MarcoBuster/yt-dlp
1
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2024-11-24 13:46:53 +01:00
Code Issues Projects Releases Wiki Activity

[nextmedia] Merge AppleDailyRealtimeNewsIE and AppleDailyAnimationNewsIE

Browse Source
This commit is contained in:
Yen Chi Hsuan 2015-05-24 02:36:47 +08:00
parent abca34cbc0
commit 9bf87ae3aa
2 changed files with 16 additions and 24 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
youtube_dl/extractor/__init__.py
View File

@ -338,8 +338,7 @@
from .nextmedia import (
NextMediaIE,
NextMediaActionNewsIE,
AppleDailyRealtimeNewsIE,
AppleDailyAnimationNewsIE
AppleDailyIE,
)
from .nfb import NFBIE
from .nfl import NFLIE

37
youtube_dl/extractor/nextmedia.py
View File

@ -89,8 +89,8 @@ def _real_extract(self, url):
return self._extract_from_nextmedia_page(news_id, url, article_page)
class AppleDailyRealtimeNewsIE(NextMediaIE):
_VALID_URL = r'http://(www|ent).appledaily.com.tw/(realtimenews|enews)/[^/]+/[^/]+/(?P<date>\d+)/(?P<id>\d+)(/.*)?'
class AppleDailyIE(NextMediaIE):
_VALID_URL = r'http://(www|ent).appledaily.com.tw/(animation|realtimenews|enews)/[^/]+/[^/]+/(?P<date>\d+)/(?P<id>\d+)(/.*)?'
_TESTS = [{
'url': 'http://ent.appledaily.com.tw/enews/article/entertainment/20150128/36354694',
'md5': 'a843ab23d150977cc55ef94f1e2c1e4d',
@ -99,7 +99,7 @@ class AppleDailyRealtimeNewsIE(NextMediaIE):
'ext': 'mp4',
'title': '周亭羽走過摩鐵陰霾2男陪吃 九把刀孤寒看醫生',
'thumbnail': 're:^https?://.*\.jpg$',
'description': 'md5:b23787119933404ce515c6356a8c355c',
'description': 'md5:2acd430e59956dc47cd7f67cb3c003f4',
'upload_date': '20150128',
}
}, {
@ -110,26 +110,10 @@ class AppleDailyRealtimeNewsIE(NextMediaIE):
'ext': 'mp4',
'title': '不滿被踩腳　山東兩大媽一路打下車',
'thumbnail': 're:^https?://.*\.jpg$',
'description': 'md5:2648aaf6fc4f401f6de35a91d111aa1d',
'description': 'md5:175b4260c1d7c085993474217e4ab1b4',
'upload_date': '20150128',
}
}]
_URL_PATTERN = r'\{url: \'(.+)\'\}'
def _fetch_title(self, page):
return self._html_search_regex(r'<h1 id="h1">([^<>]+)</h1>', page, 'news title')
def _fetch_thumbnail(self, page):
return self._html_search_regex(r"setInitialImage\(\'([^']+)'\)", page, 'video thumbnail', fatal=False)
def _fetch_timestamp(self, page):
return None
class AppleDailyAnimationNewsIE(AppleDailyRealtimeNewsIE):
_VALID_URL = 'http://www.appledaily.com.tw/animation/[^/]+/[^/]+/(?P<date>\d+)/(?P<id>\d+)(/.*)?'
_TESTS = [{
}, {
'url': 'http://www.appledaily.com.tw/animation/realtimenews/new/20150128/5003671',
'md5': '03df296d95dedc2d5886debbb80cb43f',
'info_dict': {
@ -156,8 +140,17 @@ class AppleDailyAnimationNewsIE(AppleDailyRealtimeNewsIE):
]
}]
_URL_PATTERN = r'\{url: \'(.+)\'\}'
def _fetch_title(self, page):
return self._html_search_meta('description', page, 'news title')
return (self._html_search_regex(r'<h1 id="h1">([^<>]+)</h1>', page, 'news title', default=None) or
self._html_search_meta('description', page, 'news title'))
def _fetch_thumbnail(self, page):
return self._html_search_regex(r"setInitialImage\(\'([^']+)'\)", page, 'video thumbnail', fatal=False)
def _fetch_timestamp(self, page):
return None
def _fetch_description(self, page):
return self._html_search_meta('description', page, 'news description')