[youtube] Move search URL extraction to appropriate extractor

This commit is contained in:
xarantolus 2020-06-19 21:29:47 +02:00
parent b948643f9c
commit 19f671f88b
1 changed files with 33 additions and 34 deletions

View File

@ -3153,40 +3153,7 @@ class YoutubePlaylistsIE(YoutubePlaylistsBaseInfoExtractor):
class YoutubeSearchBaseInfoExtractor(YoutubePlaylistBaseInfoExtractor): class YoutubeSearchBaseInfoExtractor(YoutubePlaylistBaseInfoExtractor):
_PLAYLIST_DATA = r'window\[\"ytInitialData\"\]\W?=\W?({.*?});' _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})(?:[^"]*"[^>]+\btitle="(?P<title>[^"]+))?'
def extract_videos_from_page_impl(self, page, ids_in_page, titles_in_page):
playlist_json = self._search_regex(self._PLAYLIST_DATA, page, 'ytInitialData')
playlist_response = self._parse_json(playlist_json, None)
result_items = try_get(
playlist_response,
lambda x: x['contents']['twoColumnSearchResultsRenderer']['primaryContents']['sectionListRenderer']['contents'][0]['itemSectionRenderer']['contents'])
# plobj either contains a 'playlistRenderer', 'videoRenderer', 'channelRenderer', 'shelfRenderer' or 'searchPyvRenderer' (promoted video/ad)
for plobj in result_items:
video_id = try_get(plobj, lambda x: x['videoRenderer']['videoId'])
video_title = try_get(plobj, lambda x: x['videoRenderer']['title']['runs'][0]['text'])
if video_id is None or video_title is None:
# we do not have a videoRenderer or it is empty
continue
video_title = video_title.strip()
try:
idx = ids_in_page.index(video_id)
if video_title and not titles_in_page[idx]:
titles_in_page[idx] = video_title
except ValueError:
ids_in_page.append(video_id)
titles_in_page.append(video_title)
def extract_videos_from_page(self, page):
ids_in_page = []
titles_in_page = []
self.extract_videos_from_page_impl(page, ids_in_page, titles_in_page)
return zip(ids_in_page, titles_in_page)
class YoutubeSearchIE(SearchInfoExtractor, YoutubeSearchBaseInfoExtractor): class YoutubeSearchIE(SearchInfoExtractor, YoutubeSearchBaseInfoExtractor):
@ -3250,6 +3217,7 @@ class YoutubeSearchURLIE(YoutubeSearchBaseInfoExtractor):
IE_DESC = 'YouTube.com search URLs' IE_DESC = 'YouTube.com search URLs'
IE_NAME = 'youtube:search_url' IE_NAME = 'youtube:search_url'
_VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?(?:search_query|q)=(?P<query>[^&]+)(?:[&]|$)' _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?(?:search_query|q)=(?P<query>[^&]+)(?:[&]|$)'
_SEARCH_DATA = r'window\[\"ytInitialData\"\]\W?=\W?({.*?});'
_TESTS = [{ _TESTS = [{
'url': 'https://www.youtube.com/results?baz=bar&search_query=youtube-dl+test+video&filters=video&lclk=video', 'url': 'https://www.youtube.com/results?baz=bar&search_query=youtube-dl+test+video&filters=video&lclk=video',
'playlist_mincount': 5, 'playlist_mincount': 5,
@ -3261,6 +3229,37 @@ class YoutubeSearchURLIE(YoutubeSearchBaseInfoExtractor):
'only_matching': True, 'only_matching': True,
}] }]
def extract_videos_from_page_impl(self, page, ids_in_page, titles_in_page):
search_response = self._parse_json(self._search_regex(self._SEARCH_DATA, page, 'ytInitialData'), None)
result_items = try_get(
search_response,
lambda x: x['contents']['twoColumnSearchResultsRenderer']['primaryContents']['sectionListRenderer']['contents'][0]['itemSectionRenderer']['contents'])
for plobj in result_items:
video_id = try_get(plobj, lambda x: x['videoRenderer']['videoId'])
video_title = try_get(plobj, lambda x: x['videoRenderer']['title']['runs'][0]['text'])
if video_id is None or video_title is None:
# we do not have a videoRenderer or it is empty
continue
video_title = video_title.strip()
try:
idx = ids_in_page.index(video_id)
if video_title and not titles_in_page[idx]:
titles_in_page[idx] = video_title
except ValueError:
ids_in_page.append(video_id)
titles_in_page.append(video_title)
def extract_videos_from_page(self, page):
ids_in_page = []
titles_in_page = []
self.extract_videos_from_page_impl(page, ids_in_page, titles_in_page)
return zip(ids_in_page, titles_in_page)
def _real_extract(self, url): def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url) mobj = re.match(self._VALID_URL, url)
query = compat_urllib_parse_unquote_plus(mobj.group('query')) query = compat_urllib_parse_unquote_plus(mobj.group('query'))