[ie/HockeyCanada] Add extractor #10002
Conversation
|
|
||
|
|
||
| class HockeyCanadaIE(InfoExtractor): | ||
| _VALID_URL = r'https://video.hockeycanada.ca/(en(?:-\w+)?|fr)/c/.+\.(?P<id>\d+)' |
There was a problem hiding this comment.
| _VALID_URL = r'https://video.hockeycanada.ca/(en(?:-\w+)?|fr)/c/.+\.(?P<id>\d+)' | |
| _VALID_URL = r'https?://video\.hockeycanada\.ca/[a-z]{2}(?:-[a-z]{2})?/c/[\w-]+\.(?P<id>\d+)' |
| 'upload_date': '20211015', | ||
| 'tags': ['English', '2021', "National Women's Team"], | ||
| 'description': 'md5:efb1cf6165b48cc3f5555c4262dd5b23', | ||
| 'thumbnail': str, |
There was a problem hiding this comment.
use regex for thumbnail values, even it's something very lenient like r're:^https?://.+\.jpg', so that we test we're returning actual image urls
| 'timestamp': 1634310409, | ||
| 'upload_date': '20211015', | ||
| 'tags': ['English', '2021', "National Women's Team"], | ||
| 'description': 'md5:efb1cf6165b48cc3f5555c4262dd5b23', |
There was a problem hiding this comment.
also if you want, you could use regex for the description values too instead of md5 hashes. up to you
| webpage = self._download_webpage(url, video_id) | ||
|
|
||
| data_url = self._html_search_regex( | ||
| r'content_api:\s*(["\'])(?P<url>.+?)\1', webpage, 'content api url', group='url') |
There was a problem hiding this comment.
Maybe make this regex a bit more strict/defined so we don't accidentally capture a huge block of JS
| media_config = traverse_obj( | ||
| self._download_json(data_url, video_id), | ||
| ('config', {lambda x: json.loads(base64.b64decode(x).decode())})) |
There was a problem hiding this comment.
I would move this API call into _real_extract
media_config = traverse_obj(
self._download_json(data_url, video_id),
('config', {base64.b64decode}, {bytes.decode}, {json.loads}, {dict}))and pass media_config to _yield_formats instead of data_url
| for media_source in traverse_obj(media_config, ('media', 'source', ..., { | ||
| 'url': ('src', {url_or_none}), | ||
| 'type': ('type', {mimetype2ext}), | ||
| })): | ||
| if not (media_url := media_source.get('url')): | ||
| continue | ||
| media_type = media_source.get('type') |
There was a problem hiding this comment.
Doing dict key traversal in the for loop is not useful if we're assigning to variables afterwards anyways IMO. We could just filter like this:
| for media_source in traverse_obj(media_config, ('media', 'source', ..., { | |
| 'url': ('src', {url_or_none}), | |
| 'type': ('type', {mimetype2ext}), | |
| })): | |
| if not (media_url := media_source.get('url')): | |
| continue | |
| media_type = media_source.get('type') | |
| for media_source in traverse_obj(media_config, ('media', 'source', lambda_, v: url_or_none(v['src']))): | |
| media_url = media_source['src'] | |
| media_type = mimetype2ext(media_source.get('type')) |
| media_type = media_source.get('type') | ||
|
|
||
| if media_type == 'm3u8': | ||
| yield from self._extract_m3u8_formats(media_url, video_id) |
There was a problem hiding this comment.
Should be non-fatal if there possibly other formats available. And specifying format ids would be nice to differentiate hls from http etc
| yield from self._extract_m3u8_formats(media_url, video_id) | |
| yield from self._extract_m3u8_formats(media_url, video_id, fatal=False, m3u8_id='hls') |
| yield from self._extract_m3u8_formats(media_url, video_id) | ||
| elif media_type == 'mp4': | ||
| fmt = { | ||
| 'url': media_url, |
There was a problem hiding this comment.
| 'url': media_url, | |
| 'format_id': 'http', | |
| 'url': media_url, |
IMPORTANT: PRs without the template will be CLOSED
Description of your pull request and other information
Fixes #1377
Template
Before submitting a pull request make sure you have:
In order to be accepted and merged into yt-dlp each piece of code must be in public domain or released under Unlicense. Check all of the following options that apply:
What is the purpose of your pull request?