed. The location is the path to the input.conf, and the line number of the key binding and the associated input command.
2012-10-13 21:09:42 +02:00
struct cmd_bind_section *owner;
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
};
input.[ch]: Move some type definitions from .h to .c These aren't used outside input.c.
2008-04-30 08:53:12 +02:00
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
struct key_name {
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
int key;
char *name;
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
};
input.[ch]: Move some type definitions from .h to .c These aren't used outside input.c.
2008-04-30 08:53:12 +02:00
input.c: simplify command definitions Change command definitions to allow writing the command definition table in more readable format.
2012-03-21 01:23:35 +01:00
/* This array defines all known commands.
* The first field is an id used to recognize the command.
* The second is the command name used in slave mode and input.conf.
* Then comes the definition of each argument, first mandatory arguments
* (ARG_INT, ARG_FLOAT, ARG_STRING) if any, then optional arguments
* (OARG_INT(default), etc) if any. The command will be given the default
* argument value if the user didn't give enough arguments to specify it.
* A command can take a maximum of MP_CMD_MAX_ARGS arguments (10).
*/
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
#define ARG_INT { .type = {"", NULL, &m_option_type_int} }
#define ARG_FLOAT { .type = {"", NULL, &m_option_type_float} }
core: change speed option/property to double The --speed option and the speed property used float. Change them to double. Change the commands that manipulate the property (speed_mult/add) to double as well. Since the cycle command shares code with the add command, we change that as well. The reason for this change is that this allows better control over speed, such as stepping by semitones. Using floats is also just plain unnecessary.
2013-08-04 23:56:20 +02:00
#define ARG_DOUBLE { .type = {"", NULL, &m_option_type_double} }
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
#define ARG_STRING { .type = {"", NULL, &m_option_type_string} }
#define ARG_CHOICE(c) { .type = {"", NULL, &m_option_type_choice, \
M_CHOICES(c)} }
commands: parse seek time arguments like time options This means a commands like "seek 13:00 absolute" actually behaves like "--start=13:00", instead of interpreting the argument as fraction as with normal float options. This is probably slightly closer to what you'd expect. As a consequence, the seek argument's type changes from float to double internally.
2013-02-24 21:16:23 +01:00
#define ARG_TIME { .type = {"", NULL, &m_option_type_time} }
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
core: change speed option/property to double The --speed option and the speed property used float. Change them to double. Change the commands that manipulate the property (speed_mult/add) to double as well. Since the cycle command shares code with the add command, we change that as well. The reason for this change is that this allows better control over speed, such as stepping by semitones. Using floats is also just plain unnecessary.
2013-08-04 23:56:20 +02:00
#define OARG_DOUBLE(def) { .type = {"", NULL, &m_option_type_double}, \
.optional = true, .v.d = def }
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
#define OARG_INT(def) { .type = {"", NULL, &m_option_type_int}, \
.optional = true, .v.i = def }
#define OARG_CHOICE(def, c) { .type = {"", NULL, &m_option_type_choice, \
M_CHOICES(c)}, \
.optional = true, .v.i = def }
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
commands: use "up" and "down" as 2nd argument for cycle command Allow the values "up" and "down" as step argument for the cycle input command. Previously, this argument was a float, which specified an arbitrary step value and direction (similar to the add command). Instead of "1" and "-1", "up" and "down" is to be used. Float values are still accepted. That capability might be removed in the future, as there's probably hardly any actual use for arbitrary step values.
2012-10-12 09:21:26 +02:00
static int parse_cycle_dir(const struct m_option *opt, struct bstr name,
struct bstr param, void *dst);
static const struct m_option_type m_option_type_cycle_dir = {
.name = "up|down",
.parse = parse_cycle_dir,
};
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
Make several mapping tables related to input processing const. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@25737 b3059339-0415-0410-9bf9-f77b7e298cf2
2008-01-13 17:13:54 +01:00
static const mp_cmd_t mp_cmds[] = {
input: warn about non-existing commands, add explicit "ignore" command If a command is not found, warn about it at loading time (just like other command parsing errors are printed at loading time). Add an explicit "ignore" command. input.conf instructs users to use this command to cancel out existing mapping. This clashed with the warning added in this commit. Make "ignore" a real command and remove the specialcasing for it from get_cmd_from_keys(). Now "ignore" is ignored because it's not handled in command.c.
2012-09-22 15:17:15 +02:00
{ MP_CMD_IGNORE, "ignore", },
input: don't warn on commands for disabled optional features Always recognize input commands for optional features (like TV commands etc.). If these features are disabled, the commands are parsed, but simply do nothing. This fixes annoying warnings on start with the default/builtin input.conf, if certain optional features are not compiled.
2012-10-13 12:31:19 +02:00
input.c: simplify command definitions Change command definitions to allow writing the command definition table in more readable format.
2012-03-21 01:23:35 +01:00
{ MP_CMD_RADIO_STEP_CHANNEL, "radio_step_channel", { ARG_INT } },
{ MP_CMD_RADIO_SET_CHANNEL, "radio_set_channel", { ARG_STRING } },
{ MP_CMD_RADIO_SET_FREQ, "radio_set_freq", { ARG_FLOAT } },
{ MP_CMD_RADIO_STEP_FREQ, "radio_step_freq", {ARG_FLOAT } },
input: don't warn on commands for disabled optional features Always recognize input commands for optional features (like TV commands etc.). If these features are disabled, the commands are parsed, but simply do nothing. This fixes annoying warnings on start with the default/builtin input.conf, if certain optional features are not compiled.
2012-10-13 12:31:19 +02:00
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
{ MP_CMD_SEEK, "seek", {
commands: parse seek time arguments like time options This means a commands like "seek 13:00 absolute" actually behaves like "--start=13:00", instead of interpreting the argument as fraction as with normal float options. This is probably slightly closer to what you'd expect. As a consequence, the seek argument's type changes from float to double internally.
2013-02-24 21:16:23 +01:00
ARG_TIME,
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
OARG_CHOICE(0, ({"relative", 0}, {"0", 0},
{"absolute-percent", 1}, {"1", 1},
{"absolute", 2}, {"2", 2})),
OARG_CHOICE(0, ({"default-precise", 0}, {"0", 0},
{"exact", 1}, {"1", 1},
{"keyframes", -1}, {"-1", -1})),
}},
core: change speed option/property to double The --speed option and the speed property used float. Change them to double. Change the commands that manipulate the property (speed_mult/add) to double as well. Since the cycle command shares code with the add command, we change that as well. The reason for this change is that this allows better control over speed, such as stepping by semitones. Using floats is also just plain unnecessary.
2013-08-04 23:56:20 +02:00
{ MP_CMD_SPEED_MULT, "speed_mult", { ARG_DOUBLE } },
input.c: simplify command definitions Change command definitions to allow writing the command definition table in more readable format.
2012-03-21 01:23:35 +01:00
{ MP_CMD_QUIT, "quit", { OARG_INT(0) } },
core: add playback resume feature (manual/opt-in) A "watch later" command is now mapped to Shift+Q. This quits the player and stores the playback state in a config file in ~/.mpv/watch_later/. When calling the player with the same file again, playback is resumed at that time position. It's also possible to make mpv save playback state always on quit with the --save-position-on-quit option. Likewise, resuming can be disabled with the --no-resume-playback option. This also attempts to save some playback parameters, like fullscreen state or track selection. This will unconditionally override config settings and command line options (which is probably not what you would expect, but in general nobody will really care about this). Some things are not backed up, because that would cause various problems. Additional subtitle files, video filters, etc. are not stored because that would be too hard and fragile. Volume/mute state are not stored because it would mess up if the system mixer is used, or if the system mixer was readjusted in the meantime. Basically, the tradeoff between perfect state restoration and complexity/fragility makes it not worth to attempt to implement it perfectly, even if the result is a little bit inconsistent.
2013-05-05 19:37:29 +02:00
{ MP_CMD_QUIT_WATCH_LATER, "quit_watch_later", },
input.c: simplify command definitions Change command definitions to allow writing the command definition table in more readable format.
2012-03-21 01:23:35 +01:00
{ MP_CMD_STOP, "stop", },
{ MP_CMD_FRAME_STEP, "frame_step", },
core: add backstep support Allows stepping back one frame via the frame_back_step inout command, bound to "," by default. This uses the precise seeking facility, and a perfect frame index built on the fly. The index is built during playback and precise seeking, and contains (as of this commit) the last 100 displayed or skipped frames. This index is used to find the PTS of the previous frame, which is then used as target for a precise seek. If no PTS is found, the core attempts to do a seek before the current frame, and skip decoded frames until the current frame is reached; this will create a sufficient index and the normal backstep algorithm can be applied. This can be rather slow. The worst case for backstepping is about the same as the worst case for precise seeking if the previous frame can be deduced from the index. If not, the worst case will be twice as slow. There's also some minor danger that the index is incorrect in case framedropping is involved. For framedropping due to --framedrop, this problem is ignored (use of --framedrop is discouraged anyway). For framedropping during precise seeking (done to make it faster), we try to not add frames to the index that are produced when this can happen. I'm not sure how well that works (or if the logic is sane), and it's sure to break with some video filters. In the worst case, backstepping might silently skip frames if you backstep after a user-initiated precise seek. (Precise seeks to do indexing are not affected.) Likewise, video filters that somehow change timing of frames and do not do this in a deterministic way (i.e. if you seek to a position, frames with different timings are produced than when the position is reached during normal playback) will make backstepping silently jump to the wrong frame. Enabling/disabling filters during playback (like for example deinterlacing) will have similar bad effects.
2013-04-24 19:31:48 +02:00
{ MP_CMD_FRAME_BACK_STEP, "frame_back_step", },
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
{ MP_CMD_PLAYLIST_NEXT, "playlist_next", {
OARG_CHOICE(0, ({"weak", 0}, {"0", 0},
{"force", 1}, {"1", 1})),
}},
{ MP_CMD_PLAYLIST_PREV, "playlist_prev", {
OARG_CHOICE(0, ({"weak", 0}, {"0", 0},
{"force", 1}, {"1", 1})),
}},
commands: sub_step takes only one argument
2012-09-27 01:07:19 +02:00
{ MP_CMD_SUB_STEP, "sub_step", { ARG_INT } },
input.c: simplify command definitions Change command definitions to allow writing the command definition table in more readable format.
2012-03-21 01:23:35 +01:00
{ MP_CMD_OSD, "osd", { OARG_INT(-1) } },
commands: add print_text input command to print text on the terminal In theory, this could take over the role of the get_property slave command, and is more general.
2012-10-11 02:24:13 +02:00
{ MP_CMD_PRINT_TEXT, "print_text", { ARG_STRING } },
commands: rename osd_show_[property_]text and osd_show_progression osd_show_[property_]text => show_text osd_show_progression => show_progress show_text, osd_show_property_text and osd_show_text both map to the code for the previous osd_show_property_text. The only special thing about osd_show_text is that you don't need to escape "$". Also, unfortunately osd_show_property_text requires escaping things twice, one time for the command parser, and the other time for the property formatting code, while osd_show_text needed only one level of escaping.
2012-09-15 02:04:57 +02:00
{ MP_CMD_SHOW_TEXT, "show_text", { ARG_STRING, OARG_INT(-1), OARG_INT(0) } },
{ MP_CMD_SHOW_PROGRESS, "show_progress", },
command: add sub_reload and sub_remove commands sub_remove remove an external subtitle track, for whatever this may be needed. sub_reload removes and re-adds an external subtitle track. Also rename sub_load to sub_add, because that seems to be more in line with sub_remove.
2012-11-15 20:26:52 +01:00
{ MP_CMD_SUB_ADD, "sub_add", { ARG_STRING } },
{ MP_CMD_SUB_REMOVE, "sub_remove", { OARG_INT(-1) } },
{ MP_CMD_SUB_RELOAD, "sub_reload", { OARG_INT(-1) } },
input: don't warn on commands for disabled optional features Always recognize input commands for optional features (like TV commands etc.). If these features are disabled, the commands are parsed, but simply do nothing. This fixes annoying warnings on start with the default/builtin input.conf, if certain optional features are not compiled.
2012-10-13 12:31:19 +02:00
input.c: simplify command definitions Change command definitions to allow writing the command definition table in more readable format.
2012-03-21 01:23:35 +01:00
{ MP_CMD_TV_START_SCAN, "tv_start_scan", },
{ MP_CMD_TV_STEP_CHANNEL, "tv_step_channel", { ARG_INT } },
{ MP_CMD_TV_STEP_NORM, "tv_step_norm", },
{ MP_CMD_TV_STEP_CHANNEL_LIST, "tv_step_chanlist", },
{ MP_CMD_TV_SET_CHANNEL, "tv_set_channel", { ARG_STRING } },
{ MP_CMD_TV_LAST_CHANNEL, "tv_last_channel", },
{ MP_CMD_TV_SET_FREQ, "tv_set_freq", { ARG_FLOAT } },
{ MP_CMD_TV_STEP_FREQ, "tv_step_freq", { ARG_FLOAT } },
{ MP_CMD_TV_SET_NORM, "tv_set_norm", { ARG_STRING } },
input: don't warn on commands for disabled optional features Always recognize input commands for optional features (like TV commands etc.). If these features are disabled, the commands are parsed, but simply do nothing. This fixes annoying warnings on start with the default/builtin input.conf, if certain optional features are not compiled.
2012-10-13 12:31:19 +02:00
input.c: simplify command definitions Change command definitions to allow writing the command definition table in more readable format.
2012-03-21 01:23:35 +01:00
{ MP_CMD_DVB_SET_CHANNEL, "dvb_set_channel", { ARG_INT, ARG_INT } },
input: don't warn on commands for disabled optional features Always recognize input commands for optional features (like TV commands etc.). If these features are disabled, the commands are parsed, but simply do nothing. This fixes annoying warnings on start with the default/builtin input.conf, if certain optional features are not compiled.
2012-10-13 12:31:19 +02:00
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
{ MP_CMD_SCREENSHOT, "screenshot", {
screenshot: change "screenshot" input command "screenshot" now maps to "screenshot subtitles" by default, instead of "screenshot video". Swap the argument order: the more useful argument should come first. Remove the compatibility aliases for numeric choices (e.g. "screenshot 1 0" won't work anymore).
2012-10-20 23:58:02 +02:00
OARG_CHOICE(2, ({"video", 0},
{"window", 1},
screenshot: allow taking screenshots with subtitles This adds a new screenshot mode "subtitles", which basically takes the video frame as decoded, and renders subtitles into it. This may fail for some pixel formats, because libswscale sucks. If this becomes ever a real problem, the code could be changed to convert the image to RGBA first (or whatever the image writer wants), and then render the subtitles into it. This would avoid the additional image copy needed with vo_xv too. But for now, it seems better to go with the current method in the common case: vo_opengl creates an image copy anyway, and drawing bitmaps to yv12 is better, as no color space conversion is involved in draw_bmp.c's up/downsampling conversion.
2012-10-20 23:32:45 +02:00
{"subtitles", 2})),
screenshot: change "screenshot" input command "screenshot" now maps to "screenshot subtitles" by default, instead of "screenshot video". Swap the argument order: the more useful argument should come first. Remove the compatibility aliases for numeric choices (e.g. "screenshot 1 0" won't work anymore).
2012-10-20 23:58:02 +02:00
OARG_CHOICE(0, ({"single", 0},
{"each-frame", 1})),
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
}},
command: add screenshot_to_file command
2013-07-08 20:34:26 +02:00
{ MP_CMD_SCREENSHOT_TO_FILE, "screenshot_to_file", {
ARG_STRING,
OARG_CHOICE(2, ({"video", 0},
{"window", 1},
{"subtitles", 2})),
}},
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
{ MP_CMD_LOADFILE, "loadfile", {
ARG_STRING,
OARG_CHOICE(0, ({"replace", 0}, {"0", 0},
{"append", 1}, {"1", 1})),
}},
{ MP_CMD_LOADLIST, "loadlist", {
ARG_STRING,
OARG_CHOICE(0, ({"replace", 0}, {"0", 0},
{"append", 1}, {"1", 1})),
}},
mplayer: turn playtree into a list, and change per-file option handling Summary: - There is no playtree anymore. It's reduced to a simple list. - Options are now always global. You can still have per-file options, but these are optional and require special syntax. - The slave command pt_step has been removed, and playlist_next and playlist_prev added. (See etc/input.conf changes.) This is a user visible incompatible change, and will break slave-mode applications. - The pt_clear slave command is renamed to playlist_clear. - Playtree entries could have multiple files. This is not the case anymore, and playlist entries have always exactly one entry. Whenever something adds more than one file (like ASX playlists or dvd:// or dvdnav:// on the command line), all files are added as separate playlist entries. Note that some of the changes are quite deep and violent. Expect regressions. The playlist parsing code in particular is of low quality. I didn't try to improve it, and merely spent to least effort necessary to keep it somehow working. (Especially ASX playlist handling.) The playtree code was complicated and bloated. It was also barely used. Most users don't even know that mplayer manages the playlist as tree, or how to use it. The most obscure features was probably specifying a tree on command line (with '{' and '}' to create/close tree nodes). It filled the player code with complexity and confused users with weird slave commands like pt_up. Replace the playtree with a simple flat playlist. Playlist parsers that actually return trees are changed to append all files to the playlist pre-order. It used to be the responsibility of the playtree code to change per-file config options. Now this is done by the player core, and the playlist code is free of such details. Options are not per-file by default anymore. This was a very obscure and complicated feature that confused even experienced users. Consider the following command line: mplayer file1.mkv file2.mkv --no-audio file3.mkv This will disable the audio for file2.mkv only, because options are per-file by default. To make the option affect all files, you're supposed to put it before the first file. This is bad, because normally you don't need per-file options. They are very rarely needed, and the only reasonable use cases I can imagine are use of the encode backend (mplayer encode branch), or for debugging. The normal use case is made harder, and the feature is perceived as bug. Even worse, correct usage is hard to explain for users. Make all options global by default. The position of an option isn't significant anymore (except for options that compensate each other, consider --shuffle --no-shuffle). One other important change is that no options are reset anymore if a new file is started. If you change settings with slave mode commands, they will not be changed by playing a new file. (Exceptions include settings that are too file specific, like audio/subtitle stream selection.) There is still some need for per-file options. Debugging and encoding are use cases that profit from per-file options. Per-file profiles (as well as per-protocol and per-VO/AO options) need the implementation related mechanisms to backup and restore options when the playback file changes. Simplify the save-slot stuff, which is possible because there is no hierarchical play tree anymore. Now there's a simple backup field. Add a way to specify per-file options on command line. Example: mplayer f1.mkv -o0 --{ -o1 f2.mkv -o2 f3.mkv --} f4.mkv -o3 will have the following options per file set: f1.mkv, f4.mkv: -o0 -o3 f2.mkv, f3.mkv: -o0 -o3 -o1 -o2 The options --{ and --} start and end per-file options. All files inside the { } will be affected by the options equally (similar to how global options and multiple files are handled). When playback of a file starts, the per-file options are set according to the command line. When playback ends, the per-file options are restored to the values when playback started.
2012-07-31 21:33:26 +02:00
{ MP_CMD_PLAYLIST_CLEAR, "playlist_clear", },
command: add some playlist manipulation commands playlist_remove and playlist_move.
2013-07-02 13:17:50 +02:00
{ MP_CMD_PLAYLIST_REMOVE, "playlist_remove", { ARG_INT } },
{ MP_CMD_PLAYLIST_MOVE, "playlist_move", { ARG_INT, ARG_INT } },
input.c: simplify command definitions Change command definitions to allow writing the command definition table in more readable format.
2012-03-21 01:23:35 +01:00
{ MP_CMD_RUN, "run", { ARG_STRING } },
add half new input support for gui git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4859 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-25 14:14:27 +01:00
input.c: simplify command definitions Change command definitions to allow writing the command definition table in more readable format.
2012-03-21 01:23:35 +01:00
{ MP_CMD_KEYDOWN_EVENTS, "key_down_event", { ARG_INT } },
commands: change input commands to make OSD usage explicit Most input commands had their own policy whether to display an OSD message for user feedback or not. Some commands had two variants, one that showed an OSD message and one that didn't (e.g. step_property_osd and step_property). Change it such that all commands show a message on the OSD. Add a "no-osd" modifier that disables OSD for that command. Rename the "step_property" and "step_property_osd" command to "switch", and rename "set_property" and "set_property_osd" to "set". Note that commands which haven't used OSD before still don't use OSD. That will possibly be fixed later. (E.g. "screenshot" could display an OSD message instead of just printing a message on the terminal.) The chapter and edition properties still produce OSD messages even with "no-osd", because they don't map so well to the property_osd_display[] mechanism.
2012-09-09 02:08:08 +02:00
{ MP_CMD_SET, "set", { ARG_STRING, ARG_STRING } },
input.c: simplify command definitions Change command definitions to allow writing the command definition table in more readable format.
2012-03-21 01:23:35 +01:00
{ MP_CMD_GET_PROPERTY, "get_property", { ARG_STRING } },
core: change speed option/property to double The --speed option and the speed property used float. Change them to double. Change the commands that manipulate the property (speed_mult/add) to double as well. Since the cycle command shares code with the add command, we change that as well. The reason for this change is that this allows better control over speed, such as stepping by semitones. Using floats is also just plain unnecessary.
2013-08-04 23:56:20 +02:00
{ MP_CMD_ADD, "add", { ARG_STRING, OARG_DOUBLE(0) } },
commands: use "up" and "down" as 2nd argument for cycle command Allow the values "up" and "down" as step argument for the cycle input command. Previously, this argument was a float, which specified an arbitrary step value and direction (similar to the add command). Instead of "1" and "-1", "up" and "down" is to be used. Float values are still accepted. That capability might be removed in the future, as there's probably hardly any actual use for arbitrary step values.
2012-10-12 09:21:26 +02:00
{ MP_CMD_CYCLE, "cycle", {
ARG_STRING,
{ .type = {"", NULL, &m_option_type_cycle_dir},
.optional = true,
core: change speed option/property to double The --speed option and the speed property used float. Change them to double. Change the commands that manipulate the property (speed_mult/add) to double as well. Since the cycle command shares code with the add command, we change that as well. The reason for this change is that this allows better control over speed, such as stepping by semitones. Using floats is also just plain unnecessary.
2013-08-04 23:56:20 +02:00
.v.d = 1 },
commands: use "up" and "down" as 2nd argument for cycle command Allow the values "up" and "down" as step argument for the cycle input command. Previously, this argument was a float, which specified an arbitrary step value and direction (similar to the add command). Instead of "1" and "-1", "up" and "down" is to be used. Float values are still accepted. That capability might be removed in the future, as there's probably hardly any actual use for arbitrary step values.
2012-10-12 09:21:26 +02:00
}},
input.c: simplify command definitions Change command definitions to allow writing the command definition table in more readable format.
2012-03-21 01:23:35 +01:00
command: add commands to enable/disable input sections For now, it's mostly for testing. It also might allow to create key binding state machines, but this sounds questionable.
2013-06-23 01:28:28 +02:00
{ MP_CMD_ENABLE_INPUT_SECTION, "enable_section", {
ARG_STRING,
OARG_CHOICE(0, ({"default", 0},
{"exclusive", 1})),
}},
{ MP_CMD_DISABLE_INPUT_SECTION, "disable_section", { ARG_STRING } },
audio/filter: use new option API Make the VF/VO/AO option parser available to audio filters. No audio filter uses this yet, but it's still a quite intrusive change. In particular, the commands for manipulating filters at runtime completely change. We delete the old code, and use the same infrastructure as for video filters. (This forces complete reinitialization of the filter chain, which hopefully isn't a problem for any use cases. The old code forced reinitialization too, but it could potentially allow a filter to cache things; e.g. consider loaded ladspa plugins and such.)
2013-07-22 14:43:58 +02:00
{ MP_CMD_AF, "af", { ARG_STRING, ARG_STRING } },
commands: add show_chapters_osd command to display chapters on OSD The command lists the chapters in the current file on the OSD. It also marks the current chapter. This is actually a cheap replacement for the chapter select libmenu functionality.
2011-10-23 05:26:30 +02:00
core: allow changing filter filters at runtime Add the "vf" command, which allows changing the video filter chain at runtime. For example, the 'y' key could be bound to toggle deinterlacing by adding 'y vf toggle yadif' to the input.conf. Reconfiguring the video filter chain normally resets the VO, so that it will be "stuck" until a new video frame is rendered. To mitigate this, a seek to the current position is issued when the filter chain is changed. This is done only if playback is paused, because normal playback will show an actual new frame quickly enough. If vdpau hardware decoding is used, filter insertion (whether it fails or not) will break the video for a while. This is because vo_vdpau resets decoding related things on vo_config().
2013-05-18 11:44:17 +02:00
{ MP_CMD_VF, "vf", { ARG_STRING, ARG_STRING } },
VO: add mechanisms to change VO commandline for VOs supporting it
2012-08-01 01:06:59 +02:00
{ MP_CMD_VO_CMDLINE, "vo_cmdline", { ARG_STRING } },
input.c: simplify command definitions Using type = -1 for terminating the command argument list was an unnecessary complication. Change it to 0 to make use of default initialization. Remove the explicit termination from the command definitions. Now the argument list is automatically terminated by C default initialization rules. Also remove other redundant {0} initializers. When an argument's default value is actually initialized to something other than 0, make the field being initialized explicit (e.g. {1} becomes {.i = 1}). Try to make use of whitespace more consistent.
2012-01-26 00:01:49 +01:00
{0}
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
};
commands: simplify legacy command-to-property bridge There are many input commands which are redundant to properties. They were parsed like normal commands, but set_property_command() in command.c handled them automatically using the property mechanism. This still required having the command specifications around, and the code in command.c was quite messy. Replace this with a text based replacement mechanism. Some corner cases are not handled: commands of form "seek_chapter 3 1" are supposed to set the "chapter" property to 3. This use is probably rare, and doesn't show up in the default input.conf. The reason compatibility is kept is because breaking input.conf is quite annoying, so a minimal effort is made to avoid this. Currently we print an annoying warning every time a legacy command is used, though. Also add a compatibility entry for "pt_step", which was removed some time ago. Variations in whitespace are not handled, but it's good enough to deal with old input.conf entries.
2012-09-06 07:19:41 +02:00
// Map legacy commands to proper commands
struct legacy_cmd {
const char *old, *new;
};
static const struct legacy_cmd legacy_cmds[] = {
commands: replace "switch" with "add" and "cycle" Now it depends on the command whether a property wraps around, or stops at min/max valid property value. For practically all properties, it's quite unambiguous what the "switch" command should have done, and there's technically no need to replace it with these new commands. More over, most properties that cycle are boolean anyway. But it seems more orthogonal to make the difference explicit, rather than hardcoding it. Having different commands also makes it more explicit to the user what these commands do, both just due to the naming, and what wrapping policy is used. The code is simpler too.
2012-09-22 06:15:36 +02:00
{"loop", "cycle loop"},
{"seek_chapter", "add chapter"},
{"switch_angle", "cycle angle"},
{"pause", "cycle pause"},
{"volume", "add volume"},
{"mute", "cycle mute"},
{"audio_delay", "add audio-delay"},
{"switch_audio", "cycle audio"},
{"balance", "add balance"},
input: make input command deprecation warnings visible Some time ago, all old special-cased commands (like "volume 1" to change volume by one) have been removed. These commands are still emulated using simple text replacement. This emulation is done to not break everyone's input.conf, especially because the input.conf provided by standard mplayer* still uses the old commands. Every use of a deprecated command prints a replacement warning, which was visible only with -v. Make these warnings visible by default. There's actually not much reason to do this, but since commands like "volume 5 1" don't work anymore, it's better to be verbose about this. Also simplify the replacement for "vo_fullscreen".
2013-03-25 20:49:11 +01:00
{"vo_fullscreen", "cycle fullscreen"},
commands: replace "switch" with "add" and "cycle" Now it depends on the command whether a property wraps around, or stops at min/max valid property value. For practically all properties, it's quite unambiguous what the "switch" command should have done, and there's technically no need to replace it with these new commands. More over, most properties that cycle are boolean anyway. But it seems more orthogonal to make the difference explicit, rather than hardcoding it. Having different commands also makes it more explicit to the user what these commands do, both just due to the naming, and what wrapping policy is used. The code is simpler too.
2012-09-22 06:15:36 +02:00
{"panscan", "add panscan"},
{"vo_ontop", "cycle ontop"},
{"vo_border", "cycle border"},
{"frame_drop", "cycle framedrop"},
{"gamma", "add gamma"},
{"brightness", "add brightness"},
{"contrast", "add contrast"},
{"saturation", "add saturation"},
{"hue", "add hue"},
{"switch_vsync", "cycle vsync"},
command: add sub_reload and sub_remove commands sub_remove remove an external subtitle track, for whatever this may be needed. sub_reload removes and re-adds an external subtitle track. Also rename sub_load to sub_add, because that seems to be more in line with sub_remove.
2012-11-15 20:26:52 +01:00
{"sub_load", "sub_add"},
commands: replace "switch" with "add" and "cycle" Now it depends on the command whether a property wraps around, or stops at min/max valid property value. For practically all properties, it's quite unambiguous what the "switch" command should have done, and there's technically no need to replace it with these new commands. More over, most properties that cycle are boolean anyway. But it seems more orthogonal to make the difference explicit, rather than hardcoding it. Having different commands also makes it more explicit to the user what these commands do, both just due to the naming, and what wrapping policy is used. The code is simpler too.
2012-09-22 06:15:36 +02:00
{"sub_select", "cycle sub"},
{"sub_pos", "add sub-pos"},
{"sub_delay", "add sub-delay"},
{"sub_visibility", "cycle sub-visibility"},
{"forced_subs_only", "cycle sub-forced-only"},
{"sub_scale", "add sub-scale"},
{"ass_use_margins", "cycle ass-use-margins"},
{"tv_set_brightness", "add tv-brightness"},
{"tv_set_hue", "add tv-hue"},
{"tv_set_saturation", "add tv-saturation"},
{"tv_set_contrast", "add tv-contrast"},
{"step_property_osd", "cycle"},
{"step_property", "no-osd cycle"},
commands: rename properties, update input.conf Use "-" instead of "_" in property names. The intent is that property names and options names should be the same (if they refer to the same thing), and options use "-" as word separator. Rename some other properties too, e.g. "switch_audio" -> "audio". Add a way to translate the old property names to the new ones, similar to the input command legacy bridge. Update input.conf. Use the new property names, and don't use legacy commands.
2012-09-15 01:10:59 +02:00
{"set_property", "no-osd set"},
{"set_property_osd", "set"},
commands: remove speed_set/speed_incr commands Redundant with set/switch commands.
2012-09-15 01:48:27 +02:00
{"speed_set", "set speed"},
commands: rename osd_show_[property_]text and osd_show_progression osd_show_[property_]text => show_text osd_show_progression => show_progress show_text, osd_show_property_text and osd_show_text both map to the code for the previous osd_show_property_text. The only special thing about osd_show_text is that you don't need to escape "$". Also, unfortunately osd_show_property_text requires escaping things twice, one time for the command parser, and the other time for the property formatting code, while osd_show_text needed only one level of escaping.
2012-09-15 02:04:57 +02:00
{"osd_show_text", "show_text"},
{"osd_show_property_text", "show_text"},
{"osd_show_progression", "show_progress"},
command: replace some show_ commands with properties show_chapters, show_tracks, and show_playlist are killed and replaced with the properties chapter-list, track-list, and playlist. The code and the output of these stays the same, this is just moving a lot of code around and reducing the number of properties. The "old" commands will still be supported for a while (to avoid making everyone angry), so handle them with the legacy layer. Add something to suppress printing the legacy warnings for these commands.
2013-05-15 02:17:47 +02:00
{"show_chapters_osd", "show_text ${chapter-list}"},
{"!show_chapters", "show_text ${chapter-list}"},
{"show_tracks_osd", "show_text ${track-list}"},
{"!show_tracks", "show_text ${track-list}"},
{"!show_playlist", "show_text ${playlist}"},
commands: make "aspect" property writeable, replaces "switch_ratio" Move the code for "switch_ratio" to the M_PROPERTY_SET case of the "aspect" property. The rules are exactly the same, e.g. setting a ratio smaller than 0.1 sets the pixel aspect ratio to 1:1. For now, we define that writing "0" sets the PAR to 1:1, and disallow -1 (possibly reserve it to reset to default aspect ratio).
2012-09-22 05:13:29 +02:00
// Approximate (can fail if user added additional whitespace)
commands: rename properties, update input.conf Use "-" instead of "_" in property names. The intent is that property names and options names should be the same (if they refer to the same thing), and options use "-" as word separator. Rename some other properties too, e.g. "switch_audio" -> "audio". Add a way to translate the old property names to the new ones, similar to the input command legacy bridge. Update input.conf. Use the new property names, and don't use legacy commands.
2012-09-15 01:10:59 +02:00
{"pt_step 1", "playlist_next"},
{"pt_step -1", "playlist_prev"},
commands: make "aspect" property writeable, replaces "switch_ratio" Move the code for "switch_ratio" to the M_PROPERTY_SET case of the "aspect" property. The rules are exactly the same, e.g. setting a ratio smaller than 0.1 sets the pixel aspect ratio to 1:1. For now, we define that writing "0" sets the PAR to 1:1, and disallow -1 (possibly reserve it to reset to default aspect ratio).
2012-09-22 05:13:29 +02:00
// Switch_ratio without argument resets aspect ratio
{"switch_ratio ", "set aspect "},
{"switch_ratio", "set aspect 0"},
commands: simplify legacy command-to-property bridge There are many input commands which are redundant to properties. They were parsed like normal commands, but set_property_command() in command.c handled them automatically using the property mechanism. This still required having the command specifications around, and the code in command.c was quite messy. Replace this with a text based replacement mechanism. Some corner cases are not handled: commands of form "seek_chapter 3 1" are supposed to set the "chapter" property to 3. This use is probably rare, and doesn't show up in the default input.conf. The reason compatibility is kept is because breaking input.conf is quite annoying, so a minimal effort is made to avoid this. Currently we print an annoying warning every time a legacy command is used, though. Also add a compatibility entry for "pt_step", which was removed some time ago. Variations in whitespace are not handled, but it's good enough to deal with old input.conf entries.
2012-09-06 07:19:41 +02:00
{0}
};
Output typos fixed. Spelling and grammar fixes in comments. Output Cosmetics git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@10399 b3059339-0415-0410-9bf9-f77b7e298cf2
2003-07-09 03:42:19 +02:00
/// The names of the keys as used in input.conf
A bug fix in the auto-repeat stuff + moved all printf to mp_msg git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@5199 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-03-19 14:30:16 +01:00
/// If you add some new keys, you also need to add them here
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
static const struct key_name key_names[] = {
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
{ ' ', "SPACE" },
Allow to add/modify keybind for '#' git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@22462 b3059339-0415-0410-9bf9-f77b7e298cf2
2007-03-05 11:15:55 +01:00
{ '#', "SHARP" },
Prefix keycode defines with MP_ Do this to reduce conflicts with <linux/input.h>, which contains some conflicting defines. This changes the meaning of MP_KEY_DOWN: KEY_DOWN is renamed to MP_KEY_DOWN (cursor down key) MP_KEY_DOWN is renamed to MP_KEY_STATE_DOWN (modifier for key down state)
2013-02-12 01:34:04 +01:00
{ MP_KEY_ENTER, "ENTER" },
{ MP_KEY_TAB, "TAB" },
{ MP_KEY_BACKSPACE, "BS" },
{ MP_KEY_DELETE, "DEL" },
{ MP_KEY_INSERT, "INS" },
{ MP_KEY_HOME, "HOME" },
{ MP_KEY_END, "END" },
{ MP_KEY_PAGE_UP, "PGUP" },
{ MP_KEY_PAGE_DOWN, "PGDWN" },
{ MP_KEY_ESC, "ESC" },
{ MP_KEY_PRINT, "PRINT" },
{ MP_KEY_RIGHT, "RIGHT" },
{ MP_KEY_LEFT, "LEFT" },
{ MP_KEY_DOWN, "DOWN" },
{ MP_KEY_UP, "UP" },
{ MP_KEY_F+1, "F1" },
{ MP_KEY_F+2, "F2" },
{ MP_KEY_F+3, "F3" },
{ MP_KEY_F+4, "F4" },
{ MP_KEY_F+5, "F5" },
{ MP_KEY_F+6, "F6" },
{ MP_KEY_F+7, "F7" },
{ MP_KEY_F+8, "F8" },
{ MP_KEY_F+9, "F9" },
{ MP_KEY_F+10, "F10" },
{ MP_KEY_F+11, "F11" },
{ MP_KEY_F+12, "F12" },
{ MP_KEY_KP0, "KP0" },
{ MP_KEY_KP1, "KP1" },
{ MP_KEY_KP2, "KP2" },
{ MP_KEY_KP3, "KP3" },
{ MP_KEY_KP4, "KP4" },
{ MP_KEY_KP5, "KP5" },
{ MP_KEY_KP6, "KP6" },
{ MP_KEY_KP7, "KP7" },
{ MP_KEY_KP8, "KP8" },
{ MP_KEY_KP9, "KP9" },
{ MP_KEY_KPDEL, "KP_DEL" },
{ MP_KEY_KPDEC, "KP_DEC" },
{ MP_KEY_KPINS, "KP_INS" },
{ MP_KEY_KPENTER, "KP_ENTER" },
{ MP_MOUSE_BTN0, "MOUSE_BTN0" },
{ MP_MOUSE_BTN1, "MOUSE_BTN1" },
{ MP_MOUSE_BTN2, "MOUSE_BTN2" },
{ MP_MOUSE_BTN3, "MOUSE_BTN3" },
{ MP_MOUSE_BTN4, "MOUSE_BTN4" },
{ MP_MOUSE_BTN5, "MOUSE_BTN5" },
{ MP_MOUSE_BTN6, "MOUSE_BTN6" },
{ MP_MOUSE_BTN7, "MOUSE_BTN7" },
{ MP_MOUSE_BTN8, "MOUSE_BTN8" },
{ MP_MOUSE_BTN9, "MOUSE_BTN9" },
{ MP_MOUSE_BTN10, "MOUSE_BTN10" },
{ MP_MOUSE_BTN11, "MOUSE_BTN11" },
{ MP_MOUSE_BTN12, "MOUSE_BTN12" },
{ MP_MOUSE_BTN13, "MOUSE_BTN13" },
{ MP_MOUSE_BTN14, "MOUSE_BTN14" },
{ MP_MOUSE_BTN15, "MOUSE_BTN15" },
{ MP_MOUSE_BTN16, "MOUSE_BTN16" },
{ MP_MOUSE_BTN17, "MOUSE_BTN17" },
{ MP_MOUSE_BTN18, "MOUSE_BTN18" },
{ MP_MOUSE_BTN19, "MOUSE_BTN19" },
{ MP_MOUSE_BTN0_DBL, "MOUSE_BTN0_DBL" },
{ MP_MOUSE_BTN1_DBL, "MOUSE_BTN1_DBL" },
{ MP_MOUSE_BTN2_DBL, "MOUSE_BTN2_DBL" },
{ MP_MOUSE_BTN3_DBL, "MOUSE_BTN3_DBL" },
{ MP_MOUSE_BTN4_DBL, "MOUSE_BTN4_DBL" },
{ MP_MOUSE_BTN5_DBL, "MOUSE_BTN5_DBL" },
{ MP_MOUSE_BTN6_DBL, "MOUSE_BTN6_DBL" },
{ MP_MOUSE_BTN7_DBL, "MOUSE_BTN7_DBL" },
{ MP_MOUSE_BTN8_DBL, "MOUSE_BTN8_DBL" },
{ MP_MOUSE_BTN9_DBL, "MOUSE_BTN9_DBL" },
{ MP_MOUSE_BTN10_DBL, "MOUSE_BTN10_DBL" },
{ MP_MOUSE_BTN11_DBL, "MOUSE_BTN11_DBL" },
{ MP_MOUSE_BTN12_DBL, "MOUSE_BTN12_DBL" },
{ MP_MOUSE_BTN13_DBL, "MOUSE_BTN13_DBL" },
{ MP_MOUSE_BTN14_DBL, "MOUSE_BTN14_DBL" },
{ MP_MOUSE_BTN15_DBL, "MOUSE_BTN15_DBL" },
{ MP_MOUSE_BTN16_DBL, "MOUSE_BTN16_DBL" },
{ MP_MOUSE_BTN17_DBL, "MOUSE_BTN17_DBL" },
{ MP_MOUSE_BTN18_DBL, "MOUSE_BTN18_DBL" },
{ MP_MOUSE_BTN19_DBL, "MOUSE_BTN19_DBL" },
{ MP_JOY_AXIS1_MINUS, "JOY_UP" },
{ MP_JOY_AXIS1_PLUS, "JOY_DOWN" },
{ MP_JOY_AXIS0_MINUS, "JOY_LEFT" },
{ MP_JOY_AXIS0_PLUS, "JOY_RIGHT" },
{ MP_JOY_AXIS0_PLUS, "JOY_AXIS0_PLUS" },
{ MP_JOY_AXIS0_MINUS, "JOY_AXIS0_MINUS" },
{ MP_JOY_AXIS1_PLUS, "JOY_AXIS1_PLUS" },
{ MP_JOY_AXIS1_MINUS, "JOY_AXIS1_MINUS" },
{ MP_JOY_AXIS2_PLUS, "JOY_AXIS2_PLUS" },
{ MP_JOY_AXIS2_MINUS, "JOY_AXIS2_MINUS" },
{ MP_JOY_AXIS3_PLUS, "JOY_AXIS3_PLUS" },
{ MP_JOY_AXIS3_MINUS, "JOY_AXIS3_MINUS" },
{ MP_JOY_AXIS4_PLUS, "JOY_AXIS4_PLUS" },
{ MP_JOY_AXIS4_MINUS, "JOY_AXIS4_MINUS" },
{ MP_JOY_AXIS5_PLUS, "JOY_AXIS5_PLUS" },
{ MP_JOY_AXIS5_MINUS, "JOY_AXIS5_MINUS" },
{ MP_JOY_AXIS6_PLUS, "JOY_AXIS6_PLUS" },
{ MP_JOY_AXIS6_MINUS, "JOY_AXIS6_MINUS" },
{ MP_JOY_AXIS7_PLUS, "JOY_AXIS7_PLUS" },
{ MP_JOY_AXIS7_MINUS, "JOY_AXIS7_MINUS" },
{ MP_JOY_AXIS8_PLUS, "JOY_AXIS8_PLUS" },
{ MP_JOY_AXIS8_MINUS, "JOY_AXIS8_MINUS" },
{ MP_JOY_AXIS9_PLUS, "JOY_AXIS9_PLUS" },
{ MP_JOY_AXIS9_MINUS, "JOY_AXIS9_MINUS" },
{ MP_JOY_BTN0, "JOY_BTN0" },
{ MP_JOY_BTN1, "JOY_BTN1" },
{ MP_JOY_BTN2, "JOY_BTN2" },
{ MP_JOY_BTN3, "JOY_BTN3" },
{ MP_JOY_BTN4, "JOY_BTN4" },
{ MP_JOY_BTN5, "JOY_BTN5" },
{ MP_JOY_BTN6, "JOY_BTN6" },
{ MP_JOY_BTN7, "JOY_BTN7" },
{ MP_JOY_BTN8, "JOY_BTN8" },
{ MP_JOY_BTN9, "JOY_BTN9" },
osx: add Apple Remote support After killing the non functional AR support in c8fd9e5 I got much complaints so this adds AR support back in (and it works). I am using the HIDRemote class by Felix Schwarz and that part of the code is under the BSD license. I slightly modified it replacing [NSApplication sharedApplication] with NSApp. The code of the class is quite complex (probably because it had to deal with all the edge cases with IOKit) but it works nicely as a black box. In a later commit I'll remove the deprecation warnings caused by HIDRemote's usage of Gestalt. Check out `etc/input.conf` for the default bindings. Apple Remote functionality is automatically compiled in when cocoa is enabled. It can be disabled at runtime with the `--no-ar` option.
2013-06-03 00:52:40 +02:00
{ MP_AR_PLAY, "AR_PLAY" },
{ MP_AR_PLAY_HOLD, "AR_PLAY_HOLD" },
{ MP_AR_CENTER, "AR_CENTER" },
{ MP_AR_CENTER_HOLD, "AR_CENTER_HOLD" },
{ MP_AR_NEXT, "AR_NEXT" },
{ MP_AR_NEXT_HOLD, "AR_NEXT_HOLD" },
{ MP_AR_PREV, "AR_PREV" },
{ MP_AR_PREV_HOLD, "AR_PREV_HOLD" },
{ MP_AR_MENU, "AR_MENU" },
{ MP_AR_MENU_HOLD, "AR_MENU_HOLD" },
{ MP_AR_VUP, "AR_VUP" },
{ MP_AR_VUP_HOLD, "AR_VUP_HOLD" },
{ MP_AR_VDOWN, "AR_VDOWN" },
{ MP_AR_VDOWN_HOLD, "AR_VDOWN_HOLD" },
osx: improve Media Keys support This commit addresses some issues with the users had with the previous implementation in commit c39efb9. Here's the changes: * Use Quartz Event Taps to remove Media Key events mpv handles from the global OS X queue. This prevents conflicts with iTunes. I did this on the main thread since it is mostly idling. It's the playloop thread that actually does all the work so there is no danger of blocking the event tap callback. * Introduce `--no-media-keys` switch so that users can disable all of mpv's media key handling at runtime (some prefer iTunes for example). * Use mpv's bindings so that users can customize what the media keys do via input.conf. Current bindings are: MK_PLAY cycle pause MK_PREV playlist_prev MK_NEXT playlist_next An additional benefit of this implementation is that it is completly handled by the `macosx_events` file instead of `macosx_application` making the project organization more straightforward.
2013-06-04 22:12:23 +02:00
{ MP_MK_PLAY, "MK_PLAY" },
{ MP_MK_PREV, "MK_PREV" },
{ MP_MK_NEXT, "MK_NEXT" },
input: add support for precise scroll axes Support horizontal and vertical axes of input devices. If the input device support precise scrolling with an input value then it should first be scaled to a standard multiplier, where 1.0 is the default. The multiplier will then applied to the following commands if possible: * MP_CMD_SEEK * MP_CMD_SPEED_MULT * MP_CMD_ADD All other commands will triggered on every axis event, without change the values specified in the config file.
2013-07-25 18:08:57 +02:00
{ MP_AXIS_UP, "AXIS_UP" },
{ MP_AXIS_DOWN, "AXIS_DOWN" },
{ MP_AXIS_LEFT, "AXIS_LEFT" },
{ MP_AXIS_RIGHT, "AXIS_RIGHT" },
Prefix keycode defines with MP_ Do this to reduce conflicts with <linux/input.h>, which contains some conflicting defines. This changes the meaning of MP_KEY_DOWN: KEY_DOWN is renamed to MP_KEY_DOWN (cursor down key) MP_KEY_DOWN is renamed to MP_KEY_STATE_DOWN (modifier for key down state)
2013-02-12 01:34:04 +01:00
{ MP_KEY_POWER, "POWER" },
{ MP_KEY_MENU, "MENU" },
{ MP_KEY_PLAY, "PLAY" },
{ MP_KEY_PAUSE, "PAUSE" },
{ MP_KEY_PLAYPAUSE, "PLAYPAUSE" },
{ MP_KEY_STOP, "STOP" },
{ MP_KEY_FORWARD, "FORWARD" },
{ MP_KEY_REWIND, "REWIND" },
{ MP_KEY_NEXT, "NEXT" },
{ MP_KEY_PREV, "PREV" },
{ MP_KEY_VOLUME_UP, "VOLUME_UP" },
{ MP_KEY_VOLUME_DOWN, "VOLUME_DOWN" },
{ MP_KEY_MUTE, "MUTE" },
input: add some more X11 multimedia/internet keys These keys can be found on various "multimedia" and "internet" keyboard. X defines many keycodes, so I'm not adding all, just what I found on my own keyboard. Other key codes can be added on request.
2013-09-01 19:59:38 +02:00
{ MP_KEY_HOMEPAGE, "HOMEPAGE" },
{ MP_KEY_WWW, "WWW" },
{ MP_KEY_MAIL, "MAIL" },
{ MP_KEY_FAVORITES, "FAVORITES" },
{ MP_KEY_SEARCH, "SEARCH" },
{ MP_KEY_SLEEP, "SLEEP" },
libvo input cleanup: remove the dependency on libinput, remove most of the crappy mappings (like O->o or ESC->q). git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@16893 b3059339-0415-0410-9bf9-f77b7e298cf2
2005-11-02 22:50:55 +01:00
// These are kept for backward compatibility
Prefix keycode defines with MP_ Do this to reduce conflicts with <linux/input.h>, which contains some conflicting defines. This changes the meaning of MP_KEY_DOWN: KEY_DOWN is renamed to MP_KEY_DOWN (cursor down key) MP_KEY_DOWN is renamed to MP_KEY_STATE_DOWN (modifier for key down state)
2013-02-12 01:34:04 +01:00
{ MP_KEY_PAUSE, "XF86_PAUSE" },
{ MP_KEY_STOP, "XF86_STOP" },
{ MP_KEY_PREV, "XF86_PREV" },
{ MP_KEY_NEXT, "XF86_NEXT" },
libvo input cleanup: remove the dependency on libinput, remove most of the crappy mappings (like O->o or ESC->q). git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@16893 b3059339-0415-0410-9bf9-f77b7e298cf2
2005-11-02 22:50:55 +01:00
input: trigger mouse_leave key bindings if mouse leaves mouse area Also, implement mouse leave events for X11. But evne on other platforms, these events will be generated if mouse crosses a section's mouse area boundaries within the mpv window.
2013-06-19 18:19:45 +02:00
{ MP_KEY_CLOSE_WIN, "CLOSE_WIN" },
{ MP_KEY_MOUSE_MOVE, "MOUSE_MOVE" },
{ MP_KEY_MOUSE_LEAVE, "MOUSE_LEAVE" },
Support for "internet" or "multimedia" keyboards in X11. Patch from Jason Tackaberry <tack at auc dot ca>. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@9823 b3059339-0415-0410-9bf9-f77b7e298cf2
2003-04-04 17:04:26 +02:00
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
{ 0, NULL }
};
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
struct key_name modifier_names[] = {
Prefix keycode defines with MP_ Do this to reduce conflicts with <linux/input.h>, which contains some conflicting defines. This changes the meaning of MP_KEY_DOWN: KEY_DOWN is renamed to MP_KEY_DOWN (cursor down key) MP_KEY_DOWN is renamed to MP_KEY_STATE_DOWN (modifier for key down state)
2013-02-12 01:34:04 +01:00
{ MP_KEY_MODIFIER_SHIFT, "Shift" },
{ MP_KEY_MODIFIER_CTRL, "Ctrl" },
{ MP_KEY_MODIFIER_ALT, "Alt" },
{ MP_KEY_MODIFIER_META, "Meta" },
input: support bindings with modifier keys for X input Add support for binding commands to modifier+key combinations like "Shift+Left" or "Ctrl+Alt+x", and support reading such combinations from the output window of X VOs. The recognized modifier names are Shift, Ctrl, Alt and Meta. Any combination of those and then a non-modifier key name, separated by '+', is accepted as a key name in input.conf. For non-special keys that produce characters shift is ignored as a modifier. For example "A" is handled as a key without modifiers even if you use shift to write the capital letter; 'a' vs 'A' already distinguishes the combinations with a normal keymap, and having separate 'a', 'Shift+A' and 'A' (written with caps lock for example) would bring more confusion than benefit. Currently reading the modifier+key combinations is only supported in the output window of those VOs that use x11_common.c event handling. It's not possible to input the key combinations in other VOs or in a terminal window.
2010-12-19 11:12:20 +01:00
{ 0 }
};
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
#define MP_MAX_FDS 10
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
struct input_fd {
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
int fd;
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
int (*read_key)(void *ctx, int fd);
int (*read_cmd)(int fd, char *dest, int size);
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
int (*close_func)(int fd);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
void *ctx;
unsigned eof : 1;
unsigned drop : 1;
unsigned dead : 1;
unsigned got_cmd : 1;
input: invert negated boolean field
2013-07-14 15:37:39 +02:00
unsigned select : 1;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
// These fields are for the cmd fds.
char *buffer;
int pos, size;
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
};
Add the possibilty to grab the keys and to filter the commands Add the commands for the osd menu git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@8197 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-11-15 00:41:44 +01:00
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
struct cmd_bind_section {
input: store number of binds, instead of using 0-termination
2013-06-23 00:40:35 +02:00
struct cmd_bind *binds;
int num_binds;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
char *section;
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
struct mp_rect mouse_area; // set at runtime, if at all
bool mouse_area_set; // mouse_area is valid and should be tested
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
struct cmd_bind_section *next;
implement "bind sections" allows to bind one input key to several slave commands (independently for each section) as shown below: RIGHT seek +10 RIGHT {tv} tv_step_channel 1 RIGHT {dvdnav} dvdnav 4 Currenlty only "tv" section added. patch by Otvos Attila oattila at chello dot hu git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@23497 b3059339-0415-0410-9bf9-f77b7e298cf2
2007-06-07 20:06:53 +02:00
};
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
#define MAX_ACTIVE_SECTIONS 5
struct active_section {
char *name;
int flags;
};
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
struct cmd_queue {
struct mp_cmd *first;
};
Create a context for input.c state Start moving static variables to a context struct. Only autorepeat state is moved to the struct in this commit. mp_input_check_interrupt now requires the context variable. Change stream functions to pass it. It's still stored in a static variable in stream/.
2008-04-30 06:15:52 +02:00
struct input_ctx {
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
#if HAVE_PTHREADS
pthread_mutex_t mutex;
#endif
m_config: refactor some things Change how m_config is initialized. Make it more uniform; now all m_config structs are intialized in exactly the same way. Make sure there's only a single m_option[] array defining the options, and keep around the pointer to the optstruct default value, and the optstruct size as well. This will allow reconstructing the option default values in the following commit. In particular, stop pretending that the handling of some special options (like --profile, --v, and some others) is in any way elegant, and make them explicit hacks. This is really more readable and easier to understand than what was before, and simplifies the code.
2013-07-27 21:26:00 +02:00
bool using_ar;
bool using_cocoa_media_keys;
Create a context for input.c state Start moving static variables to a context struct. Only autorepeat state is moved to the struct in this commit. mp_input_check_interrupt now requires the context variable. Change stream functions to pass it. It's still stored in a static variable in stream/.
2008-04-30 06:15:52 +02:00
// Autorepeat stuff
short ar_state;
Replace all calls to GetTimer()/GetTimerMS() GetTimer() is generally replaced with mp_time_us(). Both calls return microseconds, but the latter uses int64_t, us defined to never wrap, and never returns 0 or negative values. GetTimerMS() has no direct replacement. Instead the other functions are used. For some code, switch to mp_time_sec(), which returns the time as double float value in seconds. The returned time is offset to program start time, so there is enough precision left to deliver microsecond resolution for at least 100 years. Unless it's casted to a float (or the CPU reduces precision), which is why we still use mp_time_us() out of paranoia in places where precision is clearly needed. Always switch to the correct time. The whole point of the new timer calls is that they don't wrap, and storing microseconds in unsigned int variables would negate this. In some cases, remove wrap-around handling for time values.
2013-05-25 18:31:06 +02:00
int64_t last_ar;
Move input options to option struct
2008-04-30 17:57:02 +02:00
// Autorepeat config
unsigned int ar_delay;
unsigned int ar_rate;
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
// Maximum number of queued commands from keypresses (limit to avoid
// repeated slow commands piling up)
int key_fifo_size;
input: Move command bind variables to context struct
2008-04-30 07:56:45 +02:00
input.c: Move key down state to context struct
2008-04-30 09:21:50 +02:00
// these are the keys currently down
int key_down[MP_MAX_KEY_DOWN];
unsigned int num_key_down;
Replace all calls to GetTimer()/GetTimerMS() GetTimer() is generally replaced with mp_time_us(). Both calls return microseconds, but the latter uses int64_t, us defined to never wrap, and never returns 0 or negative values. GetTimerMS() has no direct replacement. Instead the other functions are used. For some code, switch to mp_time_sec(), which returns the time as double float value in seconds. The returned time is offset to program start time, so there is enough precision left to deliver microsecond resolution for at least 100 years. Unless it's casted to a float (or the CPU reduces precision), which is why we still use mp_time_us() out of paranoia in places where precision is clearly needed. Always switch to the correct time. The whole point of the new timer calls is that they don't wrap, and storing microseconds in unsigned int variables would negate this. In some cases, remove wrap-around handling for time values.
2013-05-25 18:31:06 +02:00
int64_t last_key_down;
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
struct mp_cmd *current_down_cmd;
core: remove mp_fifo indirection For some reason mp_fifo specifically handled double clicks, and other than that was a pointless wrapper around input.c functionality. Move the double click handling into input.c, and get rid of mp_fifo. Add some compatibility wrappers, because so much VO code uses these functions. Where struct mp_fifo is still used it's just a casted struct input_ctx.
2013-07-02 14:00:24 +02:00
int doubleclick_time;
int last_doubleclick_key_down;
double last_doubleclick_time;
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
// Mouse position on the consumer side (as command.c sees it)
int mouse_x, mouse_y;
input: trigger mouse_leave key bindings if mouse leaves mouse area Also, implement mouse leave events for X11. But evne on other platforms, these events will be generated if mouse crosses a section's mouse area boundaries within the mpv window.
2013-06-19 18:19:45 +02:00
char *mouse_section; // last section to receive mouse event
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
// Mouse position on the producer side (as the VO sees it)
// Unlike mouse_x/y, this can be used to resolve mouse click bindings.
int mouse_vo_x, mouse_vo_y;
input.c: Move key down state to context struct
2008-04-30 09:21:50 +02:00
input: add input test mode In input test mode, key bindings won't be executed, but are shown on the OSD. The OSD includes various information, such as the name of the key, the command itself, whether it's builtin, and the config file location it was defined. The input test mode can be enabled with "--input=test". No effort is spent trying to react to key bindings that normally exit the player; they are treated just like any other binding.
2012-10-13 21:10:20 +02:00
bool test;
Merge svn changes up to r29117
2009-04-01 01:26:34 +02:00
bool default_bindings;
input: Move command bind variables to context struct
2008-04-30 07:56:45 +02:00
// List of command binding sections
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
struct cmd_bind_section *cmd_bind_sections;
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
// List currently active command sections
struct active_section active_sections[MAX_ACTIVE_SECTIONS];
int num_active_sections;
input: Move key filedescriptor list to context struct The context variable is now passed to VOs too as it's now needed to add a callback on the X filedescriptor.
2008-04-30 10:06:55 +02:00
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
// Used to track whether we managed to read something while checking
// events sources. If yes, the sources may have more queued.
bool got_new_events;
Replace all calls to GetTimer()/GetTimerMS() GetTimer() is generally replaced with mp_time_us(). Both calls return microseconds, but the latter uses int64_t, us defined to never wrap, and never returns 0 or negative values. GetTimerMS() has no direct replacement. Instead the other functions are used. For some code, switch to mp_time_sec(), which returns the time as double float value in seconds. The returned time is offset to program start time, so there is enough precision left to deliver microsecond resolution for at least 100 years. Unless it's casted to a float (or the CPU reduces precision), which is why we still use mp_time_us() out of paranoia in places where precision is clearly needed. Always switch to the correct time. The whole point of the new timer calls is that they don't wrap, and storing microseconds in unsigned int variables would negate this. In some cases, remove wrap-around handling for time values.
2013-05-25 18:31:06 +02:00
unsigned int mouse_event_counter;
core: do mouse cursor hiding business in frontend Do this so that not every VO backend has to setup a timer for cursor hiding and interpret the --cursor-autohide option.
2013-05-16 23:17:46 +02:00
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
struct input_fd fds[MP_MAX_FDS];
unsigned int num_fds;
input: Move command filedescriptor list to context struct
2008-04-30 12:14:03 +02:00
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
struct cmd_queue key_cmd_queue;
struct cmd_queue control_cmd_queue;
ao_pulse, core: make pulse thread wake up core for more data For ao_pulse, the current latency is not a good indicator of how soon the AO requires new data to avoid underflow. Add an internal pipe that can be used to wake up the input loop from select(), and make the pulseaudio main loop (which runs in a separate thread) use this mechanism when pulse requests more data. The wakeup signal currently contains no information about the reason for the wakup, but audio buffers are always filled when the event loop wakes up. Also, request a latency of 1 second from the Pulseaudio server. The default is normally significantly higher. We don't need low latency, while higher latency helps prevent underflows reduces need for wakeups.
2012-03-25 21:58:48 +02:00
int wakeup_pipe[2];
Create a context for input.c state Start moving static variables to a context struct. Only autorepeat state is moved to the struct in this commit. mp_input_check_interrupt now requires the context variable. Change stream functions to pass it. It's still stored in a static variable in stream/.
2008-04-30 06:15:52 +02:00
};
Do not use exit_player in the signal handler, this code just can not be called from a signal handler. Instead only make the input system generate quit commands for the first CTRL+C and otherwise do getch2_disable and exit. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@25651 b3059339-0415-0410-9bf9-f77b7e298cf2
2008-01-08 22:40:44 +01:00
int async_quit_request;
options: change --input=keylist, cmdlist implementation Change the --input=keylist and --input=cmdlist suboptions to use the "print function" option type. This changes their semantics somewhat, and now some other output can appear after the printed lists (before, they called "exit(0)" directly). I'm not aware of any program parsing the output which could be affected.
2012-05-07 22:16:20 +02:00
static int print_key_list(m_option_t *cfg, char *optname, char *optparam);
static int print_cmd_list(m_option_t *cfg, char *optname, char *optparam);
Added options -input keylist and -input cmdlist to list all know keys and commands git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@5219 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-03-20 14:30:57 +01:00
input: stay in current input section if mouse button is down Normally, moving the mouse outside of the mouse area of an input section will send mouse events somewhere else (because input section mouse areas are similar to windows/widgets in real GUI toolkits). This was done even if a mouse button was held down. This is quite different from how GUI toolkits behave. Change the code so that if a mouse button is down, the mouse area of the current input section can't be left. Releasing the mouse button (while the mouse pointer is outside of the mouse area) will actually leave the mouse area. As a side-effect, this commit also tests more often whether the current mouse input section is valid. This is needed to make releasing a mouse button trigger the mouse input section change.
2013-08-30 02:24:14 +02:00
static void add_key_cmd(struct input_ctx *ictx, struct mp_cmd *cmd);
m_option: don't define OPT_BASE_STRUCT by default OPT_BASE_STRUCT defines which struct the OPT_ macros (like OPT_INT etc.) reference implicitly, since these macros take struct member names but no struct type. Normally, only cfg-mplayer.h should need this, and other places shouldn't be bothered with having to #undef it. (Some files, like demux_lavf.c, still store their options in MPOpts. In the long term, this should be removed, and handled like e.g. with VO suboptions instead.)
2013-03-01 11:27:59 +01:00
#define OPT_BASE_STRUCT struct MPOpts
A bug fix in the auto-repeat stuff + moved all printf to mp_msg git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@5199 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-03-19 14:30:16 +01:00
// Our command line options
Fix some -Wshadow warnings In general, this warning can hint to actual bugs. We don't enable it yet, because it would conflict with some unmerged code, and we should check with clang too (this commit was done by testing with gcc).
2013-07-23 00:45:23 +02:00
static const m_option_t input_config[] = {
options: change --no-config option, make it apply to input.conf as well Simplify --no-config and make it a normal flag option, and doesn't take an argument anymore. You can get the same behavior by using --no-config and then --include to explicitly load a certain config file. Make --no-config work for input.conf as well. Make it so that --input:conf=file still works in this case. As a technically unrelated change, the file argument now works as one would expect, instead of making it relatively to "~/.mpv/". This makes for simpler code and easier to understand option semantics. We can also print better error messages.
2013-02-08 23:52:06 +01:00
OPT_STRING("conf", input.config_file, CONF_GLOBAL),
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
OPT_INT("ar-delay", input.ar_delay, CONF_GLOBAL),
Move input options to option struct
2008-04-30 17:57:02 +02:00
OPT_INT("ar-rate", input.ar_rate, CONF_GLOBAL),
options: get rid of ambiguous option parsing Options parsing used to be ambiguous, as in the splitting into option and values pairs was ambiguous. Example: -option -something It wasn't clear whether -option actually takes an argument or not. The string "-something" could either be a separate option, or an argument to "-option". The code had to call the option specific parser function to resolve this. This made everything complicated and didn't even have a real use. There was only one case where this was actually used: string lists (m_option_type_string_list) and options based on it. That is because this option type actually turns a single option into a proxy for several real arguments, e.g. "vf*" can handle "-vf-add" and "-vf-clr". Options suffixed with "-clr" are the only options of this group which take no arguments. This is ambiguous only with the "old syntax" (as shown above). The "new" option syntax always puts option name and value into same argument. (E.g. "--option=--something" or "--option" "--something".) Simplify the code by making it statically known whether an option takes a parameter or not with the flag M_OPT_TYPE_OLD_SYNTAX_NO_PARAM. If it's set, the option parser assumes the option takes no argument. The only real ambiguity left, string list options that end on "-clr", are special cased in the parser. Remove some duplication of the logic in the command line parser by moving all argument splitting logic into split_opt(). (It's arguable whether that can be considered code duplication, but now the code is a bit simpler anyway. This might be subjective.) Remove the "ambiguous" parameter from all option parsing related code. Make m_config unaware of the pre-parsing concept. Make most CONF_NOCFG options also CONF_GLOBAL (except those explicitly usable as per-file options.)
2012-08-05 23:34:28 +02:00
{ "keylist", print_key_list, CONF_TYPE_PRINT_FUNC, CONF_GLOBAL | CONF_NOCFG },
{ "cmdlist", print_cmd_list, CONF_TYPE_PRINT_FUNC, CONF_GLOBAL | CONF_NOCFG },
Move input options to option struct
2008-04-30 17:57:02 +02:00
OPT_STRING("js-dev", input.js_dev, CONF_GLOBAL),
OPT_STRING("file", input.in_file, CONF_GLOBAL),
cleanup: replace OPT_FLAG_ON and OPT_MAKE_FLAGS with OPT_FLAG OPT_MAKE_FLAGS() used to emit two options (one with "no" prefixed), but that has been long removed by special casing flag options in the option parser. OPT_FLAG_ON() used to imply that there's no "no-" prefixed option, but this hasn't been the case for a while either. (Conceptually, it has been replaced by OPT_FLAG_STORE().) Remove OPT_FLAG_OFF, which was unused.
2013-02-08 21:09:18 +01:00
OPT_FLAG("default-bindings", input.default_bindings, CONF_GLOBAL),
OPT_FLAG("test", input.test, CONF_GLOBAL),
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
{ NULL, NULL, 0, 0, 0, 0, NULL}
Added key autorepeat support. Options to enable/disable joystick and lirc and set autorepeat delay and rate. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4658 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-11 12:42:08 +01:00
};
m_config: refactor some things Change how m_config is initialized. Make it more uniform; now all m_config structs are intialized in exactly the same way. Make sure there's only a single m_option[] array defining the options, and keep around the pointer to the optstruct default value, and the optstruct size as well. This will allow reconstructing the option default values in the following commit. In particular, stop pretending that the handling of some special options (like --profile, --v, and some others) is in any way elegant, and make them explicit hacks. This is really more readable and easier to understand than what was before, and simplifies the code.
2013-07-27 21:26:00 +02:00
const m_option_t mp_input_opts[] = {
Fix some -Wshadow warnings In general, this warning can hint to actual bugs. We don't enable it yet, because it would conflict with some unmerged code, and we should check with clang too (this commit was done by testing with gcc).
2013-07-23 00:45:23 +02:00
{ "input", (void *)&input_config, CONF_TYPE_SUBCONFIG, 0, 0, 0, NULL},
core: remove mp_fifo indirection For some reason mp_fifo specifically handled double clicks, and other than that was a pointless wrapper around input.c functionality. Move the double click handling into input.c, and get rid of mp_fifo. Add some compatibility wrappers, because so much VO code uses these functions. Where struct mp_fifo is still used it's just a casted struct input_ctx.
2013-07-02 14:00:24 +02:00
OPT_INTRANGE("doubleclick-time", input.doubleclick_time, 0, 0, 1000),
cleanup: replace OPT_FLAG_ON and OPT_MAKE_FLAGS with OPT_FLAG OPT_MAKE_FLAGS() used to emit two options (one with "no" prefixed), but that has been long removed by special casing flag options in the option parser. OPT_FLAG_ON() used to imply that there's no "no-" prefixed option, but this hasn't been the case for a while either. (Conceptually, it has been replaced by OPT_FLAG_STORE().) Remove OPT_FLAG_OFF, which was unused.
2013-02-08 21:09:18 +01:00
OPT_FLAG("joystick", input.use_joystick, CONF_GLOBAL),
OPT_FLAG("lirc", input.use_lirc, CONF_GLOBAL),
OPT_FLAG("lircc", input.use_lircc, CONF_GLOBAL),
fix build on linux missing ifdef in the previous commit made the build fail
2013-06-03 23:02:16 +02:00
#ifdef CONFIG_COCOA
osx: add Apple Remote support After killing the non functional AR support in c8fd9e5 I got much complaints so this adds AR support back in (and it works). I am using the HIDRemote class by Felix Schwarz and that part of the code is under the BSD license. I slightly modified it replacing [NSApplication sharedApplication] with NSApp. The code of the class is quite complex (probably because it had to deal with all the edge cases with IOKit) but it works nicely as a black box. In a later commit I'll remove the deprecation warnings caused by HIDRemote's usage of Gestalt. Check out `etc/input.conf` for the default bindings. Apple Remote functionality is automatically compiled in when cocoa is enabled. It can be disabled at runtime with the `--no-ar` option.
2013-06-03 00:52:40 +02:00
OPT_FLAG("ar", input.use_ar, CONF_GLOBAL),
osx: improve Media Keys support This commit addresses some issues with the users had with the previous implementation in commit c39efb9. Here's the changes: * Use Quartz Event Taps to remove Media Key events mpv handles from the global OS X queue. This prevents conflicts with iTunes. I did this on the main thread since it is mostly idling. It's the playloop thread that actually does all the work so there is no danger of blocking the event tap callback. * Introduce `--no-media-keys` switch so that users can disable all of mpv's media key handling at runtime (some prefer iTunes for example). * Use mpv's bindings so that users can customize what the media keys do via input.conf. Current bindings are: MK_PLAY cycle pause MK_PREV playlist_prev MK_NEXT playlist_next An additional benefit of this implementation is that it is completly handled by the `macosx_events` file instead of `macosx_application` making the project organization more straightforward.
2013-06-04 22:12:23 +02:00
OPT_FLAG("media-keys", input.use_media_keys, CONF_GLOBAL),
fix build on linux missing ifdef in the previous commit made the build fail
2013-06-03 23:02:16 +02:00
#endif
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
{ NULL, NULL, 0, 0, 0, 0, NULL}
Added key autorepeat support. Options to enable/disable joystick and lirc and set autorepeat delay and rate. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4658 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-11 12:42:08 +01:00
};
Added support for key combination and mouse buttons key code git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4590 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-08 19:10:56 +01:00
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
static int default_cmd_func(int fd, char *buf, int l);
Reverse Arpi's commit and put the right fix in place. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@7913 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-10-25 18:06:25 +02:00
input: replace internal key binds with included etc/input.conf The internal array of default key bindings is removed. Include the file etc/input.conf at compile time (using the file2header tool), and parse the default binds from etc/input.conf at startup time. This lowers maintainance overhead, and makes sure the default bindings and etc/input.conf don't deviate. Commit f30bf73bf22ed0542 already made sure etc/input.conf matches the default bindings, so this commit shouldn't change anything user-visible.
2012-08-01 22:52:28 +02:00
static const char builtin_input_conf[] =
core: move contents to mpvcore (2/2) Followup commit. Fixes all the files references.
2013-08-06 22:41:30 +02:00
#include "mpvcore/input/input_conf.h"
input: replace internal key binds with included etc/input.conf The internal array of default key bindings is removed. Include the file etc/input.conf at compile time (using the file2header tool), and parse the default binds from etc/input.conf at startup time. This lowers maintainance overhead, and makes sure the default bindings and etc/input.conf don't deviate. Commit f30bf73bf22ed0542 already made sure etc/input.conf matches the default bindings, so this commit shouldn't change anything user-visible.
2012-08-01 22:52:28 +02:00
;
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
static bool test_rect(struct mp_rect *rc, int x, int y)
{
return x >= rc->x0 && y >= rc->y0 && x < rc->x1 && y < rc->y1;
}
input: clean up messages with button combination names Previously messages that printed key/button names would have extra names in button combinations appended after the main message, resulting in output like: No bind found for key 'MOUSE_BTN1'.-MOUSE_BTN3 Add a function that creates a complete combination name and use that for all such messages. The above example changes to: No bind found for key 'MOUSE_BTN1-MOUSE_BTN3'. Other affected messages are a input.conf parsing error message and a message about a bound command being invalid.
2011-06-29 07:49:01 +02:00
static char *get_key_name(int key, char *ret)
input: support bindings with modifier keys for X input Add support for binding commands to modifier+key combinations like "Shift+Left" or "Ctrl+Alt+x", and support reading such combinations from the output window of X VOs. The recognized modifier names are Shift, Ctrl, Alt and Meta. Any combination of those and then a non-modifier key name, separated by '+', is accepted as a key name in input.conf. For non-special keys that produce characters shift is ignored as a modifier. For example "A" is handled as a key without modifiers even if you use shift to write the capital letter; 'a' vs 'A' already distinguishes the combinations with a normal keymap, and having separate 'a', 'Shift+A' and 'A' (written with caps lock for example) would bring more confusion than benefit. Currently reading the modifier+key combinations is only supported in the output window of those VOs that use x11_common.c event handling. It's not possible to input the key combinations in other VOs or in a terminal window.
2010-12-19 11:12:20 +01:00
{
for (int i = 0; modifier_names[i].name; i++) {
if (modifier_names[i].key & key) {
ret = talloc_asprintf_append_buffer(ret, "%s+",
modifier_names[i].name);
key -= modifier_names[i].key;
}
}
for (int i = 0; key_names[i].name != NULL; i++) {
if (key_names[i].key == key)
return talloc_asprintf_append_buffer(ret, "%s", key_names[i].name);
}
input: allow unicode keys and reassign internal key codes This moves all key codes above the highest valid unicode code point (which is 0x10FFFF). All key codes below MP_KEY_BASE now directly map to unicode. Configuration files (input.conf) can contain unicode characters in UTF-8 to map non-ASCII characters/keys. This shouldn't change anything user visible, except that "direct key codes" (as used in input.conf) will change their meaning.
2012-01-14 14:09:26 +01:00
// printable, and valid unicode range
if (key >= 32 && key <= 0x10FFFF)
cleanup: remove duplicated function, move escape parsing function
2013-02-16 22:51:10 +01:00
return mp_append_utf8_buffer(ret, key);
Added key autorepeat support. Options to enable/disable joystick and lirc and set autorepeat delay and rate. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4658 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-11 12:42:08 +01:00
input: support bindings with modifier keys for X input Add support for binding commands to modifier+key combinations like "Shift+Left" or "Ctrl+Alt+x", and support reading such combinations from the output window of X VOs. The recognized modifier names are Shift, Ctrl, Alt and Meta. Any combination of those and then a non-modifier key name, separated by '+', is accepted as a key name in input.conf. For non-special keys that produce characters shift is ignored as a modifier. For example "A" is handled as a key without modifiers even if you use shift to write the capital letter; 'a' vs 'A' already distinguishes the combinations with a normal keymap, and having separate 'a', 'Shift+A' and 'A' (written with caps lock for example) would bring more confusion than benefit. Currently reading the modifier+key combinations is only supported in the output window of those VOs that use x11_common.c event handling. It's not possible to input the key combinations in other VOs or in a terminal window.
2010-12-19 11:12:20 +01:00
// Print the hex key code
return talloc_asprintf_append_buffer(ret, "%#-8x", key);
}
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
input: clean up messages with button combination names Previously messages that printed key/button names would have extra names in button combinations appended after the main message, resulting in output like: No bind found for key 'MOUSE_BTN1'.-MOUSE_BTN3 Add a function that creates a complete combination name and use that for all such messages. The above example changes to: No bind found for key 'MOUSE_BTN1-MOUSE_BTN3'. Other affected messages are a input.conf parsing error message and a message about a bound command being invalid.
2011-06-29 07:49:01 +02:00
static char *get_key_combo_name(int *keys, int max)
{
char *ret = talloc_strdup(NULL, "");
input: require VOs to send key up events, redo input key lookup Making key up events implicit was sort-of a nice idea, but it's too tricky and unreliable and makes the key lookup code (interpret_keys()) hard to reason about. See e.g. previous commit for subtle bugs and issues this caused. Make key-up events explicit instead. Add key up events to all VOs. Any time MP_KEY_STATE_DOWN is used, the matching key up event must use MP_KEY_STATE_UP. Rewrite the key lookup code. It should be simpler and more robust now. (Even though the LOC increases, because the new code is less "compact".)
2013-07-01 23:54:59 +02:00
while (max > 0) {
input: clean up messages with button combination names Previously messages that printed key/button names would have extra names in button combinations appended after the main message, resulting in output like: No bind found for key 'MOUSE_BTN1'.-MOUSE_BTN3 Add a function that creates a complete combination name and use that for all such messages. The above example changes to: No bind found for key 'MOUSE_BTN1-MOUSE_BTN3'. Other affected messages are a input.conf parsing error message and a message about a bound command being invalid.
2011-06-29 07:49:01 +02:00
ret = get_key_name(*keys, ret);
if (--max && *++keys)
input: fix crash due to dangling pointer Wrong API usage, has been in the code since 2011. Rarely triggered (if at all in the current setup).
2013-03-19 13:02:42 +01:00
ret = talloc_asprintf_append_buffer(ret, "-");
input: clean up messages with button combination names Previously messages that printed key/button names would have extra names in button combinations appended after the main message, resulting in output like: No bind found for key 'MOUSE_BTN1'.-MOUSE_BTN3 Add a function that creates a complete combination name and use that for all such messages. The above example changes to: No bind found for key 'MOUSE_BTN1-MOUSE_BTN3'. Other affected messages are a input.conf parsing error message and a message about a bound command being invalid.
2011-06-29 07:49:01 +02:00
else
break;
}
return ret;
}
mplayer: turn playtree into a list, and change per-file option handling Summary: - There is no playtree anymore. It's reduced to a simple list. - Options are now always global. You can still have per-file options, but these are optional and require special syntax. - The slave command pt_step has been removed, and playlist_next and playlist_prev added. (See etc/input.conf changes.) This is a user visible incompatible change, and will break slave-mode applications. - The pt_clear slave command is renamed to playlist_clear. - Playtree entries could have multiple files. This is not the case anymore, and playlist entries have always exactly one entry. Whenever something adds more than one file (like ASX playlists or dvd:// or dvdnav:// on the command line), all files are added as separate playlist entries. Note that some of the changes are quite deep and violent. Expect regressions. The playlist parsing code in particular is of low quality. I didn't try to improve it, and merely spent to least effort necessary to keep it somehow working. (Especially ASX playlist handling.) The playtree code was complicated and bloated. It was also barely used. Most users don't even know that mplayer manages the playlist as tree, or how to use it. The most obscure features was probably specifying a tree on command line (with '{' and '}' to create/close tree nodes). It filled the player code with complexity and confused users with weird slave commands like pt_up. Replace the playtree with a simple flat playlist. Playlist parsers that actually return trees are changed to append all files to the playlist pre-order. It used to be the responsibility of the playtree code to change per-file config options. Now this is done by the player core, and the playlist code is free of such details. Options are not per-file by default anymore. This was a very obscure and complicated feature that confused even experienced users. Consider the following command line: mplayer file1.mkv file2.mkv --no-audio file3.mkv This will disable the audio for file2.mkv only, because options are per-file by default. To make the option affect all files, you're supposed to put it before the first file. This is bad, because normally you don't need per-file options. They are very rarely needed, and the only reasonable use cases I can imagine are use of the encode backend (mplayer encode branch), or for debugging. The normal use case is made harder, and the feature is perceived as bug. Even worse, correct usage is hard to explain for users. Make all options global by default. The position of an option isn't significant anymore (except for options that compensate each other, consider --shuffle --no-shuffle). One other important change is that no options are reset anymore if a new file is started. If you change settings with slave mode commands, they will not be changed by playing a new file. (Exceptions include settings that are too file specific, like audio/subtitle stream selection.) There is still some need for per-file options. Debugging and encoding are use cases that profit from per-file options. Per-file profiles (as well as per-protocol and per-VO/AO options) need the implementation related mechanisms to backup and restore options when the playback file changes. Simplify the save-slot stuff, which is possible because there is no hierarchical play tree anymore. Now there's a simple backup field. Add a way to specify per-file options on command line. Example: mplayer f1.mkv -o0 --{ -o1 f2.mkv -o2 f3.mkv --} f4.mkv -o3 will have the following options per file set: f1.mkv, f4.mkv: -o0 -o3 f2.mkv, f3.mkv: -o0 -o3 -o1 -o2 The options --{ and --} start and end per-file options. All files inside the { } will be affected by the options equally (similar to how global options and multiple files are handled). When playback of a file starts, the per-file options are set according to the command line. When playback ends, the per-file options are restored to the values when playback started.
2012-07-31 21:33:26 +02:00
bool mp_input_is_abort_cmd(int cmd_id)
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
{
switch (cmd_id) {
case MP_CMD_QUIT:
mplayer: turn playtree into a list, and change per-file option handling Summary: - There is no playtree anymore. It's reduced to a simple list. - Options are now always global. You can still have per-file options, but these are optional and require special syntax. - The slave command pt_step has been removed, and playlist_next and playlist_prev added. (See etc/input.conf changes.) This is a user visible incompatible change, and will break slave-mode applications. - The pt_clear slave command is renamed to playlist_clear. - Playtree entries could have multiple files. This is not the case anymore, and playlist entries have always exactly one entry. Whenever something adds more than one file (like ASX playlists or dvd:// or dvdnav:// on the command line), all files are added as separate playlist entries. Note that some of the changes are quite deep and violent. Expect regressions. The playlist parsing code in particular is of low quality. I didn't try to improve it, and merely spent to least effort necessary to keep it somehow working. (Especially ASX playlist handling.) The playtree code was complicated and bloated. It was also barely used. Most users don't even know that mplayer manages the playlist as tree, or how to use it. The most obscure features was probably specifying a tree on command line (with '{' and '}' to create/close tree nodes). It filled the player code with complexity and confused users with weird slave commands like pt_up. Replace the playtree with a simple flat playlist. Playlist parsers that actually return trees are changed to append all files to the playlist pre-order. It used to be the responsibility of the playtree code to change per-file config options. Now this is done by the player core, and the playlist code is free of such details. Options are not per-file by default anymore. This was a very obscure and complicated feature that confused even experienced users. Consider the following command line: mplayer file1.mkv file2.mkv --no-audio file3.mkv This will disable the audio for file2.mkv only, because options are per-file by default. To make the option affect all files, you're supposed to put it before the first file. This is bad, because normally you don't need per-file options. They are very rarely needed, and the only reasonable use cases I can imagine are use of the encode backend (mplayer encode branch), or for debugging. The normal use case is made harder, and the feature is perceived as bug. Even worse, correct usage is hard to explain for users. Make all options global by default. The position of an option isn't significant anymore (except for options that compensate each other, consider --shuffle --no-shuffle). One other important change is that no options are reset anymore if a new file is started. If you change settings with slave mode commands, they will not be changed by playing a new file. (Exceptions include settings that are too file specific, like audio/subtitle stream selection.) There is still some need for per-file options. Debugging and encoding are use cases that profit from per-file options. Per-file profiles (as well as per-protocol and per-VO/AO options) need the implementation related mechanisms to backup and restore options when the playback file changes. Simplify the save-slot stuff, which is possible because there is no hierarchical play tree anymore. Now there's a simple backup field. Add a way to specify per-file options on command line. Example: mplayer f1.mkv -o0 --{ -o1 f2.mkv -o2 f3.mkv --} f4.mkv -o3 will have the following options per file set: f1.mkv, f4.mkv: -o0 -o3 f2.mkv, f3.mkv: -o0 -o3 -o1 -o2 The options --{ and --} start and end per-file options. All files inside the { } will be affected by the options equally (similar to how global options and multiple files are handled). When playback of a file starts, the per-file options are set according to the command line. When playback ends, the per-file options are restored to the values when playback started.
2012-07-31 21:33:26 +02:00
case MP_CMD_PLAYLIST_NEXT:
case MP_CMD_PLAYLIST_PREV:
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
return true;
}
return false;
}
input: simplify previous commit a bit more Less mutable state = better.
2012-01-18 02:35:40 +01:00
static int queue_count_cmds(struct cmd_queue *queue)
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
{
input: simplify previous commit a bit more Less mutable state = better.
2012-01-18 02:35:40 +01:00
int res = 0;
for (struct mp_cmd *cmd = queue->first; cmd; cmd = cmd->queue_next)
res++;
return res;
}
static bool queue_has_abort_cmds(struct cmd_queue *queue)
{
input: make input queue thread safe If pthreads are enabled the input queue accesses are regulated by acquiring a mutex. This is useful for platforms like OS X, where the events are created in the cocoa thread and added to the queue to then be dequeued in the playloop thread.
2013-08-13 14:40:17 +02:00
bool ret = false;
for (struct mp_cmd *cmd = queue->first; cmd; cmd = cmd->queue_next)
if (mp_input_is_abort_cmd(cmd->id)) {
ret = true;
break;
}
return ret;
input: fix crashes with libmenu Note: now that libmenu has been removed, the related code in input.c has been removed. If libmenu is ever added back, a queue_remove() function that can remove any item in the input queue will be required. libmenu triggered some special case that caused crashes by double frees. In particular the queue_pop function is used, expecting to remove an element that is not necessarily always the one that queue_pop removes. Fix by adding a queue_remove functions. Rewrite the other queue functions, since these are probably buggy as well. queue_pop didn't even attempt to maintain the doubly linked list.
2011-10-24 02:28:56 +02:00
}
static void queue_remove(struct cmd_queue *queue, struct mp_cmd *cmd)
{
struct mp_cmd **p_prev = &queue->first;
while (*p_prev != cmd) {
p_prev = &(*p_prev)->queue_next;
}
// if this fails, cmd was not in the queue
assert(*p_prev == cmd);
*p_prev = cmd->queue_next;
}
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
static void queue_add(struct cmd_queue *queue, struct mp_cmd *cmd,
bool at_head)
{
input: fix crashes with libmenu Note: now that libmenu has been removed, the related code in input.c has been removed. If libmenu is ever added back, a queue_remove() function that can remove any item in the input queue will be required. libmenu triggered some special case that caused crashes by double frees. In particular the queue_pop function is used, expecting to remove an element that is not necessarily always the one that queue_pop removes. Fix by adding a queue_remove functions. Rewrite the other queue functions, since these are probably buggy as well. queue_pop didn't even attempt to maintain the doubly linked list.
2011-10-24 02:28:56 +02:00
if (at_head) {
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
cmd->queue_next = queue->first;
queue->first = cmd;
} else {
input: fix crashes with libmenu Note: now that libmenu has been removed, the related code in input.c has been removed. If libmenu is ever added back, a queue_remove() function that can remove any item in the input queue will be required. libmenu triggered some special case that caused crashes by double frees. In particular the queue_pop function is used, expecting to remove an element that is not necessarily always the one that queue_pop removes. Fix by adding a queue_remove functions. Rewrite the other queue functions, since these are probably buggy as well. queue_pop didn't even attempt to maintain the doubly linked list.
2011-10-24 02:28:56 +02:00
struct mp_cmd **p_prev = &queue->first;
while (*p_prev)
p_prev = &(*p_prev)->queue_next;
*p_prev = cmd;
cmd->queue_next = NULL;
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
}
input: make input queue thread safe If pthreads are enabled the input queue accesses are regulated by acquiring a mutex. This is useful for platforms like OS X, where the events are created in the cocoa thread and added to the queue to then be dequeued in the playloop thread.
2013-08-13 14:40:17 +02:00
}
static struct mp_cmd *queue_peek(struct cmd_queue *queue)
{
struct mp_cmd *ret = NULL;
ret = queue->first;
return ret;
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
}
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
static struct input_fd *mp_input_add_fd(struct input_ctx *ictx)
{
if (ictx->num_fds == MP_MAX_FDS) {
mp_tmsg(MSGT_INPUT, MSGL_ERR, "Too many file descriptors.\n");
return NULL;
}
struct input_fd *fd = &ictx->fds[ictx->num_fds];
*fd = (struct input_fd){
.fd = -1,
};
ictx->num_fds++;
return fd;
}
int mp_input_add_cmd_fd(struct input_ctx *ictx, int unix_fd, int select,
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
int read_func(int fd, char *dest, int size),
int close_func(int fd))
input: Move command filedescriptor list to context struct
2008-04-30 12:14:03 +02:00
{
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
if (select && unix_fd < 0) {
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
mp_msg(MSGT_INPUT, MSGL_ERR,
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
"Invalid fd %d in mp_input_add_cmd_fd", unix_fd);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
return 0;
}
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
struct input_fd *fd = mp_input_add_fd(ictx);
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
if (fd) {
fd->fd = unix_fd;
fd->select = select;
fd->read_cmd = read_func ? read_func : default_cmd_func;
fd->close_func = close_func;
}
input_unlock(ictx);
return !!fd;
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
}
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
int mp_input_add_key_fd(struct input_ctx *ictx, int unix_fd, int select,
int read_func(void *ctx, int fd),
int close_func(int fd), void *ctx)
input: Move command filedescriptor list to context struct
2008-04-30 12:14:03 +02:00
{
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
if (select && unix_fd < 0) {
mp_msg(MSGT_INPUT, MSGL_ERR,
"Invalid fd %d in mp_input_add_key_fd", unix_fd);
return 0;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
}
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
assert(read_func);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
struct input_fd *fd = mp_input_add_fd(ictx);
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
if (fd) {
fd->fd = unix_fd;
fd->select = select;
fd->read_key = read_func;
fd->close_func = close_func;
fd->ctx = ctx;
}
input_unlock(ictx);
return !!fd;
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
}
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
static void mp_input_rm_fd(struct input_ctx *ictx, int fd)
input: Move key filedescriptor list to context struct The context variable is now passed to VOs too as it's now needed to add a callback on the X filedescriptor.
2008-04-30 10:06:55 +02:00
{
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
struct input_fd *fds = ictx->fds;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
unsigned int i;
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
for (i = 0; i < ictx->num_fds; i++) {
if (fds[i].fd == fd)
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
break;
}
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
if (i == ictx->num_fds)
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
return;
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
if (fds[i].close_func)
fds[i].close_func(fds[i].fd);
talloc_free(fds[i].buffer);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
if (i + 1 < ictx->num_fds)
memmove(&fds[i], &fds[i + 1],
(ictx->num_fds - i - 1) * sizeof(struct input_fd));
ictx->num_fds--;
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
}
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
void mp_input_rm_key_fd(struct input_ctx *ictx, int fd)
input: Remove separate mp_input_add_event_fd Use the same mp_input_add_key_fd for all uses and add a context argument to its callback that was before only in the event fd callbacks. Instead of checking in input.c whether keys were inserted to the keypress FIFO during the callback do the check in the callback before returning and set return value accordingly.
2008-04-29 11:55:23 +02:00
{
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
mp_input_rm_fd(ictx, fd);
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_unlock(ictx);
Add separate event input type for terminal+vo Add an input/input.c fd type whose read function takes no arguments and returns no value. If such a function reads key or command events it'll add them to the queues itself. Use this type for terminal input which was special-cased before. The event function for X11-based VOs will use the same type later. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@24152 b3059339-0415-0410-9bf9-f77b7e298cf2
2007-08-25 06:28:14 +02:00
}
commands: use "up" and "down" as 2nd argument for cycle command Allow the values "up" and "down" as step argument for the cycle input command. Previously, this argument was a float, which specified an arbitrary step value and direction (similar to the add command). Instead of "1" and "-1", "up" and "down" is to be used. Float values are still accepted. That capability might be removed in the future, as there's probably hardly any actual use for arbitrary step values.
2012-10-12 09:21:26 +02:00
static int parse_cycle_dir(const struct m_option *opt, struct bstr name,
struct bstr param, void *dst)
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
{
core: change speed option/property to double The --speed option and the speed property used float. Change them to double. Change the commands that manipulate the property (speed_mult/add) to double as well. Since the cycle command shares code with the add command, we change that as well. The reason for this change is that this allows better control over speed, such as stepping by semitones. Using floats is also just plain unnecessary.
2013-08-04 23:56:20 +02:00
double val;
commands: use "up" and "down" as 2nd argument for cycle command Allow the values "up" and "down" as step argument for the cycle input command. Previously, this argument was a float, which specified an arbitrary step value and direction (similar to the add command). Instead of "1" and "-1", "up" and "down" is to be used. Float values are still accepted. That capability might be removed in the future, as there's probably hardly any actual use for arbitrary step values.
2012-10-12 09:21:26 +02:00
if (bstrcmp0(param, "up") == 0) {
val = +1;
} else if (bstrcmp0(param, "down") == 0) {
val = -1;
} else {
core: change speed option/property to double The --speed option and the speed property used float. Change them to double. Change the commands that manipulate the property (speed_mult/add) to double as well. Since the cycle command shares code with the add command, we change that as well. The reason for this change is that this allows better control over speed, such as stepping by semitones. Using floats is also just plain unnecessary.
2013-08-04 23:56:20 +02:00
return m_option_type_double.parse(opt, name, param, dst);
commands: use "up" and "down" as 2nd argument for cycle command Allow the values "up" and "down" as step argument for the cycle input command. Previously, this argument was a float, which specified an arbitrary step value and direction (similar to the add command). Instead of "1" and "-1", "up" and "down" is to be used. Float values are still accepted. That capability might be removed in the future, as there's probably hardly any actual use for arbitrary step values.
2012-10-12 09:21:26 +02:00
}
core: change speed option/property to double The --speed option and the speed property used float. Change them to double. Change the commands that manipulate the property (speed_mult/add) to double as well. Since the cycle command shares code with the add command, we change that as well. The reason for this change is that this allows better control over speed, such as stepping by semitones. Using floats is also just plain unnecessary.
2013-08-04 23:56:20 +02:00
*(double *)dst = val;
commands: use "up" and "down" as 2nd argument for cycle command Allow the values "up" and "down" as step argument for the cycle input command. Previously, this argument was a float, which specified an arbitrary step value and direction (similar to the add command). Instead of "1" and "-1", "up" and "down" is to be used. Float values are still accepted. That capability might be removed in the future, as there's probably hardly any actual use for arbitrary step values.
2012-10-12 09:21:26 +02:00
return 1;
}
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
static bool read_token(bstr str, bstr *out_rest, bstr *out_token)
commands: change input commands to make OSD usage explicit Most input commands had their own policy whether to display an OSD message for user feedback or not. Some commands had two variants, one that showed an OSD message and one that didn't (e.g. step_property_osd and step_property). Change it such that all commands show a message on the OSD. Add a "no-osd" modifier that disables OSD for that command. Rename the "step_property" and "step_property_osd" command to "switch", and rename "set_property" and "set_property_osd" to "set". Note that commands which haven't used OSD before still don't use OSD. That will possibly be fixed later. (E.g. "screenshot" could display an OSD message instead of just printing a message on the terminal.) The chapter and edition properties still produce OSD messages even with "no-osd", because they don't map so well to the property_osd_display[] mechanism.
2012-09-09 02:08:08 +02:00
{
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
bstr t = bstr_lstrip(str);
int next = bstrcspn(t, WHITESPACE "#");
// Handle comments
input: allow binding multiple commands to a key Separate the commands with ';'.
2013-07-08 19:27:45 +02:00
if (t.len && t.start[next] == '#')
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
t = bstr_splice(t, 0, next);
if (!t.len)
return false;
*out_token = bstr_splice(t, 0, next);
*out_rest = bstr_cut(t, next);
return true;
commands: change input commands to make OSD usage explicit Most input commands had their own policy whether to display an OSD message for user feedback or not. Some commands had two variants, one that showed an OSD message and one that didn't (e.g. step_property_osd and step_property). Change it such that all commands show a message on the OSD. Add a "no-osd" modifier that disables OSD for that command. Rename the "step_property" and "step_property_osd" command to "switch", and rename "set_property" and "set_property_osd" to "set". Note that commands which haven't used OSD before still don't use OSD. That will possibly be fixed later. (E.g. "screenshot" could display an OSD message instead of just printing a message on the terminal.) The chapter and edition properties still produce OSD messages even with "no-osd", because they don't map so well to the property_osd_display[] mechanism.
2012-09-09 02:08:08 +02:00
}
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
static bool eat_token(bstr *str, const char *tok)
commands: change input commands to make OSD usage explicit Most input commands had their own policy whether to display an OSD message for user feedback or not. Some commands had two variants, one that showed an OSD message and one that didn't (e.g. step_property_osd and step_property). Change it such that all commands show a message on the OSD. Add a "no-osd" modifier that disables OSD for that command. Rename the "step_property" and "step_property_osd" command to "switch", and rename "set_property" and "set_property_osd" to "set". Note that commands which haven't used OSD before still don't use OSD. That will possibly be fixed later. (E.g. "screenshot" could display an OSD message instead of just printing a message on the terminal.) The chapter and edition properties still produce OSD messages even with "no-osd", because they don't map so well to the property_osd_display[] mechanism.
2012-09-09 02:08:08 +02:00
{
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
bstr rest, token;
if (read_token(*str, &rest, &token) && bstrcmp0(token, tok) == 0) {
*str = rest;
return true;
}
return false;
}
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
static bool read_escaped_string(void *talloc_ctx, bstr *str, bstr *literal)
{
bstr t = *str;
char *new = talloc_strdup(talloc_ctx, "");
while (t.len) {
if (t.start[0] == '"')
break;
if (t.start[0] == '\\') {
t = bstr_cut(t, 1);
cleanup: remove duplicated function, move escape parsing function
2013-02-16 22:51:10 +01:00
if (!mp_parse_escape(&t, &new))
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
goto error;
} else {
new = talloc_strndup_append_buffer(new, t.start, 1);
t = bstr_cut(t, 1);
}
}
int len = str->len - t.len;
*literal = new ? bstr0(new) : bstr_splice(*str, 0, len);
*str = bstr_cut(*str, len);
return true;
error:
talloc_free(new);
return false;
commands: change input commands to make OSD usage explicit Most input commands had their own policy whether to display an OSD message for user feedback or not. Some commands had two variants, one that showed an OSD message and one that didn't (e.g. step_property_osd and step_property). Change it such that all commands show a message on the OSD. Add a "no-osd" modifier that disables OSD for that command. Rename the "step_property" and "step_property_osd" command to "switch", and rename "set_property" and "set_property_osd" to "set". Note that commands which haven't used OSD before still don't use OSD. That will possibly be fixed later. (E.g. "screenshot" could display an OSD message instead of just printing a message on the terminal.) The chapter and edition properties still produce OSD messages even with "no-osd", because they don't map so well to the property_osd_display[] mechanism.
2012-09-09 02:08:08 +02:00
}
input: allow binding multiple commands to a key Separate the commands with ';'.
2013-07-08 19:27:45 +02:00
// If dest is non-NULL when calling this function, append the command to the
// list formed by dest->queue_next, otherwise just set *dest = new_cmd;
static int parse_cmd(struct mp_cmd **dest, bstr str, const char *loc)
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
{
int pausing = 0;
commands: change input commands to make OSD usage explicit Most input commands had their own policy whether to display an OSD message for user feedback or not. Some commands had two variants, one that showed an OSD message and one that didn't (e.g. step_property_osd and step_property). Change it such that all commands show a message on the OSD. Add a "no-osd" modifier that disables OSD for that command. Rename the "step_property" and "step_property_osd" command to "switch", and rename "set_property" and "set_property_osd" to "set". Note that commands which haven't used OSD before still don't use OSD. That will possibly be fixed later. (E.g. "screenshot" could display an OSD message instead of just printing a message on the terminal.) The chapter and edition properties still produce OSD messages even with "no-osd", because they don't map so well to the property_osd_display[] mechanism.
2012-09-09 02:08:08 +02:00
int on_osd = MP_ON_OSD_AUTO;
input: do property expansion for all input command string arguments Also add a "raw" prefix for commands, which prevents property expansion. The idea is that if the commands are generated by a program, it doesn't have to know whether the command expands properties or not.
2013-05-18 12:56:22 +02:00
bool raw_args = false;
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
struct mp_cmd *cmd = NULL;
input: add input test mode In input test mode, key bindings won't be executed, but are shown on the OSD. The OSD includes various information, such as the name of the key, the command itself, whether it's builtin, and the config file location it was defined. The input test mode can be enabled with "--input=test". No effort is spent trying to react to key bindings that normally exit the player; they are treated just like any other binding.
2012-10-13 21:10:20 +02:00
bstr start = str;
input: allow binding multiple commands to a key Separate the commands with ';'.
2013-07-08 19:27:45 +02:00
bstr next = {0};
input: verify input.conf on loading When input.conf is loaded, verify each command and print a warning if it's invalid or uses legacy commands. This is done for both the user's and the embedded config files. The diff is a bit noisy, because mp_input_parse_cmd() is changed to take a bstr as argument instead of a char*.
2012-09-22 06:33:04 +02:00
void *tmp = talloc_new(NULL);
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
str = bstr_lstrip(str);
commands: simplify legacy command-to-property bridge There are many input commands which are redundant to properties. They were parsed like normal commands, but set_property_command() in command.c handled them automatically using the property mechanism. This still required having the command specifications around, and the code in command.c was quite messy. Replace this with a text based replacement mechanism. Some corner cases are not handled: commands of form "seek_chapter 3 1" are supposed to set the "chapter" property to 3. This use is probably rare, and doesn't show up in the default input.conf. The reason compatibility is kept is because breaking input.conf is quite annoying, so a minimal effort is made to avoid this. Currently we print an annoying warning every time a legacy command is used, though. Also add a compatibility entry for "pt_step", which was removed some time ago. Variations in whitespace are not handled, but it's good enough to deal with old input.conf entries.
2012-09-06 07:19:41 +02:00
for (const struct legacy_cmd *entry = legacy_cmds; entry->old; entry++) {
command: replace some show_ commands with properties show_chapters, show_tracks, and show_playlist are killed and replaced with the properties chapter-list, track-list, and playlist. The code and the output of these stays the same, this is just moving a lot of code around and reducing the number of properties. The "old" commands will still be supported for a while (to avoid making everyone angry), so handle them with the legacy layer. Add something to suppress printing the legacy warnings for these commands.
2013-05-15 02:17:47 +02:00
bstr old = bstr0(entry->old);
bool silent = bstr_eatstart0(&old, "!");
if (bstrcasecmp(bstr_splice(str, 0, old.len), old) == 0) {
if (!silent) {
mp_tmsg(MSGT_INPUT, MSGL_WARN, "Warning: command '%.*s' is "
"deprecated, replaced with '%s' at %s.\n",
BSTR_P(old), entry->new, loc);
}
bstr s = bstr_cut(str, old.len);
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
str = bstr0(talloc_asprintf(tmp, "%s%.*s", entry->new, BSTR_P(s)));
input: add input test mode In input test mode, key bindings won't be executed, but are shown on the OSD. The OSD includes various information, such as the name of the key, the command itself, whether it's builtin, and the config file location it was defined. The input test mode can be enabled with "--input=test". No effort is spent trying to react to key bindings that normally exit the player; they are treated just like any other binding.
2012-10-13 21:10:20 +02:00
start = str;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
break;
commands: simplify legacy command-to-property bridge There are many input commands which are redundant to properties. They were parsed like normal commands, but set_property_command() in command.c handled them automatically using the property mechanism. This still required having the command specifications around, and the code in command.c was quite messy. Replace this with a text based replacement mechanism. Some corner cases are not handled: commands of form "seek_chapter 3 1" are supposed to set the "chapter" property to 3. This use is probably rare, and doesn't show up in the default input.conf. The reason compatibility is kept is because breaking input.conf is quite annoying, so a minimal effort is made to avoid this. Currently we print an annoying warning every time a legacy command is used, though. Also add a compatibility entry for "pt_step", which was removed some time ago. Variations in whitespace are not handled, but it's good enough to deal with old input.conf entries.
2012-09-06 07:19:41 +02:00
}
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
}
input: accept input command prefixes in any order This is more consistent, and doesn't bother the user with ordering rules when new prefixes are added. Will break obscure uses of legacy commands: if the command is supposed to be translated by the legacy command bridge, and if that command uses one of the pausing* prefixes, the command can't be parsed. Well, just use the new commands in this case.
2013-05-18 12:53:38 +02:00
while (1) {
if (eat_token(&str, "pausing")) {
pausing = 1;
} else if (eat_token(&str, "pausing_keep")) {
pausing = 2;
} else if (eat_token(&str, "pausing_toggle")) {
pausing = 3;
} else if (eat_token(&str, "pausing_keep_force")) {
pausing = 4;
} else if (eat_token(&str, "no-osd")) {
on_osd = MP_ON_OSD_NO;
} else if (eat_token(&str, "osd-bar")) {
on_osd = MP_ON_OSD_BAR;
} else if (eat_token(&str, "osd-msg")) {
on_osd = MP_ON_OSD_MSG;
} else if (eat_token(&str, "osd-msg-bar")) {
on_osd = MP_ON_OSD_MSG | MP_ON_OSD_BAR;
} else if (eat_token(&str, "osd-auto")) {
// default
input: do property expansion for all input command string arguments Also add a "raw" prefix for commands, which prevents property expansion. The idea is that if the commands are generated by a program, it doesn't have to know whether the command expands properties or not.
2013-05-18 12:56:22 +02:00
} else if (eat_token(&str, "raw")) {
raw_args = true;
} else if (eat_token(&str, "expand-properties")) {
// default
input: accept input command prefixes in any order This is more consistent, and doesn't bother the user with ordering rules when new prefixes are added. Will break obscure uses of legacy commands: if the command is supposed to be translated by the legacy command bridge, and if that command uses one of the pausing* prefixes, the command can't be parsed. Well, just use the new commands in this case.
2013-05-18 12:53:38 +02:00
} else {
break;
}
commands: change input commands to make OSD usage explicit Most input commands had their own policy whether to display an OSD message for user feedback or not. Some commands had two variants, one that showed an OSD message and one that didn't (e.g. step_property_osd and step_property). Change it such that all commands show a message on the OSD. Add a "no-osd" modifier that disables OSD for that command. Rename the "step_property" and "step_property_osd" command to "switch", and rename "set_property" and "set_property_osd" to "set". Note that commands which haven't used OSD before still don't use OSD. That will possibly be fixed later. (E.g. "screenshot" could display an OSD message instead of just printing a message on the terminal.) The chapter and edition properties still produce OSD messages even with "no-osd", because they don't map so well to the property_osd_display[] mechanism.
2012-09-09 02:08:08 +02:00
}
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
int cmd_idx = 0;
while (mp_cmds[cmd_idx].name != NULL) {
if (eat_token(&str, mp_cmds[cmd_idx].name))
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
break;
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
cmd_idx++;
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
}
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
input: warn about non-existing commands, add explicit "ignore" command If a command is not found, warn about it at loading time (just like other command parsing errors are printed at loading time). Add an explicit "ignore" command. input.conf instructs users to use this command to cancel out existing mapping. This clashed with the warning added in this commit. Make "ignore" a real command and remove the specialcasing for it from get_cmd_from_keys(). Now "ignore" is ignored because it's not handled in command.c.
2012-09-22 15:17:15 +02:00
if (mp_cmds[cmd_idx].name == NULL) {
mp_tmsg(MSGT_INPUT, MSGL_ERR, "Command '%.*s' not found.\n",
BSTR_P(str));
commands: simplify legacy command-to-property bridge There are many input commands which are redundant to properties. They were parsed like normal commands, but set_property_command() in command.c handled them automatically using the property mechanism. This still required having the command specifications around, and the code in command.c was quite messy. Replace this with a text based replacement mechanism. Some corner cases are not handled: commands of form "seek_chapter 3 1" are supposed to set the "chapter" property to 3. This use is probably rare, and doesn't show up in the default input.conf. The reason compatibility is kept is because breaking input.conf is quite annoying, so a minimal effort is made to avoid this. Currently we print an annoying warning every time a legacy command is used, though. Also add a compatibility entry for "pt_step", which was removed some time ago. Variations in whitespace are not handled, but it's good enough to deal with old input.conf entries.
2012-09-06 07:19:41 +02:00
goto error;
input: warn about non-existing commands, add explicit "ignore" command If a command is not found, warn about it at loading time (just like other command parsing errors are printed at loading time). Add an explicit "ignore" command. input.conf instructs users to use this command to cancel out existing mapping. This clashed with the warning added in this commit. Make "ignore" a real command and remove the specialcasing for it from get_cmd_from_keys(). Now "ignore" is ignored because it's not handled in command.c.
2012-09-22 15:17:15 +02:00
}
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
commands: simplify legacy command-to-property bridge There are many input commands which are redundant to properties. They were parsed like normal commands, but set_property_command() in command.c handled them automatically using the property mechanism. This still required having the command specifications around, and the code in command.c was quite messy. Replace this with a text based replacement mechanism. Some corner cases are not handled: commands of form "seek_chapter 3 1" are supposed to set the "chapter" property to 3. This use is probably rare, and doesn't show up in the default input.conf. The reason compatibility is kept is because breaking input.conf is quite annoying, so a minimal effort is made to avoid this. Currently we print an annoying warning every time a legacy command is used, though. Also add a compatibility entry for "pt_step", which was removed some time ago. Variations in whitespace are not handled, but it's good enough to deal with old input.conf entries.
2012-09-06 07:19:41 +02:00
cmd = talloc_ptrtype(NULL, cmd);
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
*cmd = mp_cmds[cmd_idx];
input: add support for precise scroll axes Support horizontal and vertical axes of input devices. If the input device support precise scrolling with an input value then it should first be scaled to a standard multiplier, where 1.0 is the default. The multiplier will then applied to the following commands if possible: * MP_CMD_SEEK * MP_CMD_SPEED_MULT * MP_CMD_ADD All other commands will triggered on every axis event, without change the values specified in the config file.
2013-07-25 18:08:57 +02:00
cmd->scale = 1;
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
cmd->pausing = pausing;
cmd->on_osd = on_osd;
input: do property expansion for all input command string arguments Also add a "raw" prefix for commands, which prevents property expansion. The idea is that if the commands are generated by a program, it doesn't have to know whether the command expands properties or not.
2013-05-18 12:56:22 +02:00
cmd->raw_args = raw_args;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
for (int i = 0; i < MP_CMD_MAX_ARGS; i++) {
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
struct mp_cmd_arg *cmdarg = &cmd->args[i];
if (!cmdarg->type.type)
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
break;
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
str = bstr_lstrip(str);
input: allow binding multiple commands to a key Separate the commands with ';'.
2013-07-08 19:27:45 +02:00
if (eat_token(&str, ";")) {
next = str;
str.len = 0;
break;
}
cmd->nargs++;
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
bstr arg = {0};
input: allow quotes around any input command
2013-05-18 13:18:01 +02:00
if (bstr_eatstart0(&str, "\"")) {
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
if (!read_escaped_string(tmp, &str, &arg)) {
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
mp_tmsg(MSGT_INPUT, MSGL_ERR, "Command %s: argument %d "
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
"has broken string escapes.\n", cmd->name, i + 1);
commands: parsing: fix string argument quoting/escaping Parsing of commands with string arguments had several problems. First, escape handling modified the original string, which broke keyboard command definitions with strings containing escapes. Second, in an argument like "\\" the second quote was incorrectly considered escaped even though the preceding '\' was itself escaped. Third, an unterminated quoted string resulted in the argument being set to NULL even if this violated the declared the minimum number of arguments to the command type; this could cause a crash in the code executing the parsed commands. Fix by rewriting the string argument parsing. Also change int/float argument parsing so that the whole command is rejected on error.
2011-08-05 02:21:14 +02:00
goto error;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
}
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
if (!bstr_eatstart0(&str, "\"")) {
mp_tmsg(MSGT_INPUT, MSGL_ERR, "Command %s: argument %d is "
"unterminated.\n", cmd->name, i + 1);
commands: parsing: fix string argument quoting/escaping Parsing of commands with string arguments had several problems. First, escape handling modified the original string, which broke keyboard command definitions with strings containing escapes. Second, in an argument like "\\" the second quote was incorrectly considered escaped even though the preceding '\' was itself escaped. Third, an unterminated quoted string resulted in the argument being set to NULL even if this violated the declared the minimum number of arguments to the command type; this could cause a crash in the code executing the parsed commands. Fix by rewriting the string argument parsing. Also change int/float argument parsing so that the whole command is rejected on error.
2011-08-05 02:21:14 +02:00
goto error;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
}
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
} else {
if (!read_token(str, &str, &arg))
break;
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
if (cmdarg->optional && bstrcmp0(arg, "-") == 0)
continue;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
}
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
// Prevent option API from trying to deallocate static strings
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
cmdarg->v = ((struct mp_cmd_arg) {{0}}).v;
int r = m_option_parse(&cmdarg->type, bstr0(cmd->name), arg, &cmdarg->v);
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
if (r < 0) {
mp_tmsg(MSGT_INPUT, MSGL_ERR, "Command %s: argument %d "
"can't be parsed: %s.\n", cmd->name, i + 1,
m_option_strerror(r));
goto error;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
}
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
if (cmdarg->type.type == &m_option_type_string)
cmdarg->v.s = talloc_steal(cmd, cmdarg->v.s);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
}
input: allow binding multiple commands to a key Separate the commands with ';'.
2013-07-08 19:27:45 +02:00
if (eat_token(&str, ";")) {
next = str;
str.len = 0;
}
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
bstr dummy;
if (read_token(str, &dummy, &dummy)) {
mp_tmsg(MSGT_INPUT, MSGL_ERR, "Command %s has trailing unused "
"arguments: '%.*s'.\n", cmd->name, BSTR_P(str));
// Better make it fatal to make it clear something is wrong.
commands: parsing: fix string argument quoting/escaping Parsing of commands with string arguments had several problems. First, escape handling modified the original string, which broke keyboard command definitions with strings containing escapes. Second, in an argument like "\\" the second quote was incorrectly considered escaped even though the preceding '\' was itself escaped. Third, an unterminated quoted string resulted in the argument being set to NULL even if this violated the declared the minimum number of arguments to the command type; this could cause a crash in the code executing the parsed commands. Fix by rewriting the string argument parsing. Also change int/float argument parsing so that the whole command is rejected on error.
2011-08-05 02:21:14 +02:00
goto error;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
}
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
int min_args = 0;
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
while (min_args < MP_CMD_MAX_ARGS && cmd->args[min_args].type.type
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
&& !cmd->args[min_args].optional)
{
min_args++;
}
if (cmd->nargs < min_args) {
mp_tmsg(MSGT_INPUT, MSGL_ERR, "Command %s requires at least %d "
"arguments, we found only %d so far.\n", cmd->name, min_args,
cmd->nargs);
goto error;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
}
input: add input test mode In input test mode, key bindings won't be executed, but are shown on the OSD. The OSD includes various information, such as the name of the key, the command itself, whether it's builtin, and the config file location it was defined. The input test mode can be enabled with "--input=test". No effort is spent trying to react to key bindings that normally exit the player; they are treated just like any other binding.
2012-10-13 21:10:20 +02:00
bstr orig = (bstr) {start.start, str.start - start.start};
cmd->original = bstrdup(cmd, bstr_strip(orig));
input: allow binding multiple commands to a key Separate the commands with ';'.
2013-07-08 19:27:45 +02:00
while (*dest)
dest = &(*dest)->queue_next;
*dest = cmd;
next = bstr_strip(next);
if (next.len) {
if (parse_cmd(dest, next, loc) < 0) {
*dest = NULL;
goto error;
}
}
commands: simplify legacy command-to-property bridge There are many input commands which are redundant to properties. They were parsed like normal commands, but set_property_command() in command.c handled them automatically using the property mechanism. This still required having the command specifications around, and the code in command.c was quite messy. Replace this with a text based replacement mechanism. Some corner cases are not handled: commands of form "seek_chapter 3 1" are supposed to set the "chapter" property to 3. This use is probably rare, and doesn't show up in the default input.conf. The reason compatibility is kept is because breaking input.conf is quite annoying, so a minimal effort is made to avoid this. Currently we print an annoying warning every time a legacy command is used, though. Also add a compatibility entry for "pt_step", which was removed some time ago. Variations in whitespace are not handled, but it's good enough to deal with old input.conf entries.
2012-09-06 07:19:41 +02:00
talloc_free(tmp);
input: allow binding multiple commands to a key Separate the commands with ';'.
2013-07-08 19:27:45 +02:00
return 1;
commands: parsing: fix string argument quoting/escaping Parsing of commands with string arguments had several problems. First, escape handling modified the original string, which broke keyboard command definitions with strings containing escapes. Second, in an argument like "\\" the second quote was incorrectly considered escaped even though the preceding '\' was itself escaped. Third, an unterminated quoted string resulted in the argument being set to NULL even if this violated the declared the minimum number of arguments to the command type; this could cause a crash in the code executing the parsed commands. Fix by rewriting the string argument parsing. Also change int/float argument parsing so that the whole command is rejected on error.
2011-08-05 02:21:14 +02:00
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
error:
input: track location info for input commands If parsing a command fails, its location is printed. The location is the path to the input.conf, and the line number of the key binding and the associated input command.
2012-10-13 21:09:42 +02:00
mp_tmsg(MSGT_INPUT, MSGL_ERR, "Command was defined at %s.\n", loc);
input: handle escapes always in command parser Previously, both the command parser and property expansion (m_properties_expand_string) handled escapes with '\'. Move all escape handling into the command parser, and remove it from the property code. This removes the need to escape strings twice for commands that use property expansion. The command parser is practically rewritten: it uses m_option for the actual parsing, and reduces hackish C-string handling.
2012-09-22 14:54:57 +02:00
talloc_free(cmd);
commands: simplify legacy command-to-property bridge There are many input commands which are redundant to properties. They were parsed like normal commands, but set_property_command() in command.c handled them automatically using the property mechanism. This still required having the command specifications around, and the code in command.c was quite messy. Replace this with a text based replacement mechanism. Some corner cases are not handled: commands of form "seek_chapter 3 1" are supposed to set the "chapter" property to 3. This use is probably rare, and doesn't show up in the default input.conf. The reason compatibility is kept is because breaking input.conf is quite annoying, so a minimal effort is made to avoid this. Currently we print an annoying warning every time a legacy command is used, though. Also add a compatibility entry for "pt_step", which was removed some time ago. Variations in whitespace are not handled, but it's good enough to deal with old input.conf entries.
2012-09-06 07:19:41 +02:00
talloc_free(tmp);
input: allow binding multiple commands to a key Separate the commands with ';'.
2013-07-08 19:27:45 +02:00
return -1;
}
mp_cmd_t *mp_input_parse_cmd(bstr str, const char *loc)
{
struct mp_cmd *cmd = NULL;
if (parse_cmd(&cmd, str, loc) < 0) {
assert(!cmd);
}
// Other input.c code uses queue_next for its own purposes, so explicitly
// wrap lists in a pseudo-command.
if (cmd && cmd->queue_next) {
struct mp_cmd *list = talloc_ptrtype(NULL, list);
*list = (struct mp_cmd) {
.id = MP_CMD_COMMAND_LIST,
.name = "list",
input: actually copy sub commands This was missing from the previous commit. It worked by luck, because the sub-commands weren't freed either (as long as the original command was around), but this is proper. Also, set the original string for command lists (needed for input-test only).
2013-07-08 19:44:46 +02:00
.original = bstrdup(list, str),
input: allow binding multiple commands to a key Separate the commands with ';'.
2013-07-08 19:27:45 +02:00
};
list->args[0].v.p = cmd;
while (cmd) {
talloc_steal(list, cmd);
cmd = cmd->queue_next;
}
cmd = list;
}
return cmd;
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
}
Increase slave mode buffer so that one can load files with a deeper path (or bigger URLs). Patch by Mathieu Schroeter (mathieu DOT schroeter AT gamesover DOT ch). git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@29403 b3059339-0415-0410-9bf9-f77b7e298cf2
2009-06-26 22:21:38 +02:00
#define MP_CMD_MAX_SIZE 4096
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
static int read_cmd(struct input_fd *mp_fd, char **ret)
input.c: Remove mp_input_ prefixes from static function names The prefixes make the code unnecessarily verbose for no real benefit.
2008-04-30 08:30:51 +02:00
{
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
char *end;
*ret = NULL;
// Allocate the buffer if it doesn't exist
if (!mp_fd->buffer) {
mp_fd->buffer = talloc_size(NULL, MP_CMD_MAX_SIZE);
mp_fd->pos = 0;
mp_fd->size = MP_CMD_MAX_SIZE;
Fix the bug pointed out by Jindrich Makovicka <makovick@kmlinux.fjfi.cvut.cz> wich affected command driven input (lirc and slave mode). git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@5572 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-04-12 12:17:24 +02:00
}
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
// Get some data if needed/possible
while (!mp_fd->got_cmd && !mp_fd->eof && (mp_fd->size - mp_fd->pos > 1)) {
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
int r = mp_fd->read_cmd(mp_fd->fd, mp_fd->buffer + mp_fd->pos,
mp_fd->size - 1 - mp_fd->pos);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
// Error ?
if (r < 0) {
switch (r) {
case MP_INPUT_ERROR:
case MP_INPUT_DEAD:
mp_tmsg(MSGT_INPUT, MSGL_ERR, "Error while reading "
"command file descriptor %d: %s\n",
mp_fd->fd, strerror(errno));
case MP_INPUT_NOTHING:
return r;
case MP_INPUT_RETRY:
continue;
}
// EOF ?
} else if (r == 0) {
mp_fd->eof = 1;
break;
}
mp_fd->pos += r;
break;
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
}
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
mp_fd->got_cmd = 0;
while (1) {
int l = 0;
// Find the cmd end
mp_fd->buffer[mp_fd->pos] = '\0';
end = strchr(mp_fd->buffer, '\r');
if (end)
*end = '\n';
end = strchr(mp_fd->buffer, '\n');
// No cmd end ?
if (!end) {
// If buffer is full we must drop all until the next \n
if (mp_fd->size - mp_fd->pos <= 1) {
mp_tmsg(MSGT_INPUT, MSGL_ERR, "Command buffer of file "
"descriptor %d is full: dropping content.\n",
mp_fd->fd);
mp_fd->pos = 0;
mp_fd->drop = 1;
}
break;
}
// We already have a cmd : set the got_cmd flag
else if ((*ret)) {
mp_fd->got_cmd = 1;
break;
}
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
l = end - mp_fd->buffer;
Remove trailing whitespace from most files
2009-07-07 01:26:13 +02:00
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
// Not dropping : put the cmd in ret
if (!mp_fd->drop)
*ret = talloc_strndup(NULL, mp_fd->buffer, l);
else
mp_fd->drop = 0;
mp_fd->pos -= l + 1;
memmove(mp_fd->buffer, end + 1, mp_fd->pos);
}
if (*ret)
return 1;
else
return MP_INPUT_NOTHING;
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
}
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
static int default_cmd_func(int fd, char *buf, int l)
input.c: Remove mp_input_ prefixes from static function names The prefixes make the code unnecessarily verbose for no real benefit.
2008-04-30 08:30:51 +02:00
{
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
while (1) {
int r = read(fd, buf, l);
// Error ?
if (r < 0) {
if (errno == EINTR)
continue;
else if (errno == EAGAIN)
return MP_INPUT_NOTHING;
return MP_INPUT_ERROR;
// EOF ?
}
return r;
Reverse Arpi's commit and put the right fix in place. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@7913 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-10-25 18:06:25 +02:00
}
}
w32: silence some warnings
2013-07-13 08:42:14 +02:00
#ifndef __MINGW32__
ao_pulse, core: make pulse thread wake up core for more data For ao_pulse, the current latency is not a good indicator of how soon the AO requires new data to avoid underflow. Add an internal pipe that can be used to wake up the input loop from select(), and make the pulseaudio main loop (which runs in a separate thread) use this mechanism when pulse requests more data. The wakeup signal currently contains no information about the reason for the wakup, but audio buffers are always filled when the event loop wakes up. Also, request a latency of 1 second from the Pulseaudio server. The default is normally significantly higher. We don't need low latency, while higher latency helps prevent underflows reduces need for wakeups.
2012-03-25 21:58:48 +02:00
static int read_wakeup(void *ctx, int fd)
{
char buf[100];
read(fd, buf, sizeof(buf));
return MP_INPUT_NOTHING;
}
w32: silence some warnings
2013-07-13 08:42:14 +02:00
#endif
ao_pulse, core: make pulse thread wake up core for more data For ao_pulse, the current latency is not a good indicator of how soon the AO requires new data to avoid underflow. Add an internal pipe that can be used to wake up the input loop from select(), and make the pulseaudio main loop (which runs in a separate thread) use this mechanism when pulse requests more data. The wakeup signal currently contains no information about the reason for the wakup, but audio buffers are always filled when the event loop wakes up. Also, request a latency of 1 second from the Pulseaudio server. The default is normally significantly higher. We don't need low latency, while higher latency helps prevent underflows reduces need for wakeups.
2012-03-25 21:58:48 +02:00
input: store number of keys, instead of using 0-termination
2013-06-23 00:07:45 +02:00
static bool bind_matches_key(struct cmd_bind *bind, int n, const int *keys);
input: add input test mode In input test mode, key bindings won't be executed, but are shown on the OSD. The OSD includes various information, such as the name of the key, the command itself, whether it's builtin, and the config file location it was defined. The input test mode can be enabled with "--input=test". No effort is spent trying to react to key bindings that normally exit the player; they are treated just like any other binding.
2012-10-13 21:10:20 +02:00
static void append_bind_info(char **pmsg, struct cmd_bind *bind)
{
char *msg = *pmsg;
struct mp_cmd *cmd = mp_input_parse_cmd(bstr0(bind->cmd), bind->location);
bstr stripped = cmd ? cmd->original : bstr0(bind->cmd);
msg = talloc_asprintf_append(msg, " '%.*s'", BSTR_P(stripped));
if (!cmd)
msg = talloc_asprintf_append(msg, " (invalid)");
if (strcmp(bind->owner->section, "default") != 0)
msg = talloc_asprintf_append(msg, " in section {%s}",
bind->owner->section);
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
msg = talloc_asprintf_append(msg, " in %s", bind->location);
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
if (bind->is_builtin)
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
msg = talloc_asprintf_append(msg, " (default)");
talloc_free(cmd);
input: add input test mode In input test mode, key bindings won't be executed, but are shown on the OSD. The OSD includes various information, such as the name of the key, the command itself, whether it's builtin, and the config file location it was defined. The input test mode can be enabled with "--input=test". No effort is spent trying to react to key bindings that normally exit the player; they are treated just like any other binding.
2012-10-13 21:10:20 +02:00
*pmsg = msg;
}
static mp_cmd_t *handle_test(struct input_ctx *ictx, int n, int *keys)
{
char *key_buf = get_key_combo_name(keys, n);
// "$>" to disable property substitution when invoking "show_text"
char *msg = talloc_asprintf(NULL, "$>Key %s is bound to:\n", key_buf);
talloc_free(key_buf);
int count = 0;
for (struct cmd_bind_section *bs = ictx->cmd_bind_sections;
bs; bs = bs->next)
{
input: store number of binds, instead of using 0-termination
2013-06-23 00:40:35 +02:00
for (int i = 0; i < bs->num_binds; i++) {
if (bind_matches_key(&bs->binds[i], n, keys)) {
input: add input test mode In input test mode, key bindings won't be executed, but are shown on the OSD. The OSD includes various information, such as the name of the key, the command itself, whether it's builtin, and the config file location it was defined. The input test mode can be enabled with "--input=test". No effort is spent trying to react to key bindings that normally exit the player; they are treated just like any other binding.
2012-10-13 21:10:20 +02:00
count++;
msg = talloc_asprintf_append(msg, "%d. ", count);
input: store number of binds, instead of using 0-termination
2013-06-23 00:40:35 +02:00
append_bind_info(&msg, &bs->binds[i]);
input: add input test mode In input test mode, key bindings won't be executed, but are shown on the OSD. The OSD includes various information, such as the name of the key, the command itself, whether it's builtin, and the config file location it was defined. The input test mode can be enabled with "--input=test". No effort is spent trying to react to key bindings that normally exit the player; they are treated just like any other binding.
2012-10-13 21:10:20 +02:00
msg = talloc_asprintf_append(msg, "\n");
}
}
}
if (!count)
msg = talloc_asprintf_append(msg, "(nothing)");
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
mp_msg(MSGT_INPUT, MSGL_V, "[input] %s\n", msg);
input: add input test mode In input test mode, key bindings won't be executed, but are shown on the OSD. The OSD includes various information, such as the name of the key, the command itself, whether it's builtin, and the config file location it was defined. The input test mode can be enabled with "--input=test". No effort is spent trying to react to key bindings that normally exit the player; they are treated just like any other binding.
2012-10-13 21:10:20 +02:00
mp_cmd_t *res = mp_input_parse_cmd(bstr0("show_text \"\""), "");
res->args[0].v.s = talloc_steal(res, msg);
return res;
}
input: store number of keys, instead of using 0-termination
2013-06-23 00:07:45 +02:00
static bool bind_matches_key(struct cmd_bind *bind, int num_keys, const int *keys)
input: add input test mode In input test mode, key bindings won't be executed, but are shown on the OSD. The OSD includes various information, such as the name of the key, the command itself, whether it's builtin, and the config file location it was defined. The input test mode can be enabled with "--input=test". No effort is spent trying to react to key bindings that normally exit the player; they are treated just like any other binding.
2012-10-13 21:10:20 +02:00
{
input: store number of keys, instead of using 0-termination
2013-06-23 00:07:45 +02:00
if (bind->num_keys != num_keys)
return false;
for (int i = 0; i < num_keys; i++) {
if (bind->keys[i] != keys[i])
return false;
input: add input test mode In input test mode, key bindings won't be executed, but are shown on the OSD. The OSD includes various information, such as the name of the key, the command itself, whether it's builtin, and the config file location it was defined. The input test mode can be enabled with "--input=test". No effort is spent trying to react to key bindings that normally exit the player; they are treated just like any other binding.
2012-10-13 21:10:20 +02:00
}
input: store number of keys, instead of using 0-termination
2013-06-23 00:07:45 +02:00
return true;
input: add input test mode In input test mode, key bindings won't be executed, but are shown on the OSD. The OSD includes various information, such as the name of the key, the command itself, whether it's builtin, and the config file location it was defined. The input test mode can be enabled with "--input=test". No effort is spent trying to react to key bindings that normally exit the player; they are treated just like any other binding.
2012-10-13 21:10:20 +02:00
}
Add the possibilty to grab the keys and to filter the commands Add the commands for the osd menu git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@8197 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-11-15 00:41:44 +01:00
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
static struct cmd_bind_section *get_bind_section(struct input_ctx *ictx,
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
bstr section)
input: Move command bind variables to context struct
2008-04-30 07:56:45 +02:00
{
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
struct cmd_bind_section *bind_section = ictx->cmd_bind_sections;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
input: warn about non-existing commands, add explicit "ignore" command If a command is not found, warn about it at loading time (just like other command parsing errors are printed at loading time). Add an explicit "ignore" command. input.conf instructs users to use this command to cancel out existing mapping. This clashed with the warning added in this commit. Make "ignore" a real command and remove the specialcasing for it from get_cmd_from_keys(). Now "ignore" is ignored because it's not handled in command.c.
2012-09-22 15:17:15 +02:00
if (section.len == 0)
section = bstr0("default");
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
while (bind_section) {
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
if (bstrcmp0(section, bind_section->section) == 0)
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
return bind_section;
if (bind_section->next == NULL)
break;
bind_section = bind_section->next;
}
if (bind_section) {
bind_section->next = talloc_ptrtype(ictx, bind_section->next);
bind_section = bind_section->next;
} else {
ictx->cmd_bind_sections = talloc_ptrtype(ictx, ictx->cmd_bind_sections);
bind_section = ictx->cmd_bind_sections;
}
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
*bind_section = (struct cmd_bind_section) {
.section = bstrdup0(bind_section, section),
};
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
return bind_section;
implement "bind sections" allows to bind one input key to several slave commands (independently for each section) as shown below: RIGHT seek +10 RIGHT {tv} tv_step_channel 1 RIGHT {dvdnav} dvdnav 4 Currenlty only "tv" section added. patch by Otvos Attila oattila at chello dot hu git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@23497 b3059339-0415-0410-9bf9-f77b7e298cf2
2007-06-07 20:06:53 +02:00
}
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
static struct cmd_bind *find_bind_for_key_section(struct input_ctx *ictx,
char *section,
input: store number of binds, instead of using 0-termination
2013-06-23 00:40:35 +02:00
int num_keys, int *keys)
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
{
struct cmd_bind_section *bs = get_bind_section(ictx, bstr0(section));
input: store number of binds, instead of using 0-termination
2013-06-23 00:40:35 +02:00
if (!num_keys || !bs->num_binds)
return NULL;
// Prefer user-defined keys over builtin bindings
for (int builtin = 0; builtin < 2; builtin++) {
input: stay in current input section if mouse button is down Normally, moving the mouse outside of the mouse area of an input section will send mouse events somewhere else (because input section mouse areas are similar to windows/widgets in real GUI toolkits). This was done even if a mouse button was held down. This is quite different from how GUI toolkits behave. Change the code so that if a mouse button is down, the mouse area of the current input section can't be left. Releasing the mouse button (while the mouse pointer is outside of the mouse area) will actually leave the mouse area. As a side-effect, this commit also tests more often whether the current mouse input section is valid. This is needed to make releasing a mouse button trigger the mouse input section change.
2013-08-30 02:24:14 +02:00
if (builtin && !ictx->default_bindings)
break;
input: store number of binds, instead of using 0-termination
2013-06-23 00:40:35 +02:00
for (int n = 0; n < bs->num_binds; n++) {
if (bs->binds[n].is_builtin == (bool)builtin &&
bind_matches_key(&bs->binds[n], num_keys, keys))
return &bs->binds[n];
}
}
return NULL;
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
}
input: stay in current input section if mouse button is down Normally, moving the mouse outside of the mouse area of an input section will send mouse events somewhere else (because input section mouse areas are similar to windows/widgets in real GUI toolkits). This was done even if a mouse button was held down. This is quite different from how GUI toolkits behave. Change the code so that if a mouse button is down, the mouse area of the current input section can't be left. Releasing the mouse button (while the mouse pointer is outside of the mouse area) will actually leave the mouse area. As a side-effect, this commit also tests more often whether the current mouse input section is valid. This is needed to make releasing a mouse button trigger the mouse input section change.
2013-08-30 02:24:14 +02:00
static bool any_mouse_buttons_down(int num_keys, int *keys)
{
for (int n = 0; n < num_keys; n++) {
if (MP_KEY_IS_MOUSE_BTN_SINGLE(keys[n]))
return true;
}
return false;
}
static bool depends_on_mouse_pos(int num_keys, int *keys)
{
for (int n = 0; n < num_keys; n++) {
if (MP_KEY_DEPENDS_ON_MOUSE_POS(keys[n]))
return true;
}
return false;
}
input: don't let multi-key bindings block simple key bindings Key bindings can include mutiple keys at once (additional to key modifiers like ctrl etc.). This becomes annoying when quickly switching between two bound keys, e.g. when seeking back and forth, you might end up hitting the "left" and "right" keys at once. The user doesn't expect to invoke the key binding "left-right", but would prefer a key stroke to invoke the binding it was supposed to invoke. So if there's no binding for a multi-key combination, try to find a binding for the key last held down. This preserves the ability to define multi-key combinations, while the common case works as expected.
2013-03-30 20:07:15 +01:00
static struct cmd_bind *find_any_bind_for_key(struct input_ctx *ictx,
input: trigger mouse_leave key bindings if mouse leaves mouse area Also, implement mouse leave events for X11. But evne on other platforms, these events will be generated if mouse crosses a section's mouse area boundaries within the mpv window.
2013-06-19 18:19:45 +02:00
char *force_section,
input: don't let multi-key bindings block simple key bindings Key bindings can include mutiple keys at once (additional to key modifiers like ctrl etc.). This becomes annoying when quickly switching between two bound keys, e.g. when seeking back and forth, you might end up hitting the "left" and "right" keys at once. The user doesn't expect to invoke the key binding "left-right", but would prefer a key stroke to invoke the binding it was supposed to invoke. So if there's no binding for a multi-key combination, try to find a binding for the key last held down. This preserves the ability to define multi-key combinations, while the common case works as expected.
2013-03-30 20:07:15 +01:00
int n, int *keys)
input: Move command bind variables to context struct
2008-04-30 07:56:45 +02:00
{
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
if (force_section)
return find_bind_for_key_section(ictx, force_section, n, keys);
input: trigger mouse_leave key bindings if mouse leaves mouse area Also, implement mouse leave events for X11. But evne on other platforms, these events will be generated if mouse crosses a section's mouse area boundaries within the mpv window.
2013-06-19 18:19:45 +02:00
input: stay in current input section if mouse button is down Normally, moving the mouse outside of the mouse area of an input section will send mouse events somewhere else (because input section mouse areas are similar to windows/widgets in real GUI toolkits). This was done even if a mouse button was held down. This is quite different from how GUI toolkits behave. Change the code so that if a mouse button is down, the mouse area of the current input section can't be left. Releasing the mouse button (while the mouse pointer is outside of the mouse area) will actually leave the mouse area. As a side-effect, this commit also tests more often whether the current mouse input section is valid. This is needed to make releasing a mouse button trigger the mouse input section change.
2013-08-30 02:24:14 +02:00
bool use_mouse = depends_on_mouse_pos(n, keys);
// Check global state, because MOUSE_MOVE in particular does not include
// the global state in n/keys.
bool mouse_down = any_mouse_buttons_down(ictx->num_key_down, ictx->key_down);
// First look whether a mouse section is capturing all mouse input
// exclusively (regardless of the active section stack order).
if (use_mouse && mouse_down) {
struct cmd_bind *bind =
find_bind_for_key_section(ictx, ictx->mouse_section, n, keys);
if (bind)
return bind;
}
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
for (int i = ictx->num_active_sections - 1; i >= 0; i--) {
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
struct active_section *s = &ictx->active_sections[i];
struct cmd_bind *bind = find_bind_for_key_section(ictx, s->name, n, keys);
if (bind) {
struct cmd_bind_section *bs = bind->owner;
input: stay in current input section if mouse button is down Normally, moving the mouse outside of the mouse area of an input section will send mouse events somewhere else (because input section mouse areas are similar to windows/widgets in real GUI toolkits). This was done even if a mouse button was held down. This is quite different from how GUI toolkits behave. Change the code so that if a mouse button is down, the mouse area of the current input section can't be left. Releasing the mouse button (while the mouse pointer is outside of the mouse area) will actually leave the mouse area. As a side-effect, this commit also tests more often whether the current mouse input section is valid. This is needed to make releasing a mouse button trigger the mouse input section change.
2013-08-30 02:24:14 +02:00
if (!use_mouse || !bs->mouse_area_set ||
test_rect(&bs->mouse_area, ictx->mouse_vo_x, ictx->mouse_vo_y))
return bind;
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
}
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
if (s->flags & MP_INPUT_EXCLUSIVE)
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
break;
input: add ability to disable all default bindings for an input section Add a flags parameter to mp_input_set_section(). Add a flag that defines whether bindings in the default section are used or not. This is useful for special functionality, where the normal key bindings may have unwanted effects. For example, it shouldn't be possible to seek during encoding. However, you want to be able to cancel the encoding process gracefully. For that purpose, the "encode" section of input.conf could be made exclusive: mp_input_set_section(mpctx->input, "encode", MP_INPUT_NO_DEFAULT_SECTION); And input.conf could contain this definition: RIGHT seek 10 q {encode} quit Then only the key "q" would be bound during encoding.
2012-08-24 13:29:28 +02:00
}
input: trigger mouse_leave key bindings if mouse leaves mouse area Also, implement mouse leave events for X11. But evne on other platforms, these events will be generated if mouse crosses a section's mouse area boundaries within the mpv window.
2013-06-19 18:19:45 +02:00
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
return NULL;
input: don't let multi-key bindings block simple key bindings Key bindings can include mutiple keys at once (additional to key modifiers like ctrl etc.). This becomes annoying when quickly switching between two bound keys, e.g. when seeking back and forth, you might end up hitting the "left" and "right" keys at once. The user doesn't expect to invoke the key binding "left-right", but would prefer a key stroke to invoke the binding it was supposed to invoke. So if there's no binding for a multi-key combination, try to find a binding for the key last held down. This preserves the ability to define multi-key combinations, while the common case works as expected.
2013-03-30 20:07:15 +01:00
}
input: trigger mouse_leave key bindings if mouse leaves mouse area Also, implement mouse leave events for X11. But evne on other platforms, these events will be generated if mouse crosses a section's mouse area boundaries within the mpv window.
2013-06-19 18:19:45 +02:00
static mp_cmd_t *get_cmd_from_keys(struct input_ctx *ictx, char *force_section,
int n, int *keys)
input: don't let multi-key bindings block simple key bindings Key bindings can include mutiple keys at once (additional to key modifiers like ctrl etc.). This becomes annoying when quickly switching between two bound keys, e.g. when seeking back and forth, you might end up hitting the "left" and "right" keys at once. The user doesn't expect to invoke the key binding "left-right", but would prefer a key stroke to invoke the binding it was supposed to invoke. So if there's no binding for a multi-key combination, try to find a binding for the key last held down. This preserves the ability to define multi-key combinations, while the common case works as expected.
2013-03-30 20:07:15 +01:00
{
if (ictx->test)
return handle_test(ictx, n, keys);
input: trigger mouse_leave key bindings if mouse leaves mouse area Also, implement mouse leave events for X11. But evne on other platforms, these events will be generated if mouse crosses a section's mouse area boundaries within the mpv window.
2013-06-19 18:19:45 +02:00
struct cmd_bind *cmd = find_any_bind_for_key(ictx, force_section, n, keys);
input: don't let multi-key bindings block simple key bindings Key bindings can include mutiple keys at once (additional to key modifiers like ctrl etc.). This becomes annoying when quickly switching between two bound keys, e.g. when seeking back and forth, you might end up hitting the "left" and "right" keys at once. The user doesn't expect to invoke the key binding "left-right", but would prefer a key stroke to invoke the binding it was supposed to invoke. So if there's no binding for a multi-key combination, try to find a binding for the key last held down. This preserves the ability to define multi-key combinations, while the common case works as expected.
2013-03-30 20:07:15 +01:00
if (cmd == NULL && n > 1) {
// Hitting two keys at once, and if there's no binding for this
// combination, the key hit last should be checked.
input: trigger mouse_leave key bindings if mouse leaves mouse area Also, implement mouse leave events for X11. But evne on other platforms, these events will be generated if mouse crosses a section's mouse area boundaries within the mpv window.
2013-06-19 18:19:45 +02:00
cmd = find_any_bind_for_key(ictx, force_section, 1, (int[]){keys[n - 1]});
input: don't let multi-key bindings block simple key bindings Key bindings can include mutiple keys at once (additional to key modifiers like ctrl etc.). This becomes annoying when quickly switching between two bound keys, e.g. when seeking back and forth, you might end up hitting the "left" and "right" keys at once. The user doesn't expect to invoke the key binding "left-right", but would prefer a key stroke to invoke the binding it was supposed to invoke. So if there's no binding for a multi-key combination, try to find a binding for the key last held down. This preserves the ability to define multi-key combinations, while the common case works as expected.
2013-03-30 20:07:15 +01:00
}
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
if (cmd == NULL) {
input: clean up messages with button combination names Previously messages that printed key/button names would have extra names in button combinations appended after the main message, resulting in output like: No bind found for key 'MOUSE_BTN1'.-MOUSE_BTN3 Add a function that creates a complete combination name and use that for all such messages. The above example changes to: No bind found for key 'MOUSE_BTN1-MOUSE_BTN3'. Other affected messages are a input.conf parsing error message and a message about a bound command being invalid.
2011-06-29 07:49:01 +02:00
char *key_buf = get_key_combo_name(keys, n);
mp_tmsg(MSGT_INPUT, MSGL_WARN,
"No bind found for key '%s'.\n", key_buf);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
talloc_free(key_buf);
return NULL;
Added key autorepeat support. Options to enable/disable joystick and lirc and set autorepeat delay and rate. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4658 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-11 12:42:08 +01:00
}
input: add input test mode In input test mode, key bindings won't be executed, but are shown on the OSD. The OSD includes various information, such as the name of the key, the command itself, whether it's builtin, and the config file location it was defined. The input test mode can be enabled with "--input=test". No effort is spent trying to react to key bindings that normally exit the player; they are treated just like any other binding.
2012-10-13 21:10:20 +02:00
mp_cmd_t *ret = mp_input_parse_cmd(bstr0(cmd->cmd), cmd->location);
input: trigger mouse_leave key bindings if mouse leaves mouse area Also, implement mouse leave events for X11. But evne on other platforms, these events will be generated if mouse crosses a section's mouse area boundaries within the mpv window.
2013-06-19 18:19:45 +02:00
if (ret) {
ret->input_section = cmd->owner->section;
} else {
input: clean up messages with button combination names Previously messages that printed key/button names would have extra names in button combinations appended after the main message, resulting in output like: No bind found for key 'MOUSE_BTN1'.-MOUSE_BTN3 Add a function that creates a complete combination name and use that for all such messages. The above example changes to: No bind found for key 'MOUSE_BTN1-MOUSE_BTN3'. Other affected messages are a input.conf parsing error message and a message about a bound command being invalid.
2011-06-29 07:49:01 +02:00
char *key_buf = get_key_combo_name(keys, n);
mp_tmsg(MSGT_INPUT, MSGL_ERR,
input: track location info for input commands If parsing a command fails, its location is printed. The location is the path to the input.conf, and the line number of the key binding and the associated input command.
2012-10-13 21:09:42 +02:00
"Invalid command for bound key '%s': '%s'\n", key_buf, cmd->cmd);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
talloc_free(key_buf);
Small bugfix (memcpy too small without sizeof(int) :( ) Add move invalide bind msg so it appear only when needed :) git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4851 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-24 22:22:12 +01:00
}
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
return ret;
Added key autorepeat support. Options to enable/disable joystick and lirc and set autorepeat delay and rate. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4658 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-11 12:42:08 +01:00
}
input: stay in current input section if mouse button is down Normally, moving the mouse outside of the mouse area of an input section will send mouse events somewhere else (because input section mouse areas are similar to windows/widgets in real GUI toolkits). This was done even if a mouse button was held down. This is quite different from how GUI toolkits behave. Change the code so that if a mouse button is down, the mouse area of the current input section can't be left. Releasing the mouse button (while the mouse pointer is outside of the mouse area) will actually leave the mouse area. As a side-effect, this commit also tests more often whether the current mouse input section is valid. This is needed to make releasing a mouse button trigger the mouse input section change.
2013-08-30 02:24:14 +02:00
static void update_mouse_section(struct input_ctx *ictx)
{
struct cmd_bind *cmd =
find_any_bind_for_key(ictx, NULL, 1, (int[]){MP_KEY_MOUSE_MOVE});
char *new_section = cmd ? cmd->owner->section : "default";
char *old = ictx->mouse_section;
ictx->mouse_section = new_section;
if (strcmp(old, ictx->mouse_section) != 0) {
struct mp_cmd *cmd =
get_cmd_from_keys(ictx, old, 1, (int[]){MP_KEY_MOUSE_LEAVE});
if (cmd)
add_key_cmd(ictx, cmd);
}
}
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
static void release_down_cmd(struct input_ctx *ictx)
{
if (ictx->current_down_cmd && ictx->current_down_cmd->key_up_follows) {
ictx->current_down_cmd->key_up_follows = false;
queue_add(&ictx->key_cmd_queue, ictx->current_down_cmd, false);
} else {
talloc_free(ictx->current_down_cmd);
}
ictx->current_down_cmd = NULL;
ictx->last_key_down = 0;
ictx->ar_state = -1;
}
Add the possibilty to grab the keys and to filter the commands Add the commands for the osd menu git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@8197 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-11-15 00:41:44 +01:00
input: require VOs to send key up events, redo input key lookup Making key up events implicit was sort-of a nice idea, but it's too tricky and unreliable and makes the key lookup code (interpret_keys()) hard to reason about. See e.g. previous commit for subtle bugs and issues this caused. Make key-up events explicit instead. Add key up events to all VOs. Any time MP_KEY_STATE_DOWN is used, the matching key up event must use MP_KEY_STATE_UP. Rewrite the key lookup code. It should be simpler and more robust now. (Even though the LOC increases, because the new code is less "compact".)
2013-07-01 23:54:59 +02:00
static int find_key_down(struct input_ctx *ictx, int code)
{
code &= ~(MP_KEY_STATE_UP | MP_KEY_STATE_DOWN);
for (int j = 0; j < ictx->num_key_down; j++) {
if (ictx->key_down[j] == code)
return j;
}
return -1;
}
static void remove_key_down(struct input_ctx *ictx, int code)
Use a single select() for both key and slave input Previous code used two separate select() calls one after another, so that whenever it was running select() on one set of fds events in the other set would go unnoticed until later. Now there's a single select() which allows reacting immediately to any input source. The behavior of the new code differs somewhat from the old; for example multiple fds that stay readable are no longer handled in a round-robin fashion and the total amount the process sleeps can differ. Some tuning might be required later. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@24151 b3059339-0415-0410-9bf9-f77b7e298cf2
2007-08-25 06:28:11 +02:00
{
input: require VOs to send key up events, redo input key lookup Making key up events implicit was sort-of a nice idea, but it's too tricky and unreliable and makes the key lookup code (interpret_keys()) hard to reason about. See e.g. previous commit for subtle bugs and issues this caused. Make key-up events explicit instead. Add key up events to all VOs. Any time MP_KEY_STATE_DOWN is used, the matching key up event must use MP_KEY_STATE_UP. Rewrite the key lookup code. It should be simpler and more robust now. (Even though the LOC increases, because the new code is less "compact".)
2013-07-01 23:54:59 +02:00
int index = find_key_down(ictx, code);
if (index >= 0) {
memmove(&ictx->key_down[index], &ictx->key_down[index + 1],
(ictx->num_key_down - (index + 1)) * sizeof(int));
ictx->num_key_down -= 1;
}
}
Add the possibilty to grab the keys and to filter the commands Add the commands for the osd menu git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@8197 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-11-15 00:41:44 +01:00
input: require VOs to send key up events, redo input key lookup Making key up events implicit was sort-of a nice idea, but it's too tricky and unreliable and makes the key lookup code (interpret_keys()) hard to reason about. See e.g. previous commit for subtle bugs and issues this caused. Make key-up events explicit instead. Add key up events to all VOs. Any time MP_KEY_STATE_DOWN is used, the matching key up event must use MP_KEY_STATE_UP. Rewrite the key lookup code. It should be simpler and more robust now. (Even though the LOC increases, because the new code is less "compact".)
2013-07-01 23:54:59 +02:00
static mp_cmd_t *interpret_key(struct input_ctx *ictx, int code)
{
input: support bindings with modifier keys for X input Add support for binding commands to modifier+key combinations like "Shift+Left" or "Ctrl+Alt+x", and support reading such combinations from the output window of X VOs. The recognized modifier names are Shift, Ctrl, Alt and Meta. Any combination of those and then a non-modifier key name, separated by '+', is accepted as a key name in input.conf. For non-special keys that produce characters shift is ignored as a modifier. For example "A" is handled as a key without modifiers even if you use shift to write the capital letter; 'a' vs 'A' already distinguishes the combinations with a normal keymap, and having separate 'a', 'Shift+A' and 'A' (written with caps lock for example) would bring more confusion than benefit. Currently reading the modifier+key combinations is only supported in the output window of those VOs that use x11_common.c event handling. It's not possible to input the key combinations in other VOs or in a terminal window.
2010-12-19 11:12:20 +01:00
/* On normal keyboards shift changes the character code of non-special
* keys, so don't count the modifier separately for those. In other words
* we want to have "a" and "A" instead of "a" and "Shift+A"; but a separate
* shift modifier is still kept for special keys like arrow keys.
*/
input: require VOs to send key up events, redo input key lookup Making key up events implicit was sort-of a nice idea, but it's too tricky and unreliable and makes the key lookup code (interpret_keys()) hard to reason about. See e.g. previous commit for subtle bugs and issues this caused. Make key-up events explicit instead. Add key up events to all VOs. Any time MP_KEY_STATE_DOWN is used, the matching key up event must use MP_KEY_STATE_UP. Rewrite the key lookup code. It should be simpler and more robust now. (Even though the LOC increases, because the new code is less "compact".)
2013-07-01 23:54:59 +02:00
int unmod = code & ~MP_KEY_MODIFIER_MASK;
input: allow unicode keys and reassign internal key codes This moves all key codes above the highest valid unicode code point (which is 0x10FFFF). All key codes below MP_KEY_BASE now directly map to unicode. Configuration files (input.conf) can contain unicode characters in UTF-8 to map non-ASCII characters/keys. This shouldn't change anything user visible, except that "direct key codes" (as used in input.conf) will change their meaning.
2012-01-14 14:09:26 +01:00
if (unmod >= 32 && unmod < MP_KEY_BASE)
Prefix keycode defines with MP_ Do this to reduce conflicts with <linux/input.h>, which contains some conflicting defines. This changes the meaning of MP_KEY_DOWN: KEY_DOWN is renamed to MP_KEY_DOWN (cursor down key) MP_KEY_DOWN is renamed to MP_KEY_STATE_DOWN (modifier for key down state)
2013-02-12 01:34:04 +01:00
code &= ~MP_KEY_MODIFIER_SHIFT;
input: support bindings with modifier keys for X input Add support for binding commands to modifier+key combinations like "Shift+Left" or "Ctrl+Alt+x", and support reading such combinations from the output window of X VOs. The recognized modifier names are Shift, Ctrl, Alt and Meta. Any combination of those and then a non-modifier key name, separated by '+', is accepted as a key name in input.conf. For non-special keys that produce characters shift is ignored as a modifier. For example "A" is handled as a key without modifiers even if you use shift to write the capital letter; 'a' vs 'A' already distinguishes the combinations with a normal keymap, and having separate 'a', 'Shift+A' and 'A' (written with caps lock for example) would bring more confusion than benefit. Currently reading the modifier+key combinations is only supported in the output window of those VOs that use x11_common.c event handling. It's not possible to input the key combinations in other VOs or in a terminal window.
2010-12-19 11:12:20 +01:00
input: require VOs to send key up events, redo input key lookup Making key up events implicit was sort-of a nice idea, but it's too tricky and unreliable and makes the key lookup code (interpret_keys()) hard to reason about. See e.g. previous commit for subtle bugs and issues this caused. Make key-up events explicit instead. Add key up events to all VOs. Any time MP_KEY_STATE_DOWN is used, the matching key up event must use MP_KEY_STATE_UP. Rewrite the key lookup code. It should be simpler and more robust now. (Even though the LOC increases, because the new code is less "compact".)
2013-07-01 23:54:59 +02:00
if (!(code & MP_KEY_STATE_UP) && ictx->num_key_down >= MP_MAX_KEY_DOWN) {
mp_tmsg(MSGT_INPUT, MSGL_ERR, "Too many key down events "
"at the same time\n");
return NULL;
}
bool key_was_down = find_key_down(ictx, code) >= 0;
Prefix keycode defines with MP_ Do this to reduce conflicts with <linux/input.h>, which contains some conflicting defines. This changes the meaning of MP_KEY_DOWN: KEY_DOWN is renamed to MP_KEY_DOWN (cursor down key) MP_KEY_DOWN is renamed to MP_KEY_STATE_DOWN (modifier for key down state)
2013-02-12 01:34:04 +01:00
if (code & MP_KEY_STATE_DOWN) {
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
// Check if we don't already have this key as pushed
input: require VOs to send key up events, redo input key lookup Making key up events implicit was sort-of a nice idea, but it's too tricky and unreliable and makes the key lookup code (interpret_keys()) hard to reason about. See e.g. previous commit for subtle bugs and issues this caused. Make key-up events explicit instead. Add key up events to all VOs. Any time MP_KEY_STATE_DOWN is used, the matching key up event must use MP_KEY_STATE_UP. Rewrite the key lookup code. It should be simpler and more robust now. (Even though the LOC increases, because the new code is less "compact".)
2013-07-01 23:54:59 +02:00
if (key_was_down)
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
return NULL;
input: require VOs to send key up events, redo input key lookup Making key up events implicit was sort-of a nice idea, but it's too tricky and unreliable and makes the key lookup code (interpret_keys()) hard to reason about. See e.g. previous commit for subtle bugs and issues this caused. Make key-up events explicit instead. Add key up events to all VOs. Any time MP_KEY_STATE_DOWN is used, the matching key up event must use MP_KEY_STATE_UP. Rewrite the key lookup code. It should be simpler and more robust now. (Even though the LOC increases, because the new code is less "compact".)
2013-07-01 23:54:59 +02:00
// Cancel current down-event (there can be only one)
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
release_down_cmd(ictx);
input: require VOs to send key up events, redo input key lookup Making key up events implicit was sort-of a nice idea, but it's too tricky and unreliable and makes the key lookup code (interpret_keys()) hard to reason about. See e.g. previous commit for subtle bugs and issues this caused. Make key-up events explicit instead. Add key up events to all VOs. Any time MP_KEY_STATE_DOWN is used, the matching key up event must use MP_KEY_STATE_UP. Rewrite the key lookup code. It should be simpler and more robust now. (Even though the LOC increases, because the new code is less "compact".)
2013-07-01 23:54:59 +02:00
ictx->key_down[ictx->num_key_down] = code & ~MP_KEY_STATE_DOWN;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
ictx->num_key_down++;
input: stay in current input section if mouse button is down Normally, moving the mouse outside of the mouse area of an input section will send mouse events somewhere else (because input section mouse areas are similar to windows/widgets in real GUI toolkits). This was done even if a mouse button was held down. This is quite different from how GUI toolkits behave. Change the code so that if a mouse button is down, the mouse area of the current input section can't be left. Releasing the mouse button (while the mouse pointer is outside of the mouse area) will actually leave the mouse area. As a side-effect, this commit also tests more often whether the current mouse input section is valid. This is needed to make releasing a mouse button trigger the mouse input section change.
2013-08-30 02:24:14 +02:00
update_mouse_section(ictx);
Replace all calls to GetTimer()/GetTimerMS() GetTimer() is generally replaced with mp_time_us(). Both calls return microseconds, but the latter uses int64_t, us defined to never wrap, and never returns 0 or negative values. GetTimerMS() has no direct replacement. Instead the other functions are used. For some code, switch to mp_time_sec(), which returns the time as double float value in seconds. The returned time is offset to program start time, so there is enough precision left to deliver microsecond resolution for at least 100 years. Unless it's casted to a float (or the CPU reduces precision), which is why we still use mp_time_us() out of paranoia in places where precision is clearly needed. Always switch to the correct time. The whole point of the new timer calls is that they don't wrap, and storing microseconds in unsigned int variables would negate this. In some cases, remove wrap-around handling for time values.
2013-05-25 18:31:06 +02:00
ictx->last_key_down = mp_time_us();
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
ictx->ar_state = 0;
input: trigger mouse_leave key bindings if mouse leaves mouse area Also, implement mouse leave events for X11. But evne on other platforms, these events will be generated if mouse crosses a section's mouse area boundaries within the mpv window.
2013-06-19 18:19:45 +02:00
ictx->current_down_cmd = get_cmd_from_keys(ictx, NULL, ictx->num_key_down,
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
ictx->key_down);
if (ictx->current_down_cmd && (code & MP_KEY_EMIT_ON_UP))
ictx->current_down_cmd->key_up_follows = true;
return mp_cmd_clone(ictx->current_down_cmd);
input: require VOs to send key up events, redo input key lookup Making key up events implicit was sort-of a nice idea, but it's too tricky and unreliable and makes the key lookup code (interpret_keys()) hard to reason about. See e.g. previous commit for subtle bugs and issues this caused. Make key-up events explicit instead. Add key up events to all VOs. Any time MP_KEY_STATE_DOWN is used, the matching key up event must use MP_KEY_STATE_UP. Rewrite the key lookup code. It should be simpler and more robust now. (Even though the LOC increases, because the new code is less "compact".)
2013-07-01 23:54:59 +02:00
} else if (code & MP_KEY_STATE_UP) {
if (key_was_down) {
remove_key_down(ictx, code);
release_down_cmd(ictx);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
}
input: stay in current input section if mouse button is down Normally, moving the mouse outside of the mouse area of an input section will send mouse events somewhere else (because input section mouse areas are similar to windows/widgets in real GUI toolkits). This was done even if a mouse button was held down. This is quite different from how GUI toolkits behave. Change the code so that if a mouse button is down, the mouse area of the current input section can't be left. Releasing the mouse button (while the mouse pointer is outside of the mouse area) will actually leave the mouse area. As a side-effect, this commit also tests more often whether the current mouse input section is valid. This is needed to make releasing a mouse button trigger the mouse input section change.
2013-08-30 02:24:14 +02:00
update_mouse_section(ictx);
input: require VOs to send key up events, redo input key lookup Making key up events implicit was sort-of a nice idea, but it's too tricky and unreliable and makes the key lookup code (interpret_keys()) hard to reason about. See e.g. previous commit for subtle bugs and issues this caused. Make key-up events explicit instead. Add key up events to all VOs. Any time MP_KEY_STATE_DOWN is used, the matching key up event must use MP_KEY_STATE_UP. Rewrite the key lookup code. It should be simpler and more robust now. (Even though the LOC increases, because the new code is less "compact".)
2013-07-01 23:54:59 +02:00
return NULL;
} else {
// Press of key with no separate down/up events
if (key_was_down) {
// Mixing press events and up/down with the same key is not allowed
mp_tmsg(MSGT_INPUT, MSGL_WARN, "Mixing key presses and up/down.\n");
}
input: stay in current input section if mouse button is down Normally, moving the mouse outside of the mouse area of an input section will send mouse events somewhere else (because input section mouse areas are similar to windows/widgets in real GUI toolkits). This was done even if a mouse button was held down. This is quite different from how GUI toolkits behave. Change the code so that if a mouse button is down, the mouse area of the current input section can't be left. Releasing the mouse button (while the mouse pointer is outside of the mouse area) will actually leave the mouse area. As a side-effect, this commit also tests more often whether the current mouse input section is valid. This is needed to make releasing a mouse button trigger the mouse input section change.
2013-08-30 02:24:14 +02:00
update_mouse_section(ictx);
input: require VOs to send key up events, redo input key lookup Making key up events implicit was sort-of a nice idea, but it's too tricky and unreliable and makes the key lookup code (interpret_keys()) hard to reason about. See e.g. previous commit for subtle bugs and issues this caused. Make key-up events explicit instead. Add key up events to all VOs. Any time MP_KEY_STATE_DOWN is used, the matching key up event must use MP_KEY_STATE_UP. Rewrite the key lookup code. It should be simpler and more robust now. (Even though the LOC increases, because the new code is less "compact".)
2013-07-01 23:54:59 +02:00
// Add temporarily (include ongoing down/up events)
int num_key_down = ictx->num_key_down;
int key_down[MP_MAX_KEY_DOWN];
memcpy(key_down, ictx->key_down, num_key_down * sizeof(int));
// Assume doubleclick events never use down/up, while button events do
if (MP_KEY_IS_MOUSE_BTN_DBL(code)) {
// Don't emit "MOUSE_BTN0+MOUSE_BTN0_DBL", just "MOUSE_BTN0_DBL"
int btn = code - MP_MOUSE_BTN0_DBL + MP_MOUSE_BTN0;
if (!num_key_down || key_down[num_key_down - 1] != btn)
return NULL;
key_down[num_key_down - 1] = code;
} else {
key_down[num_key_down] = code;
num_key_down++;
}
return get_cmd_from_keys(ictx, NULL, num_key_down, key_down);
input: modify interpretation of doubleclick events The code combining button presses into multibutton commands prevented single click commands from triggering if a doubleclick event had been generated from the same button press. As a result using the mouse wheel to seek worked very badly. Special-case doubleclick events in the event interpretation code to avoid this issue. This changes the sequence of generated "keys" for press-release-press-release from MOUSE_BTN0 MOUSE_BTN0-MOUSE_BTN0_DBL MOUSE_BTN0_DBL to MOUSE_BTN0 MOUSE_BTN0_DBL MOUSE_BTN0. "Keys" like MOUSE_BTN0-MOUSE_BTN0_DBL will never be generated now; any existing configuration files using those need to be changed.
2011-04-25 13:12:04 +02:00
}
Use a single select() for both key and slave input Previous code used two separate select() calls one after another, so that whenever it was running select() on one set of fds events in the other set would go unnoticed until later. Now there's a single select() which allows reacting immediately to any input source. The behavior of the new code differs somewhat from the old; for example multiple fds that stay readable are no longer handled in a round-robin fashion and the total amount the process sleeps can differ. Some tuning might be required later. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@24151 b3059339-0415-0410-9bf9-f77b7e298cf2
2007-08-25 06:28:11 +02:00
}
Added key autorepeat support. Options to enable/disable joystick and lirc and set autorepeat delay and rate. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4658 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-11 12:42:08 +01:00
input: remove unused "paused" arguments from some functions
2010-04-25 20:47:43 +02:00
static mp_cmd_t *check_autorepeat(struct input_ctx *ictx)
Use a single select() for both key and slave input Previous code used two separate select() calls one after another, so that whenever it was running select() on one set of fds events in the other set would go unnoticed until later. Now there's a single select() which allows reacting immediately to any input source. The behavior of the new code differs somewhat from the old; for example multiple fds that stay readable are no longer handled in a round-robin fashion and the total amount the process sleeps can differ. Some tuning might be required later. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@24151 b3059339-0415-0410-9bf9-f77b7e298cf2
2007-08-25 06:28:11 +02:00
{
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
// No input : autorepeat ?
if (ictx->ar_rate > 0 && ictx->ar_state >= 0 && ictx->num_key_down > 0
&& !(ictx->key_down[ictx->num_key_down - 1] & MP_NO_REPEAT_KEY)) {
Replace all calls to GetTimer()/GetTimerMS() GetTimer() is generally replaced with mp_time_us(). Both calls return microseconds, but the latter uses int64_t, us defined to never wrap, and never returns 0 or negative values. GetTimerMS() has no direct replacement. Instead the other functions are used. For some code, switch to mp_time_sec(), which returns the time as double float value in seconds. The returned time is offset to program start time, so there is enough precision left to deliver microsecond resolution for at least 100 years. Unless it's casted to a float (or the CPU reduces precision), which is why we still use mp_time_us() out of paranoia in places where precision is clearly needed. Always switch to the correct time. The whole point of the new timer calls is that they don't wrap, and storing microseconds in unsigned int variables would negate this. In some cases, remove wrap-around handling for time values.
2013-05-25 18:31:06 +02:00
int64_t t = mp_time_us();
input: don't reset time on each key repeat Key repeats were skipped when playloop iterations took too long. Fix this by using the total times for key repeat calculation, instead of the time difference to the last key repeat event.
2013-04-24 17:42:58 +02:00
if (ictx->last_ar + 2000000 < t)
ictx->last_ar = t;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
// First time : wait delay
if (ictx->ar_state == 0
input: free all queued/allocated commands on exit These were memory leaks in theory, though not in practice (all memory is free'd on exit anyway). However, it was still annoying when leak reporting is enabled. I'm not sure if there was an actual leak in check_autorepeat(), maybe not.
2013-02-03 17:43:47 +01:00
&& (t - ictx->last_key_down) >= ictx->ar_delay * 1000)
{
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
if (!ictx->current_down_cmd) {
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
ictx->ar_state = -1;
return NULL;
}
ictx->ar_state = 1;
input: don't reset time on each key repeat Key repeats were skipped when playloop iterations took too long. Fix this by using the total times for key repeat calculation, instead of the time difference to the last key repeat event.
2013-04-24 17:42:58 +02:00
ictx->last_ar = ictx->last_key_down + ictx->ar_delay * 1000;
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
return mp_cmd_clone(ictx->current_down_cmd);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
// Then send rate / sec event
} else if (ictx->ar_state == 1
&& (t - ictx->last_ar) >= 1000000 / ictx->ar_rate) {
input: don't reset time on each key repeat Key repeats were skipped when playloop iterations took too long. Fix this by using the total times for key repeat calculation, instead of the time difference to the last key repeat event.
2013-04-24 17:42:58 +02:00
ictx->last_ar += 1000000 / ictx->ar_rate;
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
return mp_cmd_clone(ictx->current_down_cmd);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
}
Corrected the quit bug and added support for up to 10 axis git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4519 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-03 20:13:00 +01:00
}
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
return NULL;
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
}
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
static void add_key_cmd(struct input_ctx *ictx, struct mp_cmd *cmd)
{
struct cmd_queue *queue = &ictx->key_cmd_queue;
if (queue_count_cmds(queue) >= ictx->key_fifo_size &&
(!mp_input_is_abort_cmd(cmd->id) || queue_has_abort_cmds(queue)))
{
talloc_free(cmd);
return;
}
queue_add(queue, cmd, false);
}
// Whether a command can deal with redundant key up events.
static bool key_updown_ok(enum mp_command_type cmd)
{
switch (cmd) {
default:
return false;
}
}
core: remove mp_fifo indirection For some reason mp_fifo specifically handled double clicks, and other than that was a pointless wrapper around input.c functionality. Move the double click handling into input.c, and get rid of mp_fifo. Add some compatibility wrappers, because so much VO code uses these functions. Where struct mp_fifo is still used it's just a casted struct input_ctx.
2013-07-02 14:00:24 +02:00
static void mp_input_feed_key(struct input_ctx *ictx, int code)
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
{
ictx->got_new_events = true;
if (code == MP_INPUT_RELEASE_ALL) {
input: fix handling of MP_KEY_STATE_DOWN VOs can use the MP_KEY_STATE_DOWN modifier to pass key up/down events to input.c, instead of just simple key presses. This allows doing key auto- repeat handling in input.c, if the VO doesn't want to do that. One issue is that so far, this code has been used only for mouse events, even though the code was originally written with keyboard keys in mind. One difference between mouse keys and keyboard keys is that the initial key down should not generate an input command with mouse buttons (input.c did that), while keyboard events should (input.c didn't do that). Likewise, releasing a key should generate input commands for mouse buttons releases, but not for the keyboard. Change the code so mouse buttons (recognized via the MP_NO_REPEAT_KEY flag) follow the old hehavior, while other keys generate input commands on key down, but not on key release. Note that a key release event is posted either using MP_INPUT_RELEASE_ALL, or a normal key press event after having sent a an event with MP_KEY_STATE_DOWN. This is probably a bit confusing, and a MP_KEY_STATE_RELEASE should be added. Fix shift-handling with MP_KEY_STATE_DOWN as well.
2013-04-04 01:11:05 +02:00
mp_msg(MSGT_INPUT, MSGL_V, "input: release all\n");
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
ictx->num_key_down = 0;
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
release_down_cmd(ictx);
input: stay in current input section if mouse button is down Normally, moving the mouse outside of the mouse area of an input section will send mouse events somewhere else (because input section mouse areas are similar to windows/widgets in real GUI toolkits). This was done even if a mouse button was held down. This is quite different from how GUI toolkits behave. Change the code so that if a mouse button is down, the mouse area of the current input section can't be left. Releasing the mouse button (while the mouse pointer is outside of the mouse area) will actually leave the mouse area. As a side-effect, this commit also tests more often whether the current mouse input section is valid. This is needed to make releasing a mouse button trigger the mouse input section change.
2013-08-30 02:24:14 +02:00
update_mouse_section(ictx);
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
return;
}
input: require VOs to send key up events, redo input key lookup Making key up events implicit was sort-of a nice idea, but it's too tricky and unreliable and makes the key lookup code (interpret_keys()) hard to reason about. See e.g. previous commit for subtle bugs and issues this caused. Make key-up events explicit instead. Add key up events to all VOs. Any time MP_KEY_STATE_DOWN is used, the matching key up event must use MP_KEY_STATE_UP. Rewrite the key lookup code. It should be simpler and more robust now. (Even though the LOC increases, because the new code is less "compact".)
2013-07-01 23:54:59 +02:00
int unmod = code & ~MP_KEY_MODIFIER_MASK;
if (MP_KEY_DEPENDS_ON_MOUSE_POS(unmod))
ictx->mouse_event_counter++;
input: fix handling of MP_KEY_STATE_DOWN VOs can use the MP_KEY_STATE_DOWN modifier to pass key up/down events to input.c, instead of just simple key presses. This allows doing key auto- repeat handling in input.c, if the VO doesn't want to do that. One issue is that so far, this code has been used only for mouse events, even though the code was originally written with keyboard keys in mind. One difference between mouse keys and keyboard keys is that the initial key down should not generate an input command with mouse buttons (input.c did that), while keyboard events should (input.c didn't do that). Likewise, releasing a key should generate input commands for mouse buttons releases, but not for the keyboard. Change the code so mouse buttons (recognized via the MP_NO_REPEAT_KEY flag) follow the old hehavior, while other keys generate input commands on key down, but not on key release. Note that a key release event is posted either using MP_INPUT_RELEASE_ALL, or a normal key press event after having sent a an event with MP_KEY_STATE_DOWN. This is probably a bit confusing, and a MP_KEY_STATE_RELEASE should be added. Fix shift-handling with MP_KEY_STATE_DOWN as well.
2013-04-04 01:11:05 +02:00
mp_msg(MSGT_INPUT, MSGL_V, "input: key code=%#x\n", code);
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
struct mp_cmd *cmd = interpret_key(ictx, code);
if (!cmd)
return;
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
// Prevent redundant key-down events from being added to the queue. In some
// cases (like MP_CMD_SEEK commands), duplicated events might severely
// confuse the frontend.
if (cmd->key_up_follows && !key_updown_ok(cmd->id)) {
input: fix mp_cmd memory leak mp_input_feed_key() didn't free the command, if it was discarded due to the queue exceeding its maximum size.
2012-10-14 23:45:41 +02:00
talloc_free(cmd);
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
return;
input: fix mp_cmd memory leak mp_input_feed_key() didn't free the command, if it was discarded due to the queue exceeding its maximum size.
2012-10-14 23:45:41 +02:00
}
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
add_key_cmd(ictx, cmd);
}
core: remove mp_fifo indirection For some reason mp_fifo specifically handled double clicks, and other than that was a pointless wrapper around input.c functionality. Move the double click handling into input.c, and get rid of mp_fifo. Add some compatibility wrappers, because so much VO code uses these functions. Where struct mp_fifo is still used it's just a casted struct input_ctx.
2013-07-02 14:00:24 +02:00
void mp_input_put_key(struct input_ctx *ictx, int code)
{
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
core: remove mp_fifo indirection For some reason mp_fifo specifically handled double clicks, and other than that was a pointless wrapper around input.c functionality. Move the double click handling into input.c, and get rid of mp_fifo. Add some compatibility wrappers, because so much VO code uses these functions. Where struct mp_fifo is still used it's just a casted struct input_ctx.
2013-07-02 14:00:24 +02:00
double now = mp_time_sec();
int doubleclick_time = ictx->doubleclick_time;
// ignore system-doubleclick if we generate these events ourselves
int unmod = code & ~MP_KEY_MODIFIER_MASK;
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
if (doubleclick_time && MP_KEY_IS_MOUSE_BTN_DBL(unmod)) {
input_unlock(ictx);
core: remove mp_fifo indirection For some reason mp_fifo specifically handled double clicks, and other than that was a pointless wrapper around input.c functionality. Move the double click handling into input.c, and get rid of mp_fifo. Add some compatibility wrappers, because so much VO code uses these functions. Where struct mp_fifo is still used it's just a casted struct input_ctx.
2013-07-02 14:00:24 +02:00
return;
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
}
core: remove mp_fifo indirection For some reason mp_fifo specifically handled double clicks, and other than that was a pointless wrapper around input.c functionality. Move the double click handling into input.c, and get rid of mp_fifo. Add some compatibility wrappers, because so much VO code uses these functions. Where struct mp_fifo is still used it's just a casted struct input_ctx.
2013-07-02 14:00:24 +02:00
mp_input_feed_key(ictx, code);
if (code & MP_KEY_STATE_DOWN) {
code &= ~MP_KEY_STATE_DOWN;
if (ictx->last_doubleclick_key_down == code
&& now - ictx->last_doubleclick_time < doubleclick_time / 1000.0)
{
if (code >= MP_MOUSE_BTN0 && code <= MP_MOUSE_BTN2)
mp_input_feed_key(ictx, code - MP_MOUSE_BTN0 + MP_MOUSE_BTN0_DBL);
}
ictx->last_doubleclick_key_down = code;
ictx->last_doubleclick_time = now;
}
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_unlock(ictx);
core: remove mp_fifo indirection For some reason mp_fifo specifically handled double clicks, and other than that was a pointless wrapper around input.c functionality. Move the double click handling into input.c, and get rid of mp_fifo. Add some compatibility wrappers, because so much VO code uses these functions. Where struct mp_fifo is still used it's just a casted struct input_ctx.
2013-07-02 14:00:24 +02:00
}
void mp_input_put_key_utf8(struct input_ctx *ictx, int mods, struct bstr t)
{
while (t.len) {
int code = bstr_decode_utf8(t, &t);
if (code < 0)
break;
mp_input_put_key(ictx, code | mods);
}
}
input: add support for precise scroll axes Support horizontal and vertical axes of input devices. If the input device support precise scrolling with an input value then it should first be scaled to a standard multiplier, where 1.0 is the default. The multiplier will then applied to the following commands if possible: * MP_CMD_SEEK * MP_CMD_SPEED_MULT * MP_CMD_ADD All other commands will triggered on every axis event, without change the values specified in the config file.
2013-07-25 18:08:57 +02:00
void mp_input_put_axis(struct input_ctx *ictx, int direction, double value)
{
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
input: add support for precise scroll axes Support horizontal and vertical axes of input devices. If the input device support precise scrolling with an input value then it should first be scaled to a standard multiplier, where 1.0 is the default. The multiplier will then applied to the following commands if possible: * MP_CMD_SEEK * MP_CMD_SPEED_MULT * MP_CMD_ADD All other commands will triggered on every axis event, without change the values specified in the config file.
2013-07-25 18:08:57 +02:00
struct mp_cmd *cmd = interpret_key(ictx, direction);
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
if (cmd) {
cmd->scale = value;
input: add support for precise scroll axes Support horizontal and vertical axes of input devices. If the input device support precise scrolling with an input value then it should first be scaled to a standard multiplier, where 1.0 is the default. The multiplier will then applied to the following commands if possible: * MP_CMD_SEEK * MP_CMD_SPEED_MULT * MP_CMD_ADD All other commands will triggered on every axis event, without change the values specified in the config file.
2013-07-25 18:08:57 +02:00
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
ictx->got_new_events = true;
add_key_cmd(ictx, cmd);
}
input_unlock(ictx);
input: add support for precise scroll axes Support horizontal and vertical axes of input devices. If the input device support precise scrolling with an input value then it should first be scaled to a standard multiplier, where 1.0 is the default. The multiplier will then applied to the following commands if possible: * MP_CMD_SEEK * MP_CMD_SPEED_MULT * MP_CMD_ADD All other commands will triggered on every axis event, without change the values specified in the config file.
2013-07-25 18:08:57 +02:00
}
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
void mp_input_set_mouse_pos(struct input_ctx *ictx, int x, int y)
{
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
input: Ignore spurious mouse moves Windows generates WM_MOUSEMOVE messages internally whenever the window manager wants to know where the mouse is[1] and broadcasts that to everyone; w32_common doesn't check whether the position is different and just (indirectly) calls this. Do the check on input.c since it's possible some other VO or frontend also do the same thing. [1]: http://blogs.msdn.com/b/oldnewthing/archive/2003/10/01/55108.aspx
2013-07-20 23:28:43 +02:00
// we're already there
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
if (ictx->mouse_vo_x == x && ictx->mouse_vo_y == y) {
input_unlock(ictx);
input: Ignore spurious mouse moves Windows generates WM_MOUSEMOVE messages internally whenever the window manager wants to know where the mouse is[1] and broadcasts that to everyone; w32_common doesn't check whether the position is different and just (indirectly) calls this. Do the check on input.c since it's possible some other VO or frontend also do the same thing. [1]: http://blogs.msdn.com/b/oldnewthing/archive/2003/10/01/55108.aspx
2013-07-20 23:28:43 +02:00
return;
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
}
input: Ignore spurious mouse moves Windows generates WM_MOUSEMOVE messages internally whenever the window manager wants to know where the mouse is[1] and broadcasts that to everyone; w32_common doesn't check whether the position is different and just (indirectly) calls this. Do the check on input.c since it's possible some other VO or frontend also do the same thing. [1]: http://blogs.msdn.com/b/oldnewthing/archive/2003/10/01/55108.aspx
2013-07-20 23:28:43 +02:00
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
ictx->mouse_event_counter++;
ictx->mouse_vo_x = x;
ictx->mouse_vo_y = y;
input: trigger mouse_leave key bindings if mouse leaves mouse area Also, implement mouse leave events for X11. But evne on other platforms, these events will be generated if mouse crosses a section's mouse area boundaries within the mpv window.
2013-06-19 18:19:45 +02:00
input: stay in current input section if mouse button is down Normally, moving the mouse outside of the mouse area of an input section will send mouse events somewhere else (because input section mouse areas are similar to windows/widgets in real GUI toolkits). This was done even if a mouse button was held down. This is quite different from how GUI toolkits behave. Change the code so that if a mouse button is down, the mouse area of the current input section can't be left. Releasing the mouse button (while the mouse pointer is outside of the mouse area) will actually leave the mouse area. As a side-effect, this commit also tests more often whether the current mouse input section is valid. This is needed to make releasing a mouse button trigger the mouse input section change.
2013-08-30 02:24:14 +02:00
update_mouse_section(ictx);
input: trigger mouse_leave key bindings if mouse leaves mouse area Also, implement mouse leave events for X11. But evne on other platforms, these events will be generated if mouse crosses a section's mouse area boundaries within the mpv window.
2013-06-19 18:19:45 +02:00
struct mp_cmd *cmd =
get_cmd_from_keys(ictx, NULL, 1, (int[]){MP_KEY_MOUSE_MOVE});
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
if (cmd) {
cmd->mouse_move = true;
cmd->mouse_x = x;
cmd->mouse_y = y;
add_key_cmd(ictx, cmd);
}
input_unlock(ictx);
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
}
static void read_cmd_fd(struct input_ctx *ictx, struct input_fd *cmd_fd)
{
int r;
char *text;
while ((r = read_cmd(cmd_fd, &text)) >= 0) {
ictx->got_new_events = true;
input: track location info for input commands If parsing a command fails, its location is printed. The location is the path to the input.conf, and the line number of the key binding and the associated input command.
2012-10-13 21:09:42 +02:00
struct mp_cmd *cmd = mp_input_parse_cmd(bstr0(text), "<pipe>");
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
talloc_free(text);
if (cmd)
queue_add(&ictx->control_cmd_queue, cmd, false);
if (!cmd_fd->got_cmd)
return;
}
if (r == MP_INPUT_ERROR)
mp_tmsg(MSGT_INPUT, MSGL_ERR, "Error on command file descriptor %d\n",
cmd_fd->fd);
else if (r == MP_INPUT_DEAD)
cmd_fd->dead = true;
}
static void read_key_fd(struct input_ctx *ictx, struct input_fd *key_fd)
{
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
int code = key_fd->read_key(key_fd->ctx, key_fd->fd);
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
if (code >= 0 || code == MP_INPUT_RELEASE_ALL) {
mp_input_feed_key(ictx, code);
return;
}
if (code == MP_INPUT_ERROR)
mp_tmsg(MSGT_INPUT, MSGL_ERR,
"Error on key input file descriptor %d\n", key_fd->fd);
else if (code == MP_INPUT_DEAD) {
mp_tmsg(MSGT_INPUT, MSGL_ERR,
"Dead key input on file descriptor %d\n", key_fd->fd);
key_fd->dead = true;
}
}
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
static void read_fd(struct input_ctx *ictx, struct input_fd *fd)
Use a single select() for both key and slave input Previous code used two separate select() calls one after another, so that whenever it was running select() on one set of fds events in the other set would go unnoticed until later. Now there's a single select() which allows reacting immediately to any input source. The behavior of the new code differs somewhat from the old; for example multiple fds that stay readable are no longer handled in a round-robin fashion and the total amount the process sleeps can differ. Some tuning might be required later. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@24151 b3059339-0415-0410-9bf9-f77b7e298cf2
2007-08-25 06:28:11 +02:00
{
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
if (fd->read_cmd) {
read_cmd_fd(ictx, fd);
} else {
read_key_fd(ictx, fd);
input: adjust wait time for key-repeat Basically, these are additional timers that can expire without making the central select() exit.
2013-04-24 17:02:52 +02:00
}
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
}
static void remove_dead_fds(struct input_ctx *ictx)
{
for (int i = 0; i < ictx->num_fds; i++) {
if (ictx->fds[i].dead) {
mp_input_rm_fd(ictx, ictx->fds[i].fd);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
i--;
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
}
}
}
Invert logic for HAVE_NO_POSIX_SELECT to HAVE_POSIX_SELECT to be in line with the other HAVE_ #defines throughout the codebase. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@22769 b3059339-0415-0410-9bf9-f77b7e298cf2
2007-03-21 09:24:20 +01:00
#ifdef HAVE_POSIX_SELECT
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
static void input_wait_read(struct input_ctx *ictx, int time)
{
input.c: Revert code uglifications done for gcc 2.95 It's unlikely that this branch will ever compile with gcc 2.95 again, so it's pointless to keep code uglifications done to work around its limitations.
2008-04-29 11:24:02 +02:00
fd_set fds;
Use a single select() for both key and slave input Previous code used two separate select() calls one after another, so that whenever it was running select() on one set of fds events in the other set would go unnoticed until later. Now there's a single select() which allows reacting immediately to any input source. The behavior of the new code differs somewhat from the old; for example multiple fds that stay readable are no longer handled in a round-robin fashion and the total amount the process sleeps can differ. Some tuning might be required later. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@24151 b3059339-0415-0410-9bf9-f77b7e298cf2
2007-08-25 06:28:11 +02:00
FD_ZERO(&fds);
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
int max_fd = 0;
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
for (int i = 0; i < ictx->num_fds; i++) {
input: invert negated boolean field
2013-07-14 15:37:39 +02:00
if (!ictx->fds[i].select)
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
continue;
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
if (ictx->fds[i].fd > max_fd)
max_fd = ictx->fds[i].fd;
FD_SET(ictx->fds[i].fd, &fds);
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
}
struct timeval tv, *time_val;
core: do mouse cursor hiding business in frontend Do this so that not every VO backend has to setup a timer for cursor hiding and interpret the --cursor-autohide option.
2013-05-16 23:17:46 +02:00
tv.tv_sec = time / 1000;
tv.tv_usec = (time % 1000) * 1000;
time_val = &tv;
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
if (select(max_fd + 1, &fds, NULL, NULL, time_val) < 0) {
if (errno != EINTR)
mp_tmsg(MSGT_INPUT, MSGL_ERR, "Select error: %s\n",
strerror(errno));
FD_ZERO(&fds);
Use a single select() for both key and slave input Previous code used two separate select() calls one after another, so that whenever it was running select() on one set of fds events in the other set would go unnoticed until later. Now there's a single select() which allows reacting immediately to any input source. The behavior of the new code differs somewhat from the old; for example multiple fds that stay readable are no longer handled in a round-robin fashion and the total amount the process sleeps can differ. Some tuning might be required later. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@24151 b3059339-0415-0410-9bf9-f77b7e298cf2
2007-08-25 06:28:11 +02:00
}
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
for (int i = 0; i < ictx->num_fds; i++) {
input: invert negated boolean field
2013-07-14 15:37:39 +02:00
if (ictx->fds[i].select && !FD_ISSET(ictx->fds[i].fd, &fds))
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
continue;
read_fd(ictx, &ictx->fds[i]);
}
}
Use a single select() for both key and slave input Previous code used two separate select() calls one after another, so that whenever it was running select() on one set of fds events in the other set would go unnoticed until later. Now there's a single select() which allows reacting immediately to any input source. The behavior of the new code differs somewhat from the old; for example multiple fds that stay readable are no longer handled in a round-robin fashion and the total amount the process sleeps can differ. Some tuning might be required later. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@24151 b3059339-0415-0410-9bf9-f77b7e298cf2
2007-08-25 06:28:11 +02:00
#else
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
static void input_wait_read(struct input_ctx *ictx, int time)
{
core: do mouse cursor hiding business in frontend Do this so that not every VO backend has to setup a timer for cursor hiding and interpret the --cursor-autohide option.
2013-05-16 23:17:46 +02:00
if (time > 0)
Replace calls to usec_sleep() This is just dumb sed replacement to mp_sleep_us(). Also remove the now unused usec_sleep() wrapper.
2013-05-25 19:56:52 +02:00
mp_sleep_us(time * 1000);
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
for (int i = 0; i < ictx->num_fds; i++)
read_fd(ictx, &ictx->fds[i]);
}
Added comments support in input.conf (comments lines with #) And fixed a possible bug git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4764 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-20 15:45:18 +01:00
allow select()less compilation git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@9837 b3059339-0415-0410-9bf9-f77b7e298cf2
2003-04-04 22:56:29 +02:00
#endif
Use a single select() for both key and slave input Previous code used two separate select() calls one after another, so that whenever it was running select() on one set of fds events in the other set would go unnoticed until later. Now there's a single select() which allows reacting immediately to any input source. The behavior of the new code differs somewhat from the old; for example multiple fds that stay readable are no longer handled in a round-robin fashion and the total amount the process sleeps can differ. Some tuning might be required later. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@24151 b3059339-0415-0410-9bf9-f77b7e298cf2
2007-08-25 06:28:11 +02:00
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
/**
* \param time time to wait at most for an event in milliseconds
*/
static void read_events(struct input_ctx *ictx, int time)
{
if (ictx->num_key_down) {
time = FFMIN(time, 1000 / ictx->ar_rate);
time = FFMIN(time, ictx->ar_delay);
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
}
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
time = FFMAX(time, 0);
input: never wait if there are new events in the input queue Move the reading loop from read_all_fd_events to read_events. If got_new_events is set when calling read_events, don't actually wait and set the timeout to 0. (Note that not waiting is sort of transparent to the caller: the caller is just supposed to execute the event loop again, and then it will actually wait. mplayer.c handles this correctly.) This might reduce latency with some input sources.
2013-07-14 15:44:11 +02:00
while (1) {
if (ictx->got_new_events)
time = 0;
ictx->got_new_events = false;
remove_dead_fds(ictx);
if (time) {
for (int i = 0; i < ictx->num_fds; i++) {
if (!ictx->fds[i].select)
read_fd(ictx, &ictx->fds[i]);
}
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
}
input: never wait if there are new events in the input queue Move the reading loop from read_all_fd_events to read_events. If got_new_events is set when calling read_events, don't actually wait and set the timeout to 0. (Note that not waiting is sort of transparent to the caller: the caller is just supposed to execute the event loop again, and then it will actually wait. mplayer.c handles this correctly.) This might reduce latency with some input sources.
2013-07-14 15:44:11 +02:00
if (ictx->got_new_events)
time = 0;
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
input: never wait if there are new events in the input queue Move the reading loop from read_all_fd_events to read_events. If got_new_events is set when calling read_events, don't actually wait and set the timeout to 0. (Note that not waiting is sort of transparent to the caller: the caller is just supposed to execute the event loop again, and then it will actually wait. mplayer.c handles this correctly.) This might reduce latency with some input sources.
2013-07-14 15:44:11 +02:00
input_wait_read(ictx, time);
// Read until all input FDs are empty
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
if (!ictx->got_new_events)
input: never wait if there are new events in the input queue Move the reading loop from read_all_fd_events to read_events. If got_new_events is set when calling read_events, don't actually wait and set the timeout to 0. (Note that not waiting is sort of transparent to the caller: the caller is just supposed to execute the event loop again, and then it will actually wait. mplayer.c handles this correctly.) This might reduce latency with some input sources.
2013-07-14 15:44:11 +02:00
break;
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
}
}
Use a single select() for both key and slave input Previous code used two separate select() calls one after another, so that whenever it was running select() on one set of fds events in the other set would go unnoticed until later. Now there's a single select() which allows reacting immediately to any input source. The behavior of the new code differs somewhat from the old; for example multiple fds that stay readable are no longer handled in a round-robin fashion and the total amount the process sleeps can differ. Some tuning might be required later. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@24151 b3059339-0415-0410-9bf9-f77b7e298cf2
2007-08-25 06:28:11 +02:00
OSX, input: implement wakeup in response to Cocoa events Add code to wake up the select() call in input.c when an OSX event is available and a Cocoa OpenGL backend is initialized. Fixes the slow response to input or other events in Cocoa-based VOs during long select() sleeps (e.g., when mplayer2 is paused) introduced by commit 7040968.
2012-04-03 08:13:12 +02:00
static void read_all_events(struct input_ctx *ictx, int time)
{
mp_msg, getch2: unix tty background support Now, when backgrounded, mpv plays and outputs messages to stdout, but statusline is not output. Background<->foreground transitions are detected by signals and polling the process groups.
2013-01-23 15:47:49 +01:00
getch2_poll();
input: never wait if there are new events in the input queue Move the reading loop from read_all_fd_events to read_events. If got_new_events is set when calling read_events, don't actually wait and set the timeout to 0. (Note that not waiting is sort of transparent to the caller: the caller is just supposed to execute the event loop again, and then it will actually wait. mplayer.c handles this correctly.) This might reduce latency with some input sources.
2013-07-14 15:44:11 +02:00
read_events(ictx, time);
OSX, input: implement wakeup in response to Cocoa events Add code to wake up the select() call in input.c when an OSX event is available and a Cocoa OpenGL backend is initialized. Fixes the slow response to input or other events in Cocoa-based VOs during long select() sleeps (e.g., when mplayer2 is paused) introduced by commit 7040968.
2012-04-03 08:13:12 +02:00
}
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
int mp_input_queue_cmd(struct input_ctx *ictx, mp_cmd_t *cmd)
input: Move cmd_queue to context struct Menu instances now also need a input context pointer to queue commands.
2008-04-30 12:00:59 +02:00
{
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
ictx->got_new_events = true;
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
if (cmd)
queue_add(&ictx->control_cmd_queue, cmd, false);
input_unlock(ictx);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
return 1;
Fixed bug with comments in input.conf parser Added a command queue to let mplayer send command to himself git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4822 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-23 22:13:35 +01:00
}
Do not loose commands while paused. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@13604 b3059339-0415-0410-9bf9-f77b7e298cf2
2004-10-10 19:39:07 +02:00
/**
* \param peek_only when set, the returned command stays in the queue.
* Do not free the returned cmd whe you set this!
*/
input: remove unused "paused" arguments from some functions
2010-04-25 20:47:43 +02:00
mp_cmd_t *mp_input_get_cmd(struct input_ctx *ictx, int time, int peek_only)
Create a context for input.c state Start moving static variables to a context struct. Only autorepeat state is moved to the struct in this commit. mp_input_check_interrupt now requires the context variable. Change stream functions to pass it. It's still stored in a static variable in stream/.
2008-04-30 06:15:52 +02:00
{
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
input: free all queued/allocated commands on exit These were memory leaks in theory, though not in practice (all memory is free'd on exit anyway). However, it was still annoying when leak reporting is enabled. I'm not sure if there was an actual leak in check_autorepeat(), maybe not.
2013-02-03 17:43:47 +01:00
if (async_quit_request) {
struct mp_cmd *cmd = mp_input_parse_cmd(bstr0("quit 1"), "");
queue_add(&ictx->control_cmd_queue, cmd, true);
}
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
input: simplify previous commit a bit more Less mutable state = better.
2012-01-18 02:35:40 +01:00
if (ictx->control_cmd_queue.first || ictx->key_cmd_queue.first)
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
time = 0;
read_all_events(ictx, time);
struct cmd_queue *queue = &ictx->control_cmd_queue;
input: simplify previous commit a bit more Less mutable state = better.
2012-01-18 02:35:40 +01:00
if (!queue->first)
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
queue = &ictx->key_cmd_queue;
input: simplify previous commit a bit more Less mutable state = better.
2012-01-18 02:35:40 +01:00
if (!queue->first) {
input: minor simplification to mp_input_get_cmd()
2012-10-14 23:44:49 +02:00
struct mp_cmd *repeated = check_autorepeat(ictx);
if (repeated)
queue_add(queue, repeated, false);
}
input: make input queue thread safe If pthreads are enabled the input queue accesses are regulated by acquiring a mutex. This is useful for platforms like OS X, where the events are created in the cocoa thread and added to the queue to then be dequeued in the playloop thread.
2013-08-13 14:40:17 +02:00
struct mp_cmd *ret = queue_peek(queue);
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
if (ret && !peek_only) {
input: minor simplification to mp_input_get_cmd()
2012-10-14 23:44:49 +02:00
queue_remove(queue, ret);
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
if (ret->mouse_move) {
ictx->mouse_x = ret->mouse_x;
ictx->mouse_y = ret->mouse_y;
}
}
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_unlock(ictx);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
return ret;
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
}
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
void mp_input_get_mouse_pos(struct input_ctx *ictx, int *x, int *y)
{
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
*x = ictx->mouse_x;
*y = ictx->mouse_y;
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_unlock(ictx);
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
}
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
void mp_cmd_free(mp_cmd_t *cmd)
{
input: Use talloc for struct mp_cmd and things under it
2008-04-30 12:47:21 +02:00
talloc_free(cmd);
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
}
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
mp_cmd_t *mp_cmd_clone(mp_cmd_t *cmd)
{
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
mp_cmd_t *ret;
int i;
Added key autorepeat support. Options to enable/disable joystick and lirc and set autorepeat delay and rate. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4658 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-11 12:42:08 +01:00
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
if (!cmd)
return NULL;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
ret = talloc_memdup(NULL, cmd, sizeof(mp_cmd_t));
ret->name = talloc_strdup(ret, cmd->name);
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
for (i = 0; i < MP_CMD_MAX_ARGS; i++) {
if (cmd->args[i].type.type == &m_option_type_string)
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
ret->args[i].v.s = talloc_strdup(ret, cmd->args[i].v.s);
}
Added key autorepeat support. Options to enable/disable joystick and lirc and set autorepeat delay and rate. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4658 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-11 12:42:08 +01:00
input: actually copy sub commands This was missing from the previous commit. It worked by luck, because the sub-commands weren't freed either (as long as the original command was around), but this is proper. Also, set the original string for command lists (needed for input-test only).
2013-07-08 19:44:46 +02:00
if (cmd->id == MP_CMD_COMMAND_LIST) {
input: fix dangling pointer issues with multi-part commands This didn't setup the linked list of sub-commands for multi-part commands ("a ; b ; c") correctly. Also, the new commands were attached to the allocation of the _old_ command instead of the new one. (Wow, whatever the hell I was (not) thinking when I wrote this code.)
2013-07-24 19:43:31 +02:00
struct mp_cmd *prev = NULL;
input: actually copy sub commands This was missing from the previous commit. It worked by luck, because the sub-commands weren't freed either (as long as the original command was around), but this is proper. Also, set the original string for command lists (needed for input-test only).
2013-07-08 19:44:46 +02:00
for (struct mp_cmd *sub = cmd->args[0].v.p; sub; sub = sub->queue_next) {
sub = mp_cmd_clone(sub);
input: fix dangling pointer issues with multi-part commands This didn't setup the linked list of sub-commands for multi-part commands ("a ; b ; c") correctly. Also, the new commands were attached to the allocation of the _old_ command instead of the new one. (Wow, whatever the hell I was (not) thinking when I wrote this code.)
2013-07-24 19:43:31 +02:00
talloc_steal(ret, sub);
if (prev) {
prev->queue_next = sub;
} else {
ret->args[0].v.p = sub;
}
prev = sub;
input: actually copy sub commands This was missing from the previous commit. It worked by luck, because the sub-commands weren't freed either (as long as the original command was around), but this is proper. Also, set the original string for command lists (needed for input-test only).
2013-07-08 19:44:46 +02:00
}
}
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
return ret;
Added key autorepeat support. Options to enable/disable joystick and lirc and set autorepeat delay and rate. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4658 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-11 12:42:08 +01:00
}
input: support bindings with modifier keys for X input Add support for binding commands to modifier+key combinations like "Shift+Left" or "Ctrl+Alt+x", and support reading such combinations from the output window of X VOs. The recognized modifier names are Shift, Ctrl, Alt and Meta. Any combination of those and then a non-modifier key name, separated by '+', is accepted as a key name in input.conf. For non-special keys that produce characters shift is ignored as a modifier. For example "A" is handled as a key without modifiers even if you use shift to write the capital letter; 'a' vs 'A' already distinguishes the combinations with a normal keymap, and having separate 'a', 'Shift+A' and 'A' (written with caps lock for example) would bring more confusion than benefit. Currently reading the modifier+key combinations is only supported in the output window of those VOs that use x11_common.c event handling. It's not possible to input the key combinations in other VOs or in a terminal window.
2010-12-19 11:12:20 +01:00
int mp_input_get_key_from_name(const char *name)
input.c: Remove mp_input_ prefixes from static function names The prefixes make the code unnecessarily verbose for no real benefit.
2008-04-30 08:30:51 +02:00
{
input: support bindings with modifier keys for X input Add support for binding commands to modifier+key combinations like "Shift+Left" or "Ctrl+Alt+x", and support reading such combinations from the output window of X VOs. The recognized modifier names are Shift, Ctrl, Alt and Meta. Any combination of those and then a non-modifier key name, separated by '+', is accepted as a key name in input.conf. For non-special keys that produce characters shift is ignored as a modifier. For example "A" is handled as a key without modifiers even if you use shift to write the capital letter; 'a' vs 'A' already distinguishes the combinations with a normal keymap, and having separate 'a', 'Shift+A' and 'A' (written with caps lock for example) would bring more confusion than benefit. Currently reading the modifier+key combinations is only supported in the output window of those VOs that use x11_common.c event handling. It's not possible to input the key combinations in other VOs or in a terminal window.
2010-12-19 11:12:20 +01:00
int modifiers = 0;
const char *p;
cleanup: silence most of the clang warnings
2011-07-07 23:41:42 +02:00
while ((p = strchr(name, '+'))) {
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
for (struct key_name *m = modifier_names; m->name; m++)
bstr: rename bstr() function to bstr0(), and typedef bstr to struct bstr Replace all uses of bstr() with bstr0(). Also remove the ridiculous C++ workaround.
2012-07-28 23:47:42 +02:00
if (!bstrcasecmp(bstr0(m->name),
bstr: rename BSTR() -> bstr() Rename the BSTR() function to bstr(). The former caused a conflict with some Windows OS name, and it's no longer a macro so uppercase naming is less appropriate.
2011-07-27 07:38:12 +02:00
(struct bstr){(char *)name, p - name})) {
input: support bindings with modifier keys for X input Add support for binding commands to modifier+key combinations like "Shift+Left" or "Ctrl+Alt+x", and support reading such combinations from the output window of X VOs. The recognized modifier names are Shift, Ctrl, Alt and Meta. Any combination of those and then a non-modifier key name, separated by '+', is accepted as a key name in input.conf. For non-special keys that produce characters shift is ignored as a modifier. For example "A" is handled as a key without modifiers even if you use shift to write the capital letter; 'a' vs 'A' already distinguishes the combinations with a normal keymap, and having separate 'a', 'Shift+A' and 'A' (written with caps lock for example) would bring more confusion than benefit. Currently reading the modifier+key combinations is only supported in the output window of those VOs that use x11_common.c event handling. It's not possible to input the key combinations in other VOs or in a terminal window.
2010-12-19 11:12:20 +01:00
modifiers |= m->key;
goto found;
}
if (!strcmp(name, "+"))
return '+' + modifiers;
return -1;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
found:
input: support bindings with modifier keys for X input Add support for binding commands to modifier+key combinations like "Shift+Left" or "Ctrl+Alt+x", and support reading such combinations from the output window of X VOs. The recognized modifier names are Shift, Ctrl, Alt and Meta. Any combination of those and then a non-modifier key name, separated by '+', is accepted as a key name in input.conf. For non-special keys that produce characters shift is ignored as a modifier. For example "A" is handled as a key without modifiers even if you use shift to write the capital letter; 'a' vs 'A' already distinguishes the combinations with a normal keymap, and having separate 'a', 'Shift+A' and 'A' (written with caps lock for example) would bring more confusion than benefit. Currently reading the modifier+key combinations is only supported in the output window of those VOs that use x11_common.c event handling. It's not possible to input the key combinations in other VOs or in a terminal window.
2010-12-19 11:12:20 +01:00
name = p + 1;
}
input: allow unicode keys and reassign internal key codes This moves all key codes above the highest valid unicode code point (which is 0x10FFFF). All key codes below MP_KEY_BASE now directly map to unicode. Configuration files (input.conf) can contain unicode characters in UTF-8 to map non-ASCII characters/keys. This shouldn't change anything user visible, except that "direct key codes" (as used in input.conf) will change their meaning.
2012-01-14 14:09:26 +01:00
bstr: rename bstr() function to bstr0(), and typedef bstr to struct bstr Replace all uses of bstr() with bstr0(). Also remove the ridiculous C++ workaround.
2012-07-28 23:47:42 +02:00
struct bstr bname = bstr0(name);
input: allow unicode keys and reassign internal key codes This moves all key codes above the highest valid unicode code point (which is 0x10FFFF). All key codes below MP_KEY_BASE now directly map to unicode. Configuration files (input.conf) can contain unicode characters in UTF-8 to map non-ASCII characters/keys. This shouldn't change anything user visible, except that "direct key codes" (as used in input.conf) will change their meaning.
2012-01-14 14:09:26 +01:00
struct bstr rest;
int code = bstr_decode_utf8(bname, &rest);
if (code >= 0 && rest.len == 0)
return code + modifiers;
if (bstr_startswith0(bname, "0x"))
input: support bindings with modifier keys for X input Add support for binding commands to modifier+key combinations like "Shift+Left" or "Ctrl+Alt+x", and support reading such combinations from the output window of X VOs. The recognized modifier names are Shift, Ctrl, Alt and Meta. Any combination of those and then a non-modifier key name, separated by '+', is accepted as a key name in input.conf. For non-special keys that produce characters shift is ignored as a modifier. For example "A" is handled as a key without modifiers even if you use shift to write the capital letter; 'a' vs 'A' already distinguishes the combinations with a normal keymap, and having separate 'a', 'Shift+A' and 'A' (written with caps lock for example) would bring more confusion than benefit. Currently reading the modifier+key combinations is only supported in the output window of those VOs that use x11_common.c event handling. It's not possible to input the key combinations in other VOs or in a terminal window.
2010-12-19 11:12:20 +01:00
return strtol(name, NULL, 16) + modifiers;
for (int i = 0; key_names[i].name != NULL; i++) {
if (strcasecmp(key_names[i].name, name) == 0)
return key_names[i].key + modifiers;
}
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
input: support bindings with modifier keys for X input Add support for binding commands to modifier+key combinations like "Shift+Left" or "Ctrl+Alt+x", and support reading such combinations from the output window of X VOs. The recognized modifier names are Shift, Ctrl, Alt and Meta. Any combination of those and then a non-modifier key name, separated by '+', is accepted as a key name in input.conf. For non-special keys that produce characters shift is ignored as a modifier. For example "A" is handled as a key without modifiers even if you use shift to write the capital letter; 'a' vs 'A' already distinguishes the combinations with a normal keymap, and having separate 'a', 'Shift+A' and 'A' (written with caps lock for example) would bring more confusion than benefit. Currently reading the modifier+key combinations is only supported in the output window of those VOs that use x11_common.c event handling. It's not possible to input the key combinations in other VOs or in a terminal window.
2010-12-19 11:12:20 +01:00
return -1;
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
}
input: store number of keys, instead of using 0-termination
2013-06-23 00:07:45 +02:00
static int get_input_from_name(char *name, int *out_num_keys, int *keys)
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
{
char *end, *ptr;
int n = 0;
ptr = name;
n = 0;
for (end = strchr(ptr, '-'); ptr != NULL; end = strchr(ptr, '-')) {
if (end && end[1] != '\0') {
if (end[1] == '-')
end = &end[1];
end[0] = '\0';
}
keys[n] = mp_input_get_key_from_name(ptr);
if (keys[n] < 0)
return 0;
n++;
if (end && end[1] != '\0' && n < MP_MAX_KEY_DOWN)
ptr = &end[1];
else
break;
Added support for key combination and mouse buttons key code git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4590 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-08 19:10:56 +01:00
}
input: store number of keys, instead of using 0-termination
2013-06-23 00:07:45 +02:00
*out_num_keys = n;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
return 1;
Added support for key combination and mouse buttons key code git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4590 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-08 19:10:56 +01:00
}
input: store number of binds, instead of using 0-termination
2013-06-23 00:40:35 +02:00
static void bind_dealloc(struct cmd_bind *bind)
{
talloc_free(bind->cmd);
talloc_free(bind->location);
}
input: warn about non-existing commands, add explicit "ignore" command If a command is not found, warn about it at loading time (just like other command parsing errors are printed at loading time). Add an explicit "ignore" command. input.conf instructs users to use this command to cancel out existing mapping. This clashed with the warning added in this commit. Make "ignore" a real command and remove the specialcasing for it from get_cmd_from_keys(). Now "ignore" is ignored because it's not handled in command.c.
2012-09-22 15:17:15 +02:00
static void bind_keys(struct input_ctx *ictx, bool builtin, bstr section,
input: store number of keys, instead of using 0-termination
2013-06-23 00:07:45 +02:00
const int *keys, int num_keys, bstr command,
input: track location info for input commands If parsing a command fails, its location is printed. The location is the path to the input.conf, and the line number of the key binding and the associated input command.
2012-10-13 21:09:42 +02:00
const char *loc)
input: Move command bind variables to context struct
2008-04-30 07:56:45 +02:00
{
input: store number of binds, instead of using 0-termination
2013-06-23 00:40:35 +02:00
struct cmd_bind_section *bs = get_bind_section(ictx, section);
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
struct cmd_bind *bind = NULL;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
input: store number of keys, instead of using 0-termination
2013-06-23 00:07:45 +02:00
assert(num_keys <= MP_MAX_KEY_DOWN);
input: store number of binds, instead of using 0-termination
2013-06-23 00:40:35 +02:00
for (int n = 0; n < bs->num_binds; n++) {
struct cmd_bind *b = &bs->binds[n];
if (bind_matches_key(b, num_keys, keys) && b->is_builtin == builtin) {
bind = b;
break;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
}
}
if (!bind) {
input: store number of binds, instead of using 0-termination
2013-06-23 00:40:35 +02:00
struct cmd_bind empty = {{0}};
MP_TARRAY_APPEND(bs, bs->binds, bs->num_binds, empty);
bind = &bs->binds[bs->num_binds - 1];
Really take tab as a space in cmd parsing. Now config file override the defaults bindings :) git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4843 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-24 17:31:27 +01:00
}
input: store number of binds, instead of using 0-termination
2013-06-23 00:40:35 +02:00
bind_dealloc(bind);
*bind = (struct cmd_bind) {
.cmd = bstrdup0(bs->binds, command),
.location = talloc_strdup(bs->binds, loc),
.owner = bs,
.is_builtin = builtin,
.num_keys = num_keys,
};
input: store number of keys, instead of using 0-termination
2013-06-23 00:07:45 +02:00
memcpy(bind->keys, keys, num_keys * sizeof(bind->keys[0]));
Really take tab as a space in cmd parsing. Now config file override the defaults bindings :) git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4843 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-02-24 17:31:27 +01:00
}
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
// restrict_section: every entry is forced to this section name
// if NULL, load normally and allow any sections
input: track location info for input commands If parsing a command fails, its location is printed. The location is the path to the input.conf, and the line number of the key binding and the associated input command.
2012-10-13 21:09:42 +02:00
static int parse_config(struct input_ctx *ictx, bool builtin, bstr data,
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
const char *location, const char *restrict_section)
input: Move command bind variables to context struct
2008-04-30 07:56:45 +02:00
{
input: store number of keys, instead of using 0-termination
2013-06-23 00:07:45 +02:00
int n_binds = 0;
input: track location info for input commands If parsing a command fails, its location is printed. The location is the path to the input.conf, and the line number of the key binding and the associated input command.
2012-10-13 21:09:42 +02:00
int line_no = 0;
char *cur_loc = NULL;
input: fix input.conf parse errors Commit df899f59be removing a write outside a buffer triggered another problem, as for some reason the code did not 0-terminate its read buffer in the specific case that it had encountered an EOF, and as a result could parse contents left in the buffer for a second time. Usually this resulted in parsing error messages. Fix the problem by rewriting the offending code in a less hacky form.
2011-07-20 03:41:51 +02:00
input: simplify input.conf parsing Now input.conf is loaded into memory at once, instead of streaming the file into the parser. The real reason for this change is that I want to be able to read the config file from memory. (Using fmemopen() would have been simpler, but that is available on sane platforms only.)
2012-08-01 21:50:24 +02:00
while (data.len) {
input: track location info for input commands If parsing a command fails, its location is printed. The location is the path to the input.conf, and the line number of the key binding and the associated input command.
2012-10-13 21:09:42 +02:00
line_no++;
if (cur_loc)
talloc_free(cur_loc);
cur_loc = talloc_asprintf(NULL, "%s:%d", location, line_no);
input: simplify input.conf parsing Now input.conf is loaded into memory at once, instead of streaming the file into the parser. The real reason for this change is that I want to be able to read the config file from memory. (Using fmemopen() would have been simpler, but that is available on sane platforms only.)
2012-08-01 21:50:24 +02:00
bstr line = bstr_strip_linebreaks(bstr_getline(data, &data));
line = bstr_lstrip(line);
if (line.len == 0 || bstr_startswith0(line, "#"))
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
continue;
input: fix input.conf parse errors Commit df899f59be removing a write outside a buffer triggered another problem, as for some reason the code did not 0-terminate its read buffer in the specific case that it had encountered an EOF, and as a result could parse contents left in the buffer for a second time. Usually this resulted in parsing error messages. Fix the problem by rewriting the offending code in a less hacky form.
2011-07-20 03:41:51 +02:00
struct bstr command;
// Find the key name starting a line
input: change bind_keys() to use bstr This seems rather pointless considering there are still stupid bstrdup0()s left, but maybe this is the right direction.
2012-08-01 23:41:21 +02:00
struct bstr keyname = bstr_split(line, WHITESPACE, &command);
input: fix input.conf parse errors Commit df899f59be removing a write outside a buffer triggered another problem, as for some reason the code did not 0-terminate its read buffer in the specific case that it had encountered an EOF, and as a result could parse contents left in the buffer for a second time. Usually this resulted in parsing error messages. Fix the problem by rewriting the offending code in a less hacky form.
2011-07-20 03:41:51 +02:00
command = bstr_strip(command);
if (command.len == 0) {
mp_tmsg(MSGT_INPUT, MSGL_ERR,
input: track location info for input commands If parsing a command fails, its location is printed. The location is the path to the input.conf, and the line number of the key binding and the associated input command.
2012-10-13 21:09:42 +02:00
"Unfinished key binding: %.*s at %s\n", BSTR_P(line),
cur_loc);
input: fix input.conf parse errors Commit df899f59be removing a write outside a buffer triggered another problem, as for some reason the code did not 0-terminate its read buffer in the specific case that it had encountered an EOF, and as a result could parse contents left in the buffer for a second time. Usually this resulted in parsing error messages. Fix the problem by rewriting the offending code in a less hacky form.
2011-07-20 03:41:51 +02:00
continue;
}
char *name = bstrdup0(NULL, keyname);
input: store number of keys, instead of using 0-termination
2013-06-23 00:07:45 +02:00
int keys[MP_MAX_KEY_DOWN];
int num_keys = 0;
if (!get_input_from_name(name, &num_keys, keys)) {
input: fix input.conf parse errors Commit df899f59be removing a write outside a buffer triggered another problem, as for some reason the code did not 0-terminate its read buffer in the specific case that it had encountered an EOF, and as a result could parse contents left in the buffer for a second time. Usually this resulted in parsing error messages. Fix the problem by rewriting the offending code in a less hacky form.
2011-07-20 03:41:51 +02:00
talloc_free(name);
mp_tmsg(MSGT_INPUT, MSGL_ERR,
input: track location info for input commands If parsing a command fails, its location is printed. The location is the path to the input.conf, and the line number of the key binding and the associated input command.
2012-10-13 21:09:42 +02:00
"Unknown key '%.*s' at %s\n", BSTR_P(keyname), cur_loc);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
continue;
input: support bindings with modifier keys for X input Add support for binding commands to modifier+key combinations like "Shift+Left" or "Ctrl+Alt+x", and support reading such combinations from the output window of X VOs. The recognized modifier names are Shift, Ctrl, Alt and Meta. Any combination of those and then a non-modifier key name, separated by '+', is accepted as a key name in input.conf. For non-special keys that produce characters shift is ignored as a modifier. For example "A" is handled as a key without modifiers even if you use shift to write the capital letter; 'a' vs 'A' already distinguishes the combinations with a normal keymap, and having separate 'a', 'Shift+A' and 'A' (written with caps lock for example) would bring more confusion than benefit. Currently reading the modifier+key combinations is only supported in the output window of those VOs that use x11_common.c event handling. It's not possible to input the key combinations in other VOs or in a terminal window.
2010-12-19 11:12:20 +01:00
}
input: fix input.conf parse errors Commit df899f59be removing a write outside a buffer triggered another problem, as for some reason the code did not 0-terminate its read buffer in the specific case that it had encountered an EOF, and as a result could parse contents left in the buffer for a second time. Usually this resulted in parsing error messages. Fix the problem by rewriting the offending code in a less hacky form.
2011-07-20 03:41:51 +02:00
talloc_free(name);
input: warn about non-existing commands, add explicit "ignore" command If a command is not found, warn about it at loading time (just like other command parsing errors are printed at loading time). Add an explicit "ignore" command. input.conf instructs users to use this command to cancel out existing mapping. This clashed with the warning added in this commit. Make "ignore" a real command and remove the specialcasing for it from get_cmd_from_keys(). Now "ignore" is ignored because it's not handled in command.c.
2012-09-22 15:17:15 +02:00
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
bstr section = bstr0(restrict_section);
if (!section.len) {
if (bstr_startswith0(command, "{")) {
int p = bstrchr(command, '}');
if (p != -1) {
section = bstr_strip(bstr_splice(command, 1, p));
command = bstr_lstrip(bstr_cut(command, p + 1));
}
input: warn about non-existing commands, add explicit "ignore" command If a command is not found, warn about it at loading time (just like other command parsing errors are printed at loading time). Add an explicit "ignore" command. input.conf instructs users to use this command to cancel out existing mapping. This clashed with the warning added in this commit. Make "ignore" a real command and remove the specialcasing for it from get_cmd_from_keys(). Now "ignore" is ignored because it's not handled in command.c.
2012-09-22 15:17:15 +02:00
}
}
input: store number of keys, instead of using 0-termination
2013-06-23 00:07:45 +02:00
bind_keys(ictx, builtin, section, keys, num_keys, command, cur_loc);
input: fix input.conf parse errors Commit df899f59be removing a write outside a buffer triggered another problem, as for some reason the code did not 0-terminate its read buffer in the specific case that it had encountered an EOF, and as a result could parse contents left in the buffer for a second time. Usually this resulted in parsing error messages. Fix the problem by rewriting the offending code in a less hacky form.
2011-07-20 03:41:51 +02:00
n_binds++;
input: verify input.conf on loading When input.conf is loaded, verify each command and print a warning if it's invalid or uses legacy commands. This is done for both the user's and the embedded config files. The diff is a bit noisy, because mp_input_parse_cmd() is changed to take a bstr as argument instead of a char*.
2012-09-22 06:33:04 +02:00
// Print warnings if invalid commands are encountered.
input: track location info for input commands If parsing a command fails, its location is printed. The location is the path to the input.conf, and the line number of the key binding and the associated input command.
2012-10-13 21:09:42 +02:00
talloc_free(mp_input_parse_cmd(command, cur_loc));
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
}
input: simplify input.conf parsing Now input.conf is loaded into memory at once, instead of streaming the file into the parser. The real reason for this change is that I want to be able to read the config file from memory. (Using fmemopen() would have been simpler, but that is available on sane platforms only.)
2012-08-01 21:50:24 +02:00
input: track location info for input commands If parsing a command fails, its location is printed. The location is the path to the input.conf, and the line number of the key binding and the associated input command.
2012-10-13 21:09:42 +02:00
talloc_free(cur_loc);
input: simplify input.conf parsing Now input.conf is loaded into memory at once, instead of streaming the file into the parser. The real reason for this change is that I want to be able to read the config file from memory. (Using fmemopen() would have been simpler, but that is available on sane platforms only.)
2012-08-01 21:50:24 +02:00
return n_binds;
}
options: change --no-config option, make it apply to input.conf as well Simplify --no-config and make it a normal flag option, and doesn't take an argument anymore. You can get the same behavior by using --no-config and then --include to explicitly load a certain config file. Make --no-config work for input.conf as well. Make it so that --input:conf=file still works in this case. As a technically unrelated change, the file argument now works as one would expect, instead of making it relatively to "~/.mpv/". This makes for simpler code and easier to understand option semantics. We can also print better error messages.
2013-02-08 23:52:06 +01:00
static int parse_config_file(struct input_ctx *ictx, char *file, bool warn)
input: simplify input.conf parsing Now input.conf is loaded into memory at once, instead of streaming the file into the parser. The real reason for this change is that I want to be able to read the config file from memory. (Using fmemopen() would have been simpler, but that is available on sane platforms only.)
2012-08-01 21:50:24 +02:00
{
input: silence warning if input.conf is missing It's silly to print a warning if an optional config file is missing. Don't print anything at the default message level if an input config is not found. Unfortunately, the behavior is the same for explicitly passed input config files (with --input=conf=file.conf).
2012-11-15 12:40:39 +01:00
if (!mp_path_exists(file)) {
options: change --no-config option, make it apply to input.conf as well Simplify --no-config and make it a normal flag option, and doesn't take an argument anymore. You can get the same behavior by using --no-config and then --include to explicitly load a certain config file. Make --no-config work for input.conf as well. Make it so that --input:conf=file still works in this case. As a technically unrelated change, the file argument now works as one would expect, instead of making it relatively to "~/.mpv/". This makes for simpler code and easier to understand option semantics. We can also print better error messages.
2013-02-08 23:52:06 +01:00
mp_msg(MSGT_INPUT, warn ? MSGL_ERR : MSGL_V,
"Input config file %s not found.\n", file);
input: silence warning if input.conf is missing It's silly to print a warning if an optional config file is missing. Don't print anything at the default message level if an input config is not found. Unfortunately, the behavior is the same for explicitly passed input config files (with --input=conf=file.conf).
2012-11-15 12:40:39 +01:00
return 0;
}
core: change open_stream and demux_open signature This removes the dependency on DEMUXER_TYPE_* and the file_format parameter from the stream open functions. Remove some of the playlist handling code. It looks like this was needed only for loading linked mov files with demux_mov (which was removed long ago). Delete a minor bit of dead network-related code from stream.c as well.
2013-07-11 21:10:42 +02:00
stream_t *s = stream_open(file, NULL);
input: simplify input.conf parsing Now input.conf is loaded into memory at once, instead of streaming the file into the parser. The real reason for this change is that I want to be able to read the config file from memory. (Using fmemopen() would have been simpler, but that is available on sane platforms only.)
2012-08-01 21:50:24 +02:00
if (!s) {
options: change --no-config option, make it apply to input.conf as well Simplify --no-config and make it a normal flag option, and doesn't take an argument anymore. You can get the same behavior by using --no-config and then --include to explicitly load a certain config file. Make --no-config work for input.conf as well. Make it so that --input:conf=file still works in this case. As a technically unrelated change, the file argument now works as one would expect, instead of making it relatively to "~/.mpv/". This makes for simpler code and easier to understand option semantics. We can also print better error messages.
2013-02-08 23:52:06 +01:00
mp_msg(MSGT_INPUT, MSGL_ERR, "Can't open input config file %s.\n", file);
input: simplify input.conf parsing Now input.conf is loaded into memory at once, instead of streaming the file into the parser. The real reason for this change is that I want to be able to read the config file from memory. (Using fmemopen() would have been simpler, but that is available on sane platforms only.)
2012-08-01 21:50:24 +02:00
return 0;
}
stream: remove padding parameter from stream_read_complete() Seems like a completely unnecessary complication. Instead, always add a 1 byte padding (could be extended if a caller needs it), and clear it. Also add some documentation. There was some, but it was outdated and incomplete.
2013-06-11 12:16:42 +02:00
bstr res = stream_read_complete(s, NULL, 1000000);
input: simplify input.conf parsing Now input.conf is loaded into memory at once, instead of streaming the file into the parser. The real reason for this change is that I want to be able to read the config file from memory. (Using fmemopen() would have been simpler, but that is available on sane platforms only.)
2012-08-01 21:50:24 +02:00
free_stream(s);
mp_msg(MSGT_INPUT, MSGL_V, "Parsing input config file %s\n", file);
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
int n_binds = parse_config(ictx, false, res, file, NULL);
input: simplify input.conf parsing Now input.conf is loaded into memory at once, instead of streaming the file into the parser. The real reason for this change is that I want to be able to read the config file from memory. (Using fmemopen() would have been simpler, but that is available on sane platforms only.)
2012-08-01 21:50:24 +02:00
talloc_free(res.start);
mp_msg(MSGT_INPUT, MSGL_V, "Input config file %s parsed: %d binds\n",
file, n_binds);
return 1;
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
}
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
// If name is NULL, return "default".
// Return a statically allocated name of the section (i.e. return value never
// gets deallocated).
static char *normalize_section(struct input_ctx *ictx, char *name)
{
return get_bind_section(ictx, bstr0(name))->section;
}
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
void mp_input_disable_section(struct input_ctx *ictx, char *name)
{
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
name = normalize_section(ictx, name);
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
// Remove old section, or make sure it's on top if re-enabled
for (int i = ictx->num_active_sections - 1; i >= 0; i--) {
struct active_section *as = &ictx->active_sections[i];
if (strcmp(as->name, name) == 0) {
for (int x = i; i < ictx->num_active_sections - 1; i++)
ictx->active_sections[x] = ictx->active_sections[x + 1];
ictx->num_active_sections--;
}
}
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_unlock(ictx);
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
}
void mp_input_enable_section(struct input_ctx *ictx, char *name, int flags)
{
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
name = normalize_section(ictx, name);
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
mp_input_disable_section(ictx, name);
if (ictx->num_active_sections < MAX_ACTIVE_SECTIONS) {
ictx->active_sections[ictx->num_active_sections++] =
(struct active_section) {name, flags};
}
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_unlock(ictx);
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
}
void mp_input_disable_all_sections(struct input_ctx *ictx)
{
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
ictx->num_active_sections = 0;
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_unlock(ictx);
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
}
void mp_input_set_section_mouse_area(struct input_ctx *ictx, char *name,
int x0, int y0, int x1, int y1)
input: Move command bind variables to context struct
2008-04-30 07:56:45 +02:00
{
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
struct cmd_bind_section *s = get_bind_section(ictx, bstr0(name));
s->mouse_area = (struct mp_rect){x0, y0, x1, y1};
s->mouse_area_set = x0 != x1 && y0 != y1;
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_unlock(ictx);
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
}
bool mp_input_test_mouse_active(struct input_ctx *ictx, int x, int y)
{
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
bool res = false;
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
for (int i = 0; i < ictx->num_active_sections; i++) {
char *name = ictx->active_sections[i].name;
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
struct cmd_bind_section *s = get_bind_section(ictx, bstr0(name));
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
if (s->mouse_area_set && test_rect(&s->mouse_area, x, y)) {
res = true;
break;
}
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
}
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_unlock(ictx);
return res;
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
}
bool mp_input_test_dragging(struct input_ctx *ictx, int x, int y)
{
return mp_input_test_mouse_active(ictx, x, y);
implement "bind sections" allows to bind one input key to several slave commands (independently for each section) as shown below: RIGHT seek +10 RIGHT {tv} tv_step_channel 1 RIGHT {dvdnav} dvdnav 4 Currenlty only "tv" section added. patch by Otvos Attila oattila at chello dot hu git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@23497 b3059339-0415-0410-9bf9-f77b7e298cf2
2007-06-07 20:06:53 +02:00
}
input: store number of binds, instead of using 0-termination
2013-06-23 00:40:35 +02:00
// builtin: if true, remove all builtin binds, else remove all user binds
static void remove_binds(struct cmd_bind_section *bs, bool builtin)
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
{
input: store number of binds, instead of using 0-termination
2013-06-23 00:40:35 +02:00
for (int n = bs->num_binds - 1; n >= 0; n--) {
if (bs->binds[n].is_builtin == builtin) {
bind_dealloc(&bs->binds[n]);
assert(bs->num_binds >= 1);
bs->binds[n] = bs->binds[bs->num_binds - 1];
bs->num_binds--;
input: don't keep separate sections for builtin key bindings Instead mark individual key bindings as builtin. Not sure whether this is conceptually simpler or more complicated. For one, it requires the annoying remove_binds() function to wipe existing bindings instead of just killing the section, on the other hand it gets rid of almost all special handling of builtin vs. normal sections.
2013-06-22 23:26:44 +02:00
}
}
}
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
void mp_input_define_section(struct input_ctx *ictx, char *name, char *location,
char *contents, bool builtin)
input: Move command bind variables to context struct
2008-04-30 07:56:45 +02:00
{
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
if (!name || !name[0])
return; // parse_config() changes semantics with restrict_section==empty
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
if (contents) {
parse_config(ictx, builtin, bstr0(contents), location, name);
} else {
// Disable:
mp_input_disable_section(ictx, name);
// Delete:
input: store number of binds, instead of using 0-termination
2013-06-23 00:40:35 +02:00
struct cmd_bind_section *bs = get_bind_section(ictx, bstr0(name));
remove_binds(bs, builtin);
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
}
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_unlock(ictx);
implement "bind sections" allows to bind one input key to several slave commands (independently for each section) as shown below: RIGHT seek +10 RIGHT {tv} tv_step_channel 1 RIGHT {dvdnav} dvdnav 4 Currenlty only "tv" section added. patch by Otvos Attila oattila at chello dot hu git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@23497 b3059339-0415-0410-9bf9-f77b7e298cf2
2007-06-07 20:06:53 +02:00
}
m_config: refactor some things Change how m_config is initialized. Make it more uniform; now all m_config structs are intialized in exactly the same way. Make sure there's only a single m_option[] array defining the options, and keep around the pointer to the optstruct default value, and the optstruct size as well. This will allow reconstructing the option default values in the following commit. In particular, stop pretending that the handling of some special options (like --profile, --v, and some others) is in any way elegant, and make them explicit hacks. This is really more readable and easier to understand than what was before, and simplifies the code.
2013-07-27 21:26:00 +02:00
struct input_ctx *mp_input_init(struct MPOpts *opts)
Create a context for input.c state Start moving static variables to a context struct. Only autorepeat state is moved to the struct in this commit. mp_input_check_interrupt now requires the context variable. Change stream functions to pass it. It's still stored in a static variable in stream/.
2008-04-30 06:15:52 +02:00
{
m_config: refactor some things Change how m_config is initialized. Make it more uniform; now all m_config structs are intialized in exactly the same way. Make sure there's only a single m_option[] array defining the options, and keep around the pointer to the optstruct default value, and the optstruct size as well. This will allow reconstructing the option default values in the following commit. In particular, stop pretending that the handling of some special options (like --profile, --v, and some others) is in any way elegant, and make them explicit hacks. This is really more readable and easier to understand than what was before, and simplifies the code.
2013-07-27 21:26:00 +02:00
struct input_conf *input_conf = &opts->input;
Create a context for input.c state Start moving static variables to a context struct. Only autorepeat state is moved to the struct in this commit. mp_input_check_interrupt now requires the context variable. Change stream functions to pass it. It's still stored in a static variable in stream/.
2008-04-30 06:15:52 +02:00
struct input_ctx *ictx = talloc_ptrtype(NULL, ictx);
*ictx = (struct input_ctx){
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
.key_fifo_size = input_conf->key_fifo_size,
core: remove mp_fifo indirection For some reason mp_fifo specifically handled double clicks, and other than that was a pointless wrapper around input.c functionality. Move the double click handling into input.c, and get rid of mp_fifo. Add some compatibility wrappers, because so much VO code uses these functions. Where struct mp_fifo is still used it's just a casted struct input_ctx.
2013-07-02 14:00:24 +02:00
.doubleclick_time = input_conf->doubleclick_time,
Create a context for input.c state Start moving static variables to a context struct. Only autorepeat state is moved to the struct in this commit. mp_input_check_interrupt now requires the context variable. Change stream functions to pass it. It's still stored in a static variable in stream/.
2008-04-30 06:15:52 +02:00
.ar_state = -1,
Move input options to option struct
2008-04-30 17:57:02 +02:00
.ar_delay = input_conf->ar_delay,
.ar_rate = input_conf->ar_rate,
Merge svn changes up to r29117
2009-04-01 01:26:34 +02:00
.default_bindings = input_conf->default_bindings,
input: stay in current input section if mouse button is down Normally, moving the mouse outside of the mouse area of an input section will send mouse events somewhere else (because input section mouse areas are similar to windows/widgets in real GUI toolkits). This was done even if a mouse button was held down. This is quite different from how GUI toolkits behave. Change the code so that if a mouse button is down, the mouse area of the current input section can't be left. Releasing the mouse button (while the mouse pointer is outside of the mouse area) will actually leave the mouse area. As a side-effect, this commit also tests more often whether the current mouse input section is valid. This is needed to make releasing a mouse button trigger the mouse input section change.
2013-08-30 02:24:14 +02:00
.mouse_section = "default",
input: add input test mode In input test mode, key bindings won't be executed, but are shown on the OSD. The OSD includes various information, such as the name of the key, the command itself, whether it's builtin, and the config file location it was defined. The input test mode can be enabled with "--input=test". No effort is spent trying to react to key bindings that normally exit the player; they are treated just like any other binding.
2012-10-13 21:10:20 +02:00
.test = input_conf->test,
ao_pulse, core: make pulse thread wake up core for more data For ao_pulse, the current latency is not a good indicator of how soon the AO requires new data to avoid underflow. Add an internal pipe that can be used to wake up the input loop from select(), and make the pulseaudio main loop (which runs in a separate thread) use this mechanism when pulse requests more data. The wakeup signal currently contains no information about the reason for the wakup, but audio buffers are always filled when the event loop wakes up. Also, request a latency of 1 second from the Pulseaudio server. The default is normally significantly higher. We don't need low latency, while higher latency helps prevent underflows reduces need for wakeups.
2012-03-25 21:58:48 +02:00
.wakeup_pipe = {-1, -1},
Create a context for input.c state Start moving static variables to a context struct. Only autorepeat state is moved to the struct in this commit. mp_input_check_interrupt now requires the context variable. Change stream functions to pass it. It's still stored in a static variable in stream/.
2008-04-30 06:15:52 +02:00
};
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
#if HAVE_PTHREADS
pthread_mutexattr_t attr;
pthread_mutexattr_init(&attr);
pthread_mutexattr_settype(&attr, PTHREAD_MUTEX_RECURSIVE);
pthread_mutex_init(&ictx->mutex, &attr);
pthread_mutexattr_destroy(&attr);
#endif
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
mp_input_enable_section(ictx, NULL, 0);
input: refactor: store builtin key binds along user user binds Builtin (i.e. default) binds are still separately handled, and this commit shouldn't change any user-visible behavior.
2012-08-01 22:45:35 +02:00
input: comment all default bindings in input.conf I suspect most users will just copy etc/input.conf when they want to remap some default bindings. But usually this means the user even copies bindings he doesn't care about, and it's better if the user maps only the bindings in his input.conf the user intends to remap. Comment all bindings in etc/input.conf. Since this file also defines the builtin defaults and is baked into the mpv binary, we have to do something to get them anyway, even though they are commented. Do this by having input.c "uncomment" the bindings in the baked in input.conf. (Of course this is done only for the builtin config, not configs loaded from disk.)
2013-09-01 17:06:11 +02:00
// "Uncomment" the default key bindings in etc/input.conf and add them.
// All lines that do not start with '# ' are parsed.
bstr builtin = bstr0(builtin_input_conf);
while (builtin.len) {
bstr line = bstr_getline(builtin, &builtin);
bstr_eatstart0(&line, "#");
if (!bstr_startswith0(line, " "))
parse_config(ictx, true, line, "<builtin>", NULL);
}
Create a context for input.c state Start moving static variables to a context struct. Only autorepeat state is moved to the struct in this commit. mp_input_check_interrupt now requires the context variable. Change stream functions to pass it. It's still stored in a static variable in stream/.
2008-04-30 06:15:52 +02:00
ao_pulse, core: make pulse thread wake up core for more data For ao_pulse, the current latency is not a good indicator of how soon the AO requires new data to avoid underflow. Add an internal pipe that can be used to wake up the input loop from select(), and make the pulseaudio main loop (which runs in a separate thread) use this mechanism when pulse requests more data. The wakeup signal currently contains no information about the reason for the wakup, but audio buffers are always filled when the event loop wakes up. Also, request a latency of 1 second from the Pulseaudio server. The default is normally significantly higher. We don't need low latency, while higher latency helps prevent underflows reduces need for wakeups.
2012-03-25 21:58:48 +02:00
#ifndef __MINGW32__
long ret = pipe(ictx->wakeup_pipe);
for (int i = 0; i < 2 && ret >= 0; i++) {
ret = fcntl(ictx->wakeup_pipe[i], F_GETFL);
if (ret < 0)
break;
ret = fcntl(ictx->wakeup_pipe[i], F_SETFL, ret | O_NONBLOCK);
}
if (ret < 0)
mp_msg(MSGT_INPUT, MSGL_ERR,
"Failed to initialize wakeup pipe: %s\n", strerror(errno));
else
mp_input_add_key_fd(ictx, ictx->wakeup_pipe[0], true, read_wakeup,
NULL, NULL);
#endif
options: change --no-config option, make it apply to input.conf as well Simplify --no-config and make it a normal flag option, and doesn't take an argument anymore. You can get the same behavior by using --no-config and then --include to explicitly load a certain config file. Make --no-config work for input.conf as well. Make it so that --input:conf=file still works in this case. As a technically unrelated change, the file argument now works as one would expect, instead of making it relatively to "~/.mpv/". This makes for simpler code and easier to understand option semantics. We can also print better error messages.
2013-02-08 23:52:06 +01:00
bool config_ok = false;
if (input_conf->config_file)
config_ok = parse_config_file(ictx, input_conf->config_file, true);
m_config: refactor some things Change how m_config is initialized. Make it more uniform; now all m_config structs are intialized in exactly the same way. Make sure there's only a single m_option[] array defining the options, and keep around the pointer to the optstruct default value, and the optstruct size as well. This will allow reconstructing the option default values in the following commit. In particular, stop pretending that the handling of some special options (like --profile, --v, and some others) is in any way elegant, and make them explicit hacks. This is really more readable and easier to understand than what was before, and simplifies the code.
2013-07-27 21:26:00 +02:00
if (!config_ok && opts->load_config) {
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
// Try global conf dir
options: change --no-config option, make it apply to input.conf as well Simplify --no-config and make it a normal flag option, and doesn't take an argument anymore. You can get the same behavior by using --no-config and then --include to explicitly load a certain config file. Make --no-config work for input.conf as well. Make it so that --input:conf=file still works in this case. As a technically unrelated change, the file argument now works as one would expect, instead of making it relatively to "~/.mpv/". This makes for simpler code and easier to understand option semantics. We can also print better error messages.
2013-02-08 23:52:06 +01:00
char *file = mp_find_config_file("input.conf");
config_ok = file && parse_config_file(ictx, file, false);
talloc_free(file);
}
if (!config_ok) {
mp_msg(MSGT_INPUT, MSGL_V, "Falling back on default (hardcoded) "
"input config\n");
Corrected my mistake in the last memfix patch. Bug was discovered by Shachar Raindel <shacharr@gmail.com>. The basic patch idea is also from him. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@12943 b3059339-0415-0410-9bf9-f77b7e298cf2
2004-08-03 14:21:14 +02:00
}
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
Rename a bunch of miscellaneous preprocessor directives. Switch them from a HAVE_ to a CONFIG_ prefix. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@27423 b3059339-0415-0410-9bf9-f77b7e298cf2
2008-08-07 10:58:07 +02:00
#ifdef CONFIG_JOYSTICK
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
if (input_conf->use_joystick) {
int fd = mp_input_joystick_init(input_conf->js_dev);
if (fd < 0)
mp_tmsg(MSGT_INPUT, MSGL_ERR, "Can't init input joystick\n");
else
mp_input_add_key_fd(ictx, fd, 1, mp_input_joystick_read,
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
close, NULL);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
}
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
#endif
Rename a bunch of miscellaneous preprocessor directives. Switch them from a HAVE_ to a CONFIG_ prefix. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@27423 b3059339-0415-0410-9bf9-f77b7e298cf2
2008-08-07 10:58:07 +02:00
#ifdef CONFIG_LIRC
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
if (input_conf->use_lirc) {
int fd = mp_input_lirc_init();
if (fd > 0)
mp_input_add_cmd_fd(ictx, fd, 0, mp_input_lirc_read,
mp_input_lirc_close);
}
Added lirc support in input git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4433 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-31 10:39:11 +01:00
#endif
Rename a bunch of miscellaneous preprocessor directives. Switch them from a HAVE_ to a CONFIG_ prefix. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@27423 b3059339-0415-0410-9bf9-f77b7e298cf2
2008-08-07 10:58:07 +02:00
#ifdef CONFIG_LIRCC
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
if (input_conf->use_lircc) {
Rename to "mpv" This changes the name of this project to mpv. Most user-visible mentions of "MPlayer" and "mplayer" are changed to "mpv". The binary name and the default config file location are changed as well. The new default config file location is: ~/.mpv/ Remove etc/mplayer.desktop. Apparently this was for the MPlayer GUI, which has been removed from mplayer2 ages ago. We don't have a logo, and the MS Windows resource files sort-of require one, so leave etc/mplayer.ico/.xpm as-is. Remove the debian and rpm packaging scripts. These contained outdated dependencies and likely were more harmful than useful. (Patches which add working and well-tested packaging are welcome.)
2012-10-11 02:04:08 +02:00
int fd = lircc_init("mpv", NULL);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
if (fd >= 0)
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
mp_input_add_cmd_fd(ictx, fd, 1, NULL, lircc_cleanup);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
}
lirccd support by Fredrik Tolf <fredrik@dolda2000.cjb.net> git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@10216 b3059339-0415-0410-9bf9-f77b7e298cf2
2003-05-30 20:23:55 +02:00
#endif
osx: add Apple Remote support After killing the non functional AR support in c8fd9e5 I got much complaints so this adds AR support back in (and it works). I am using the HIDRemote class by Felix Schwarz and that part of the code is under the BSD license. I slightly modified it replacing [NSApplication sharedApplication] with NSApp. The code of the class is quite complex (probably because it had to deal with all the edge cases with IOKit) but it works nicely as a black box. In a later commit I'll remove the deprecation warnings caused by HIDRemote's usage of Gestalt. Check out `etc/input.conf` for the default bindings. Apple Remote functionality is automatically compiled in when cocoa is enabled. It can be disabled at runtime with the `--no-ar` option.
2013-06-03 00:52:40 +02:00
#ifdef CONFIG_COCOA
if (input_conf->use_ar) {
osx: improve Media Keys support This commit addresses some issues with the users had with the previous implementation in commit c39efb9. Here's the changes: * Use Quartz Event Taps to remove Media Key events mpv handles from the global OS X queue. This prevents conflicts with iTunes. I did this on the main thread since it is mostly idling. It's the playloop thread that actually does all the work so there is no danger of blocking the event tap callback. * Introduce `--no-media-keys` switch so that users can disable all of mpv's media key handling at runtime (some prefer iTunes for example). * Use mpv's bindings so that users can customize what the media keys do via input.conf. Current bindings are: MK_PLAY cycle pause MK_PREV playlist_prev MK_NEXT playlist_next An additional benefit of this implementation is that it is completly handled by the `macosx_events` file instead of `macosx_application` making the project organization more straightforward.
2013-06-04 22:12:23 +02:00
cocoa_init_apple_remote();
m_config: refactor some things Change how m_config is initialized. Make it more uniform; now all m_config structs are intialized in exactly the same way. Make sure there's only a single m_option[] array defining the options, and keep around the pointer to the optstruct default value, and the optstruct size as well. This will allow reconstructing the option default values in the following commit. In particular, stop pretending that the handling of some special options (like --profile, --v, and some others) is in any way elegant, and make them explicit hacks. This is really more readable and easier to understand than what was before, and simplifies the code.
2013-07-27 21:26:00 +02:00
ictx->using_ar = true;
osx: improve Media Keys support This commit addresses some issues with the users had with the previous implementation in commit c39efb9. Here's the changes: * Use Quartz Event Taps to remove Media Key events mpv handles from the global OS X queue. This prevents conflicts with iTunes. I did this on the main thread since it is mostly idling. It's the playloop thread that actually does all the work so there is no danger of blocking the event tap callback. * Introduce `--no-media-keys` switch so that users can disable all of mpv's media key handling at runtime (some prefer iTunes for example). * Use mpv's bindings so that users can customize what the media keys do via input.conf. Current bindings are: MK_PLAY cycle pause MK_PREV playlist_prev MK_NEXT playlist_next An additional benefit of this implementation is that it is completly handled by the `macosx_events` file instead of `macosx_application` making the project organization more straightforward.
2013-06-04 22:12:23 +02:00
}
if (input_conf->use_media_keys) {
cocoa_init_media_keys();
m_config: refactor some things Change how m_config is initialized. Make it more uniform; now all m_config structs are intialized in exactly the same way. Make sure there's only a single m_option[] array defining the options, and keep around the pointer to the optstruct default value, and the optstruct size as well. This will allow reconstructing the option default values in the following commit. In particular, stop pretending that the handling of some special options (like --profile, --v, and some others) is in any way elegant, and make them explicit hacks. This is really more readable and easier to understand than what was before, and simplifies the code.
2013-07-27 21:26:00 +02:00
ictx->using_cocoa_media_keys = true;
osx: add Apple Remote support After killing the non functional AR support in c8fd9e5 I got much complaints so this adds AR support back in (and it works). I am using the HIDRemote class by Felix Schwarz and that part of the code is under the BSD license. I slightly modified it replacing [NSApplication sharedApplication] with NSApp. The code of the class is quite complex (probably because it had to deal with all the edge cases with IOKit) but it works nicely as a black box. In a later commit I'll remove the deprecation warnings caused by HIDRemote's usage of Gestalt. Check out `etc/input.conf` for the default bindings. Apple Remote functionality is automatically compiled in when cocoa is enabled. It can be disabled at runtime with the `--no-ar` option.
2013-06-03 00:52:40 +02:00
}
#endif
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
if (input_conf->in_file) {
windows support: unicode filenames Windows uses a legacy codepage for char* / runtime functions accepting char *. Using UTF-8 as the codepage with setlocale() is explicitly forbidden. Work this around by overriding the MSVCRT functions with wrapper macros, that assume UTF-8 and use "proper" API calls like _wopen etc. to deal with unicode filenames. All code that uses standard functions that take or return filenames must now include osdep/io.h. stat() can't be overridden, because MinGW-w64 itself defines "stat" as a macro. Change code to use use mp_stat() instead. This is not perfectly clean, but still somewhat sane, and much better than littering the rest of the mplayer code with MinGW specific hacks. It's also a bit fragile, but that's actually little different from the previous situation. Also, MinGW is unlikely to ever include a nice way of dealing with this.
2012-02-03 08:05:11 +01:00
int mode = O_RDONLY;
#ifndef __MINGW32__
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
// Use RDWR for FIFOs to ensure they stay open over multiple accesses.
windows support: unicode filenames Windows uses a legacy codepage for char* / runtime functions accepting char *. Using UTF-8 as the codepage with setlocale() is explicitly forbidden. Work this around by overriding the MSVCRT functions with wrapper macros, that assume UTF-8 and use "proper" API calls like _wopen etc. to deal with unicode filenames. All code that uses standard functions that take or return filenames must now include osdep/io.h. stat() can't be overridden, because MinGW-w64 itself defines "stat" as a macro. Change code to use use mp_stat() instead. This is not perfectly clean, but still somewhat sane, and much better than littering the rest of the mplayer code with MinGW specific hacks. It's also a bit fragile, but that's actually little different from the previous situation. Also, MinGW is unlikely to ever include a nice way of dealing with this.
2012-02-03 08:05:11 +01:00
// Note that on Windows due to how the API works, using RDONLY should
// be ok.
struct stat st;
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
if (stat(input_conf->in_file, &st) == 0 && S_ISFIFO(st.st_mode))
windows support: unicode filenames Windows uses a legacy codepage for char* / runtime functions accepting char *. Using UTF-8 as the codepage with setlocale() is explicitly forbidden. Work this around by overriding the MSVCRT functions with wrapper macros, that assume UTF-8 and use "proper" API calls like _wopen etc. to deal with unicode filenames. All code that uses standard functions that take or return filenames must now include osdep/io.h. stat() can't be overridden, because MinGW-w64 itself defines "stat" as a macro. Change code to use use mp_stat() instead. This is not perfectly clean, but still somewhat sane, and much better than littering the rest of the mplayer code with MinGW specific hacks. It's also a bit fragile, but that's actually little different from the previous situation. Also, MinGW is unlikely to ever include a nice way of dealing with this.
2012-02-03 08:05:11 +01:00
mode = O_RDWR;
mode |= O_NONBLOCK;
#endif
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
int in_file_fd = open(input_conf->in_file, mode);
if (in_file_fd >= 0)
cleanup: remove some unnecessary input.[ch] typedefs Remove some unnecessary typedefs and remove pointless mp_ prefix from some internal struct names. Change the type of the "close_func" pointers from "void f(int fd)" to "int f(int fd)" so that using standard close() there is valid. Delete some useless assert() statements. Move internal MP_MAX_KEY_DOWN define from input.h to input.c.
2011-07-16 17:17:48 +02:00
mp_input_add_cmd_fd(ictx, in_file_fd, 1, NULL, close);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
else
mp_tmsg(MSGT_INPUT, MSGL_ERR, "Can't open %s: %s\n",
input_conf->in_file, strerror(errno));
}
OSX, input: implement wakeup in response to Cocoa events Add code to wake up the select() call in input.c when an OSX event is available and a Cocoa OpenGL backend is initialized. Fixes the slow response to input or other events in Cocoa-based VOs during long select() sleeps (e.g., when mplayer2 is paused) introduced by commit 7040968.
2012-04-03 08:13:12 +02:00
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
return ictx;
Added lirc support in input git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4433 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-31 10:39:11 +01:00
}
input: free all queued/allocated commands on exit These were memory leaks in theory, though not in practice (all memory is free'd on exit anyway). However, it was still annoying when leak reporting is enabled. I'm not sure if there was an actual leak in check_autorepeat(), maybe not.
2013-02-03 17:43:47 +01:00
static void clear_queue(struct cmd_queue *queue)
{
while (queue->first) {
struct mp_cmd *item = queue->first;
queue_remove(queue, item);
talloc_free(item);
}
}
m_config: refactor some things Change how m_config is initialized. Make it more uniform; now all m_config structs are intialized in exactly the same way. Make sure there's only a single m_option[] array defining the options, and keep around the pointer to the optstruct default value, and the optstruct size as well. This will allow reconstructing the option default values in the following commit. In particular, stop pretending that the handling of some special options (like --profile, --v, and some others) is in any way elegant, and make them explicit hacks. This is really more readable and easier to understand than what was before, and simplifies the code.
2013-07-27 21:26:00 +02:00
void mp_input_uninit(struct input_ctx *ictx)
Create a context for input.c state Start moving static variables to a context struct. Only autorepeat state is moved to the struct in this commit. mp_input_check_interrupt now requires the context variable. Change stream functions to pass it. It's still stored in a static variable in stream/.
2008-04-30 06:15:52 +02:00
{
Remove INITIALIZED_INPUT from mpctx->initialized_flags The input functions are never uninited except at exit. Move the uninit to player exit function and remove the flag.
2008-04-30 08:14:06 +02:00
if (!ictx)
return;
osx: add Apple Remote support After killing the non functional AR support in c8fd9e5 I got much complaints so this adds AR support back in (and it works). I am using the HIDRemote class by Felix Schwarz and that part of the code is under the BSD license. I slightly modified it replacing [NSApplication sharedApplication] with NSApp. The code of the class is quite complex (probably because it had to deal with all the edge cases with IOKit) but it works nicely as a black box. In a later commit I'll remove the deprecation warnings caused by HIDRemote's usage of Gestalt. Check out `etc/input.conf` for the default bindings. Apple Remote functionality is automatically compiled in when cocoa is enabled. It can be disabled at runtime with the `--no-ar` option.
2013-06-03 00:52:40 +02:00
#ifdef CONFIG_COCOA
m_config: refactor some things Change how m_config is initialized. Make it more uniform; now all m_config structs are intialized in exactly the same way. Make sure there's only a single m_option[] array defining the options, and keep around the pointer to the optstruct default value, and the optstruct size as well. This will allow reconstructing the option default values in the following commit. In particular, stop pretending that the handling of some special options (like --profile, --v, and some others) is in any way elegant, and make them explicit hacks. This is really more readable and easier to understand than what was before, and simplifies the code.
2013-07-27 21:26:00 +02:00
if (ictx->using_ar) {
osx: improve Media Keys support This commit addresses some issues with the users had with the previous implementation in commit c39efb9. Here's the changes: * Use Quartz Event Taps to remove Media Key events mpv handles from the global OS X queue. This prevents conflicts with iTunes. I did this on the main thread since it is mostly idling. It's the playloop thread that actually does all the work so there is no danger of blocking the event tap callback. * Introduce `--no-media-keys` switch so that users can disable all of mpv's media key handling at runtime (some prefer iTunes for example). * Use mpv's bindings so that users can customize what the media keys do via input.conf. Current bindings are: MK_PLAY cycle pause MK_PREV playlist_prev MK_NEXT playlist_next An additional benefit of this implementation is that it is completly handled by the `macosx_events` file instead of `macosx_application` making the project organization more straightforward.
2013-06-04 22:12:23 +02:00
cocoa_uninit_apple_remote();
}
m_config: refactor some things Change how m_config is initialized. Make it more uniform; now all m_config structs are intialized in exactly the same way. Make sure there's only a single m_option[] array defining the options, and keep around the pointer to the optstruct default value, and the optstruct size as well. This will allow reconstructing the option default values in the following commit. In particular, stop pretending that the handling of some special options (like --profile, --v, and some others) is in any way elegant, and make them explicit hacks. This is really more readable and easier to understand than what was before, and simplifies the code.
2013-07-27 21:26:00 +02:00
if (ictx->using_cocoa_media_keys) {
osx: improve Media Keys support This commit addresses some issues with the users had with the previous implementation in commit c39efb9. Here's the changes: * Use Quartz Event Taps to remove Media Key events mpv handles from the global OS X queue. This prevents conflicts with iTunes. I did this on the main thread since it is mostly idling. It's the playloop thread that actually does all the work so there is no danger of blocking the event tap callback. * Introduce `--no-media-keys` switch so that users can disable all of mpv's media key handling at runtime (some prefer iTunes for example). * Use mpv's bindings so that users can customize what the media keys do via input.conf. Current bindings are: MK_PLAY cycle pause MK_PREV playlist_prev MK_NEXT playlist_next An additional benefit of this implementation is that it is completly handled by the `macosx_events` file instead of `macosx_application` making the project organization more straightforward.
2013-06-04 22:12:23 +02:00
cocoa_uninit_media_keys();
osx: add Apple Remote support After killing the non functional AR support in c8fd9e5 I got much complaints so this adds AR support back in (and it works). I am using the HIDRemote class by Felix Schwarz and that part of the code is under the BSD license. I slightly modified it replacing [NSApplication sharedApplication] with NSApp. The code of the class is quite complex (probably because it had to deal with all the edge cases with IOKit) but it works nicely as a black box. In a later commit I'll remove the deprecation warnings caused by HIDRemote's usage of Gestalt. Check out `etc/input.conf` for the default bindings. Apple Remote functionality is automatically compiled in when cocoa is enabled. It can be disabled at runtime with the `--no-ar` option.
2013-06-03 00:52:40 +02:00
}
#endif
input: use only one array for input sources Removes some code duplication. Also restructure the input waiting code a bit: split the select() loop into a input_wait_read() function. On systems which do not have POSIX select(), this function has an alternate implementation, which waits unconditionally.
2013-07-07 16:53:35 +02:00
for (int i = 0; i < ictx->num_fds; i++) {
if (ictx->fds[i].close_func)
ictx->fds[i].close_func(ictx->fds[i].fd);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
}
input: free all queued/allocated commands on exit These were memory leaks in theory, though not in practice (all memory is free'd on exit anyway). However, it was still annoying when leak reporting is enabled. I'm not sure if there was an actual leak in check_autorepeat(), maybe not.
2013-02-03 17:43:47 +01:00
for (int i = 0; i < 2; i++) {
ao_pulse, core: make pulse thread wake up core for more data For ao_pulse, the current latency is not a good indicator of how soon the AO requires new data to avoid underflow. Add an internal pipe that can be used to wake up the input loop from select(), and make the pulseaudio main loop (which runs in a separate thread) use this mechanism when pulse requests more data. The wakeup signal currently contains no information about the reason for the wakup, but audio buffers are always filled when the event loop wakes up. Also, request a latency of 1 second from the Pulseaudio server. The default is normally significantly higher. We don't need low latency, while higher latency helps prevent underflows reduces need for wakeups.
2012-03-25 21:58:48 +02:00
if (ictx->wakeup_pipe[i] != -1)
close(ictx->wakeup_pipe[i]);
input: free all queued/allocated commands on exit These were memory leaks in theory, though not in practice (all memory is free'd on exit anyway). However, it was still annoying when leak reporting is enabled. I'm not sure if there was an actual leak in check_autorepeat(), maybe not.
2013-02-03 17:43:47 +01:00
}
clear_queue(&ictx->key_cmd_queue);
clear_queue(&ictx->control_cmd_queue);
input: handle mouse movement differently Before this commit, mouse movement events emitted a special command ("set_mouse_pos"), which was specially handled in command.c. This was once special-cased to the dvdnav and menu code, and did nothing after libmenu and dvdnav were removed. Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"), which then can be bound to an arbitrary command. The mouse position is now managed in input.c. A command which actually needs the mouse position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos() to query it. The former returns raw window-space coordinates, while the latter returns coordinates transformed to OSD- space. (Both are the same for most VOs, except vo_xv and vo_x11, which can't render OSD in window-space. These require extra code for mapping mouse position.) As of this commit, there is still nothing that uses mouse movement, so MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the mouse (much like MOUSE_BTN0). Extend the concept of input sections. Allow multiple sections to be active at once, and organize them as stack. Bindings from the top of the stack are preferred to lower ones. Each section has a mouse input section associated, inside which mouse events are associated with the bindings. If the mouse pointer is outside of a section's mouse area, mouse events will be dispatched to an input section lower on the stack of active sections. This is intended for scripting, which is to be added later. Two scripts could occupy different areas of the screen without conflicting with each other. (If it turns out that this mechanism is useless, we'll just remove it again.)
2013-04-26 02:13:30 +02:00
talloc_free(ictx->current_down_cmd);
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
pthread_mutex_destroy(&ictx->mutex);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
talloc_free(ictx);
A new configurable input system and joystick support for this system git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@4419 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-01-30 13:46:03 +01:00
}
options: change --input=keylist, cmdlist implementation Change the --input=keylist and --input=cmdlist suboptions to use the "print function" option type. This changes their semantics somewhat, and now some other output can appear after the printed lists (before, they called "exit(0)" directly). I'm not aware of any program parsing the output which could be affected.
2012-05-07 22:16:20 +02:00
static int print_key_list(m_option_t *cfg, char *optname, char *optparam)
input.c: Remove mp_input_ prefixes from static function names The prefixes make the code unnecessarily verbose for no real benefit.
2008-04-30 08:30:51 +02:00
{
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
int i;
printf("\n");
for (i = 0; key_names[i].name != NULL; i++)
printf("%s\n", key_names[i].name);
options: change --input=keylist, cmdlist implementation Change the --input=keylist and --input=cmdlist suboptions to use the "print function" option type. This changes their semantics somewhat, and now some other output can appear after the printed lists (before, they called "exit(0)" directly). I'm not aware of any program parsing the output which could be affected.
2012-05-07 22:16:20 +02:00
return M_OPT_EXIT;
Added options -input keylist and -input cmdlist to list all know keys and commands git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@5219 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-03-20 14:30:57 +01:00
}
options: change --input=keylist, cmdlist implementation Change the --input=keylist and --input=cmdlist suboptions to use the "print function" option type. This changes their semantics somewhat, and now some other output can appear after the printed lists (before, they called "exit(0)" directly). I'm not aware of any program parsing the output which could be affected.
2012-05-07 22:16:20 +02:00
static int print_cmd_list(m_option_t *cfg, char *optname, char *optparam)
input.c: Remove mp_input_ prefixes from static function names The prefixes make the code unnecessarily verbose for no real benefit.
2008-04-30 08:30:51 +02:00
{
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
const mp_cmd_t *cmd;
int i, j;
for (i = 0; (cmd = &mp_cmds[i])->name != NULL; i++) {
printf("%-20.20s", cmd->name);
commands: add choice type to input commands Allow using the choice type (as it used for command line) for arguments of input commands. Change the magic integer arguments of some commands (like seek) to use choices instead. The old numeric values are still allowed (but only those which made sense before, not arbitrary integers). In order to do this, remove the input.c specific types (like MP_CMD_ARG_INT) completely and specify commands using the m_option types. Also, add the special choice "-" to some arguments. It's supposed to signify the default value, so arguments can be easily skipped. Maybe the choice option should recognize this and not change the previous value, but we'll leave this for later. For now, leave compatibility integer values for all new choice arguments, e.g. "0" maps to 0. We could let the choice option type do this automatically, but we don't, because we want user input values and internal mplayer values decoupled in general. The compatibility options will be removed one day, too. Also, remove optional args for strings - would require either annoying additional code, or copying strings twice. It's not used, so remove it.
2012-09-24 21:57:43 +02:00
for (j = 0; j < MP_CMD_MAX_ARGS && cmd->args[j].type.type; j++) {
const char *type = cmd->args[j].type.type->name;
input.c: simplify command definitions Change command definitions to allow writing the command definition table in more readable format.
2012-03-21 01:23:35 +01:00
if (cmd->args[j].optional)
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
printf(" [%s]", type);
else
printf(" %s", type);
}
printf("\n");
Added options -input keylist and -input cmdlist to list all know keys and commands git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@5219 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-03-20 14:30:57 +01:00
}
options: change --input=keylist, cmdlist implementation Change the --input=keylist and --input=cmdlist suboptions to use the "print function" option type. This changes their semantics somewhat, and now some other output can appear after the printed lists (before, they called "exit(0)" directly). I'm not aware of any program parsing the output which could be affected.
2012-05-07 22:16:20 +02:00
return M_OPT_EXIT;
Added options -input keylist and -input cmdlist to list all know keys and commands git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@5219 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-03-20 14:30:57 +01:00
}
mpdemux.c|h moved to libinput, mpdemux_check_interrupt() -> mp_input_check_interrupt() git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@7863 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-10-23 16:46:20 +02:00
ao_pulse, core: make pulse thread wake up core for more data For ao_pulse, the current latency is not a good indicator of how soon the AO requires new data to avoid underflow. Add an internal pipe that can be used to wake up the input loop from select(), and make the pulseaudio main loop (which runs in a separate thread) use this mechanism when pulse requests more data. The wakeup signal currently contains no information about the reason for the wakup, but audio buffers are always filled when the event loop wakes up. Also, request a latency of 1 second from the Pulseaudio server. The default is normally significantly higher. We don't need low latency, while higher latency helps prevent underflows reduces need for wakeups.
2012-03-25 21:58:48 +02:00
void mp_input_wakeup(struct input_ctx *ictx)
{
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
// Safe without locking
ao_pulse, core: make pulse thread wake up core for more data For ao_pulse, the current latency is not a good indicator of how soon the AO requires new data to avoid underflow. Add an internal pipe that can be used to wake up the input loop from select(), and make the pulseaudio main loop (which runs in a separate thread) use this mechanism when pulse requests more data. The wakeup signal currently contains no information about the reason for the wakup, but audio buffers are always filled when the event loop wakes up. Also, request a latency of 1 second from the Pulseaudio server. The default is normally significantly higher. We don't need low latency, while higher latency helps prevent underflows reduces need for wakeups.
2012-03-25 21:58:48 +02:00
if (ictx->wakeup_pipe[1] >= 0)
write(ictx->wakeup_pipe[1], &(char){0}, 1);
}
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
static bool test_abort(struct input_ctx *ictx)
{
if (async_quit_request || queue_has_abort_cmds(&ictx->key_cmd_queue) ||
queue_has_abort_cmds(&ictx->control_cmd_queue))
{
mp_tmsg(MSGT_INPUT, MSGL_WARN, "Received command to move to "
"another file. Aborting current processing.\n");
return true;
}
return false;
}
Document that time argument is in milliseconds. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@31252 b3059339-0415-0410-9bf9-f77b7e298cf2
2010-05-28 20:44:41 +02:00
/**
* \param time time to wait for an interruption in milliseconds
*/
Create a context for input.c state Start moving static variables to a context struct. Only autorepeat state is moved to the struct in this commit. mp_input_check_interrupt now requires the context variable. Change stream functions to pass it. It's still stored in a static variable in stream/.
2008-04-30 06:15:52 +02:00
int mp_input_check_interrupt(struct input_ctx *ictx, int time)
{
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
bool res = test_abort(ictx);
if (!res) {
input: rework event reading and command queuing Rework much of the logic related to reading from event sources and queuing commands. The two biggest architecture changes are: - The code buffering keycodes in mp_fifo.c is gone. Instead key input is now immediately fed to input.c and interpreted as commands, and then the commands are buffered instead. - mp_input_get_cmd() now always tries to read every available event from every event source and convert them to (buffered) commands. Before it would only process new events until one new command became available. Some relevant behavior changes: - Before commands could be lost when stream code called mp_input_check_interrupt() which read commands (to see if they were of types that triggered aborts during slow IO tasks) and then threw them away. This was especially an issue if cache was enabled and slow to read. Fixed - now it's possible to check whether there are queued commands which will abort playback of the current file without throwing other commands away. - mp_input_check_interrupt() now prints a message if it returns true. This is especially useful because the failures caused by aborted stream reads can trigger error messages from other code that was doing the read; the new message makes it more obvious what the cause of the subsequent error messages is. - It's now possible to again avoid making stdin non-blocking (which caused some issues) without reintroducing extra latency. The change will be done in a subsequent commit. - Event sources that do not support select() should now have somewhat lower latency in certain situations as they will be checked both before and after select()/sleep in input reading; before the sleep always happened first even if such sources already had queued input. Before the key fifo was also handled in this manner (first key triggered select, but if multiple were read then rest could be delayed; however in most cases this didn't add latency in practice as after central code started doing command handling it queried for further commands with a max sleep time of 0). - Key fifo limiting is more accurate now: it now counts actual commands intead of keycodes, and all queued keys are read immediately from input devices so they can be counted correctly. - Since keypresses are now interpreted immediately, commands which change keybindings will no longer affect following keypresses that have already been read before the command is executed. This should not be an issue in practice with current keybinding behavior.
2011-07-17 03:47:50 +02:00
read_all_events(ictx, time);
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
res = test_abort(ictx);
cleanup: reformat input.c
2011-04-25 10:38:46 +02:00
}
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_unlock(ictx);
return res;
mpdemux.c|h moved to libinput, mpdemux_check_interrupt() -> mp_input_check_interrupt() git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@7863 b3059339-0415-0410-9bf9-f77b7e298cf2
2002-10-23 16:46:20 +02:00
}
core: do mouse cursor hiding business in frontend Do this so that not every VO backend has to setup a timer for cursor hiding and interpret the --cursor-autohide option.
2013-05-16 23:17:46 +02:00
Replace all calls to GetTimer()/GetTimerMS() GetTimer() is generally replaced with mp_time_us(). Both calls return microseconds, but the latter uses int64_t, us defined to never wrap, and never returns 0 or negative values. GetTimerMS() has no direct replacement. Instead the other functions are used. For some code, switch to mp_time_sec(), which returns the time as double float value in seconds. The returned time is offset to program start time, so there is enough precision left to deliver microsecond resolution for at least 100 years. Unless it's casted to a float (or the CPU reduces precision), which is why we still use mp_time_us() out of paranoia in places where precision is clearly needed. Always switch to the correct time. The whole point of the new timer calls is that they don't wrap, and storing microseconds in unsigned int variables would negate this. In some cases, remove wrap-around handling for time values.
2013-05-25 18:31:06 +02:00
unsigned int mp_input_get_mouse_event_counter(struct input_ctx *ictx)
core: do mouse cursor hiding business in frontend Do this so that not every VO backend has to setup a timer for cursor hiding and interpret the --cursor-autohide option.
2013-05-16 23:17:46 +02:00
{
mplayer: don't hide mouse cursor if mouse is inside mouse area
2013-06-19 18:45:02 +02:00
// Make the frontend always display the mouse cursor (as long as it's not
// forced invisible) if mouse input is desired.
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
input_lock(ictx);
mplayer: don't hide mouse cursor if mouse is inside mouse area
2013-06-19 18:45:02 +02:00
if (mp_input_test_mouse_active(ictx, ictx->mouse_x, ictx->mouse_y))
ictx->mouse_event_counter++;
input: lock for accessing struct input_ctx The previous code was locking only the input queue. That was too weak since it didn't protect the input_ctx data structure. So remove the locking on the queue and lock all the public functions that interact with the input_ctx. The private functions and public functions that do not act on the input_ctx (there are quite some functions doing mp_cmd manipulations for instance) are not locked. Some changes by wm4. Use a recursive mutex, and restructure some code to be less annoying with locks, such as converting code to single return, or making use of the recursive mutex.
2013-08-26 19:15:35 +02:00
int ret = ictx->mouse_event_counter;
input_unlock(ictx);
return ret;
core: do mouse cursor hiding business in frontend Do this so that not every VO backend has to setup a timer for cursor hiding and interpret the --cursor-autohide option.
2013-05-16 23:17:46 +02:00
}