filter: Added --files-from-raw flag

--files-from parses input files by ignoring comments starting with # and ;
and stripping whitespace from start and end of strings.

The --files-from-raw flag was added that reads every line from the file ignoring
comment characters and not stripping whitespace while maintaining
backwards compatibility.

Fixes #3762

cmd/lsf/lsf.go

@@ -132,13 +132,13 @@ Eg
     "this file contains a comma, in the file name.txt",6
 
 Note that the --absolute parameter is useful for making lists of files
-to pass to an rclone copy with the --files-from flag.
+to pass to an rclone copy with the --files-from-raw flag.
 
 For example to find all the files modified within one day and copy
 those only (without traversing the whole directory structure):
 
     rclone lsf --absolute --files-only --max-age 1d /path/to/local > new_files
-    rclone copy --files-from new_files /path/to/local remote:path
+    rclone copy --files-from-raw new_files /path/to/local remote:path
 
 ` + lshelp.Help,
 	Run: func(command *cobra.Command, args []string) {

docs/content/docs.md

@@ -1542,6 +1542,7 @@ For the filtering options
   * `--include`
   * `--include-from`
   * `--files-from`
+  * `--files-from-raw`
   * `--min-size`
   * `--max-size`
   * `--min-age`

docs/content/filtering.md

@@ -17,8 +17,9 @@ Each path as it passes through rclone is matched against the include
 and exclude rules like `--include`, `--exclude`, `--include-from`,
 `--exclude-from`, `--filter`, or `--filter-from`. The simplest way to
 try them out is using the `ls` command, or `--dry-run` together with
-`-v`. `--filter-from`, `--exclude-from`, `--include-from`, `--files-from`
+`-v`. `--filter-from`, `--exclude-from`, `--include-from`, `--files-from`,
-understand `-` as a file name to mean read from standard input.
+`--files-from-raw` understand `-` as a file name to mean read from standard
+input.
 
 ## Patterns ##
 
@@ -179,6 +180,7 @@ type.
   * `--exclude-from`
   * `--filter`
   * `--filter-from`
+  * `--filter-from-raw`
 
 **Important** You should not use `--include*` together with `--exclude*`. 
 It may produce different results than you expected. In that case try to use: `--filter*`.
@@ -306,9 +308,9 @@ This reads a list of file names from the file passed in and **only**
 these files are transferred.  The **filtering rules are ignored**
 completely if you use this option.
 
-`--files-from` expects a list of files as it's input. [rclone lsf](/commands/rclone_lsf/)
+`--files-from` expects a list of files as it's input. Leading / trailing
-has a compatible format that can be used to export file lists from
+whitespace is stripped from the input lines and lines starting with `#`
-remotes.
+and `;` are ignored.
 
 Rclone will traverse the file system if you use `--files-from`,
 effectively using the files in `--files-from` as a set of filters.
@@ -324,7 +326,8 @@ are read in the order that they are placed on the command line.
 
 Paths within the `--files-from` file will be interpreted as starting
 with the root specified in the command.  Leading `/` characters are
-ignored.
+ignored. See [--files-from-raw](#files-from-raw-read-list-of-source-file-names-without-any-processing)
+if you need the input to be processed in a raw manner.
 
 For example, suppose you had `files-from.txt` with this content:
 
@@ -384,6 +387,13 @@ In this case there will be an extra `home` directory on the remote:
     /home/user1/dir/file  → remote:backup/home/user1/dir/file
     /home/user2/stuff     → remote:backup/home/user2/stuff
 
+### `--files-from-raw` - Read list of source-file names without any processing ###
+This option is same as `--files-from` with the only difference being that the input
+is read in a raw manner. This means that lines with leading/trailing whitespace and
+lines starting with `;` or `#` are read without any processing. [rclone lsf](/commands/rclone_lsf/)
+has a compatible format that can be used to export file lists from remotes, which
+can then be used as an input to `--files-from-raw`.
+
 ### `--min-size` - Don't transfer any file smaller than this ###
 
 This option controls the minimum size file which will be transferred.

fs/filter/filter.go

@@ -88,6 +88,7 @@ type Opt struct {
 	IncludeRule    []string
 	IncludeFrom    []string
 	FilesFrom      []string
+	FilesFromRaw   []string
 	MinAge         fs.Duration
 	MaxAge         fs.Duration
 	MinSize        fs.SizeSuffix
@@ -150,7 +151,7 @@ func NewFilter(opt *Opt) (f *Filter, err error) {
 		addImplicitExclude = true
 	}
 	for _, rule := range f.Opt.IncludeFrom {
-		err := forEachLine(rule, func(line string) error {
+		err := forEachLine(rule, false, func(line string) error {
 			return f.Add(true, line)
 		})
 		if err != nil {
@@ -166,7 +167,7 @@ func NewFilter(opt *Opt) (f *Filter, err error) {
 		foundExcludeRule = true
 	}
 	for _, rule := range f.Opt.ExcludeFrom {
-		err := forEachLine(rule, func(line string) error {
+		err := forEachLine(rule, false, func(line string) error {
 			return f.Add(false, line)
 		})
 		if err != nil {
@@ -186,25 +187,42 @@ func NewFilter(opt *Opt) (f *Filter, err error) {
 		}
 	}
 	for _, rule := range f.Opt.FilterFrom {
-		err := forEachLine(rule, f.AddRule)
+		err := forEachLine(rule, false, f.AddRule)
 		if err != nil {
 			return nil, err
 		}
 	}
 
 	inActive := f.InActive()
+
 	for _, rule := range f.Opt.FilesFrom {
 		if !inActive {
-			return nil, fmt.Errorf("The usage of --files-from overrides all other filters, it should be used alone")
+			return nil, fmt.Errorf("The usage of --files-from overrides all other filters, it should be used alone or with --files-from-raw")
 		}
 		f.initAddFile() // init to show --files-from set even if no files within
-		err := forEachLine(rule, func(line string) error {
+		err := forEachLine(rule, false, func(line string) error {
 			return f.AddFile(line)
 		})
 		if err != nil {
 			return nil, err
 		}
 	}
+
+	for _, rule := range f.Opt.FilesFromRaw {
+		// --files-from-raw can be used with --files-from, hence we do
+		// not need to get the value of f.InActive again
+		if !inActive {
+			return nil, fmt.Errorf("The usage of --files-from-raw overrides all other filters, it should be used alone or with --files-from")
+		}
+		f.initAddFile() // init to show --files-from set even if no files within
+		err := forEachLine(rule, true, func(line string) error {
+			return f.AddFile(line)
+		})
+		if err != nil {
+			return nil, err
+		}
+	}
+
 	if addImplicitExclude {
 		err = f.Add(false, "/**")
 		if err != nil {
@@ -463,8 +481,8 @@ func (f *Filter) IncludeObject(ctx context.Context, o fs.Object) bool {
 
 // forEachLine calls fn on every line in the file pointed to by path
 //
-// It ignores empty lines and lines starting with '#' or ';'
+// It ignores empty lines and lines starting with '#' or ';' if raw is false
-func forEachLine(path string, fn func(string) error) (err error) {
+func forEachLine(path string, raw bool, fn func(string) error) (err error) {
 	var scanner *bufio.Scanner
 	if path == "-" {
 		scanner = bufio.NewScanner(os.Stdin)
@@ -478,10 +496,12 @@ func forEachLine(path string, fn func(string) error) (err error) {
 	}
 	for scanner.Scan() {
 		line := scanner.Text()
+		if !raw {
 			line = strings.TrimSpace(line)
 			if len(line) == 0 || line[0] == '#' || line[0] == ';' {
 				continue
 			}
+		}
 		err := fn(line)
 		if err != nil {
 			return err

fs/filter/filter_test.go

@@ -65,6 +65,29 @@ func TestNewFilterForbiddenMixOfFilesFromAndFilterRule(t *testing.T) {
 	require.Contains(t, err.Error(), "The usage of --files-from overrides all other filters")
 }
 
+func TestNewFilterForbiddenMixOfFilesFromRawAndFilterRule(t *testing.T) {
+	Opt := DefaultOpt
+
+	// Set up the input
+	Opt.FilterRule = []string{"- filter1", "- filter1b"}
+	Opt.FilesFromRaw = []string{testFile(t, "#comment\nfiles1\nfiles2\n")}
+
+	rm := func(p string) {
+		err := os.Remove(p)
+		if err != nil {
+			t.Logf("error removing %q: %v", p, err)
+		}
+	}
+	// Reset the input
+	defer func() {
+		rm(Opt.FilesFromRaw[0])
+	}()
+
+	_, err := NewFilter(&Opt)
+	require.Error(t, err)
+	require.Contains(t, err.Error(), "The usage of --files-from-raw overrides all other filters")
+}
+
 func TestNewFilterWithFilesFromAlone(t *testing.T) {
 	Opt := DefaultOpt
 
@@ -93,6 +116,34 @@ func TestNewFilterWithFilesFromAlone(t *testing.T) {
 	}
 }
 
+func TestNewFilterWithFilesFromRaw(t *testing.T) {
+	Opt := DefaultOpt
+
+	// Set up the input
+	Opt.FilesFromRaw = []string{testFile(t, "#comment\nfiles1\nfiles2\n")}
+
+	rm := func(p string) {
+		err := os.Remove(p)
+		if err != nil {
+			t.Logf("error removing %q: %v", p, err)
+		}
+	}
+	// Reset the input
+	defer func() {
+		rm(Opt.FilesFromRaw[0])
+	}()
+
+	f, err := NewFilter(&Opt)
+	require.NoError(t, err)
+	assert.Len(t, f.files, 3)
+	for _, name := range []string{"#comment", "files1", "files2"} {
+		_, ok := f.files[name]
+		if !ok {
+			t.Errorf("Didn't find file %q in f.files", name)
+		}
+	}
+}
+
 func TestNewFilterFullExceptFilesFromOpt(t *testing.T) {
 	Opt := DefaultOpt
 
@@ -517,7 +568,7 @@ func TestFilterAddDirRuleOrFileRule(t *testing.T) {
 	}
 }
 
-func testFilterForEachLine(t *testing.T, useStdin bool) {
+func testFilterForEachLine(t *testing.T, useStdin, raw bool) {
 	file := testFile(t, `; comment
 one
 # another comment
@@ -546,20 +597,33 @@ five
 		}()
 		fileName = "-"
 	}
-	err := forEachLine(fileName, func(s string) error {
+	err := forEachLine(fileName, raw, func(s string) error {
 		lines = append(lines, s)
 		return nil
 	})
 	require.NoError(t, err)
+	if raw {
+		assert.Equal(t, "; comment,one,# another comment,,,two, # indented comment,three  ,four    ,five,  six  ",
+			strings.Join(lines, ","))
+	} else {
 		assert.Equal(t, "one,two,three,four,five,six", strings.Join(lines, ","))
+	}
 }
 
 func TestFilterForEachLine(t *testing.T) {
-	testFilterForEachLine(t, false)
+	testFilterForEachLine(t, false, false)
 }
 
 func TestFilterForEachLineStdin(t *testing.T) {
-	testFilterForEachLine(t, true)
+	testFilterForEachLine(t, true, false)
+}
+
+func TestFilterForEachLineWithRaw(t *testing.T) {
+	testFilterForEachLine(t, false, true)
+}
+
+func TestFilterForEachLineStdinWithRaw(t *testing.T) {
+	testFilterForEachLine(t, true, true)
 }
 
 func TestFilterMatchesFromDocs(t *testing.T) {

fs/filter/filterflags/filterflags.go

@@ -31,6 +31,7 @@ func AddFlags(flagSet *pflag.FlagSet) {
 	flags.StringArrayVarP(flagSet, &Opt.IncludeRule, "include", "", nil, "Include files matching pattern")
 	flags.StringArrayVarP(flagSet, &Opt.IncludeFrom, "include-from", "", nil, "Read include patterns from file (use - to read from stdin)")
 	flags.StringArrayVarP(flagSet, &Opt.FilesFrom, "files-from", "", nil, "Read list of source-file names from file (use - to read from stdin)")
+	flags.StringArrayVarP(flagSet, &Opt.FilesFromRaw, "files-from-raw", "", nil, "Read list of source-file names from file without any processing of lines (use - to read from stdin)")
 	flags.FVarP(flagSet, &Opt.MinAge, "min-age", "", "Only transfer files older than this in s or suffix ms|s|m|h|d|w|M|y")
 	flags.FVarP(flagSet, &Opt.MaxAge, "max-age", "", "Only transfer files younger than this in s or suffix ms|s|m|h|d|w|M|y")
 	flags.FVarP(flagSet, &Opt.MinSize, "min-size", "", "Only transfer files bigger than this in k or suffix b|k|M|G")