[extractor/generic] Accept values for `fragment_query`, `variant_query` (#6600)

Closes #6593
Authored by: bashonly
This commit is contained in:
bashonly 2023-03-23 11:28:23 -05:00 committed by GitHub
parent 6994afc030
commit 5cc0a8fd2e
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 14 additions and 10 deletions

View File

@ -1797,8 +1797,8 @@ The following extractors use this feature:
* `approximate_date`: Extract approximate `upload_date` and `timestamp` in flat-playlist. This may cause date-based filters to be slightly off * `approximate_date`: Extract approximate `upload_date` and `timestamp` in flat-playlist. This may cause date-based filters to be slightly off
#### generic #### generic
* `fragment_query`: Passthrough any query in mpd/m3u8 manifest URLs to their fragments. Does not apply to ffmpeg * `fragment_query`: Passthrough any query in mpd/m3u8 manifest URLs to their fragments if no value is provided, or else apply the query string given as `fragment_query=VALUE`. Does not apply to ffmpeg
* `variant_query`: Passthrough the master m3u8 URL query to its variant playlist URLs * `variant_query`: Passthrough the master m3u8 URL query to its variant playlist URLs if no value is provided, or else apply the query string given as `variant_query=VALUE`
* `hls_key`: An HLS AES-128 key URI *or* key (as hex), and optionally the IV (as hex), in the form of `(URI|KEY)[,IV]`; e.g. `generic:hls_key=ABCDEF1234567980,0xFEDCBA0987654321`. Passing any of these values will force usage of the native HLS downloader and override the corresponding values found in the m3u8 playlist * `hls_key`: An HLS AES-128 key URI *or* key (as hex), and optionally the IV (as hex), in the form of `(URI|KEY)[,IV]`; e.g. `generic:hls_key=ABCDEF1234567980,0xFEDCBA0987654321`. Passing any of these values will force usage of the native HLS downloader and override the corresponding values found in the m3u8 playlist
#### funimation #### funimation

View File

@ -24,7 +24,6 @@ from ..utils import (
mimetype2ext, mimetype2ext,
orderedSet, orderedSet,
parse_duration, parse_duration,
parse_qs,
parse_resolution, parse_resolution,
smuggle_url, smuggle_url,
str_or_none, str_or_none,
@ -2187,18 +2186,23 @@ class GenericIE(InfoExtractor):
self._downloader.write_debug(f'Identified {num} {name}{format_field(note, None, "; %s")}') self._downloader.write_debug(f'Identified {num} {name}{format_field(note, None, "; %s")}')
def _extra_manifest_info(self, info, manifest_url): def _extra_manifest_info(self, info, manifest_url):
if self._configuration_arg('fragment_query'): fragment_query = self._configuration_arg('fragment_query', [None], casesense=True)[0]
query_string = urllib.parse.urlparse(manifest_url).query if fragment_query is not None:
if query_string: fragment_query = self._configuration_arg('fragment_query', casesense=True)[0]
info['extra_param_to_segment_url'] = query_string info['extra_param_to_segment_url'] = (
urllib.parse.urlparse(fragment_query).query or fragment_query
or urllib.parse.urlparse(manifest_url).query or None)
hex_or_none = lambda x: x if re.fullmatch(r'(0x)?[\da-f]+', x, re.IGNORECASE) else None hex_or_none = lambda x: x if re.fullmatch(r'(0x)?[\da-f]+', x, re.IGNORECASE) else None
info['hls_aes'] = traverse_obj(self._configuration_arg('hls_key'), { info['hls_aes'] = traverse_obj(self._configuration_arg('hls_key', casesense=True), {
'uri': (0, {url_or_none}), 'key': (0, {hex_or_none}), 'iv': (1, {hex_or_none}), 'uri': (0, {url_or_none}), 'key': (0, {hex_or_none}), 'iv': (1, {hex_or_none}),
}) or None }) or None
if self._configuration_arg('variant_query'): variant_query = self._configuration_arg('variant_query', [None], casesense=True)[0]
query = parse_qs(manifest_url) if variant_query is not None:
query = urllib.parse.parse_qs(
urllib.parse.urlparse(variant_query).query or variant_query
or urllib.parse.urlparse(manifest_url).query)
for fmt in self._downloader._get_formats(info): for fmt in self._downloader._get_formats(info):
fmt['url'] = update_url_query(fmt['url'], query) fmt['url'] = update_url_query(fmt['url'], query)