jcgruenhage/youtube-dl
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

Merge pull request #7320 from remitamine/adobetv

Browse source 
[adobetv] improve extraction and add support specific language video,show and channel extraction
This commit is contained in:
remitamine 2015-12-04 08:54:06 +01:00
commit 3047121c63
2 changed files with 116 additions and 51 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
youtube_dl/extractor/__init__.py
View file

@ -6,6 +6,8 @@ from .academicearth import AcademicEarthCourseIE
from .addanime import AddAnimeIE
from .adobetv import (
AdobeTVIE,
AdobeTVShowIE,
AdobeTVChannelIE,
AdobeTVVideoIE,
)
from .adultswim import AdultSwimIE

165
youtube_dl/extractor/adobetv.py
View file

@ -1,23 +1,32 @@
from __future__ import unicode_literals
import re
from .common import InfoExtractor
from ..compat import compat_str
from ..utils import (
parse_duration,
unified_strdate,
str_to_int,
int_or_none,
float_or_none,
ISO639Utils,
determine_ext,
)
class AdobeTVIE(InfoExtractor):
_VALID_URL = r'https?://tv\.adobe\.com/watch/[^/]+/(?P<id>[^/]+)'
class AdobeTVBaseIE(InfoExtractor):
_API_BASE_URL = 'http://tv.adobe.com/api/v4/'
class AdobeTVIE(AdobeTVBaseIE):
_VALID_URL = r'https?://tv\.adobe\.com/(?:(?P<language>fr|de|es|jp)/)?watch/(?P<show_urlname>[^/]+)/(?P<id>[^/]+)'
_TEST = {
'url': 'http://tv.adobe.com/watch/the-complete-picture-with-julieanne-kost/quick-tip-how-to-draw-a-circle-around-an-object-in-photoshop/',
'md5': '9bc5727bcdd55251f35ad311ca74fa1e',
'info_dict': {
'id': 'quick-tip-how-to-draw-a-circle-around-an-object-in-photoshop',
'id': '10981',
'ext': 'mp4',
'title': 'Quick Tip - How to Draw a Circle Around an Object in Photoshop',
'description': 'md5:99ec318dc909d7ba2a1f2b038f7d2311',
@ -29,50 +38,106 @@ class AdobeTVIE(InfoExtractor):
}
def _real_extract(self, url):
video_id = self._match_id(url)
webpage = self._download_webpage(url, video_id)
language, show_urlname, urlname = re.match(self._VALID_URL, url).groups()
if not language:
language = 'en'
player = self._parse_json(
self._search_regex(r'html5player:\s*({.+?})\s*\n', webpage, 'player'),
video_id)
title = player.get('title') or self._search_regex(
r'data-title="([^"]+)"', webpage, 'title')
description = self._og_search_description(webpage)
thumbnail = self._og_search_thumbnail(webpage)
upload_date = unified_strdate(
self._html_search_meta('datepublished', webpage, 'upload date'))
duration = parse_duration(
self._html_search_meta('duration', webpage, 'duration') or
self._search_regex(
r'Runtime:\s*(\d{2}:\d{2}:\d{2})',
webpage, 'duration', fatal=False))
view_count = str_to_int(self._search_regex(
r'<div class="views">\s*Views?:\s*([\d,.]+)\s*</div>',
webpage, 'view count'))
video_data = self._download_json(
self._API_BASE_URL + 'episode/get/?language=%s&show_urlname=%s&urlname=%s&disclosure=standard' % (language, show_urlname, urlname),
urlname)['data'][0]
formats = [{
'url': source['src'],
'format_id': source.get('quality') or source['src'].split('-')[-1].split('.')[0] or None,
'tbr': source.get('bitrate'),
} for source in player['sources']]
'url': source['url'],
'format_id': source.get('quality_level') or source['url'].split('-')[-1].split('.')[0] or None,
'width': int_or_none(source.get('width')),
'height': int_or_none(source.get('height')),
'tbr': int_or_none(source.get('video_data_rate')),
} for source in video_data['videos']]
self._sort_formats(formats)
return {
'id': video_id,
'title': title,
'description': description,
'thumbnail': thumbnail,
'upload_date': upload_date,
'duration': duration,
'view_count': view_count,
'id': compat_str(video_data['id']),
'title': video_data['title'],
'description': video_data.get('description'),
'thumbnail': video_data.get('thumbnail'),
'upload_date': unified_strdate(video_data.get('start_date')),
'duration': parse_duration(video_data.get('duration')),
'view_count': str_to_int(video_data.get('playcount')),
'formats': formats,
}
class AdobeTVPlaylistBaseIE(AdobeTVBaseIE):
def _parse_page_data(self, page_data):
return [self.url_result(self._get_element_url(element_data)) for element_data in page_data]
def _extract_playlist_entries(self, url, display_id):
page = self._download_json(url, display_id)
entries = self._parse_page_data(page['data'])
for page_num in range(2, page['paging']['pages'] + 1):
entries.extend(self._parse_page_data(
self._download_json(url + '&page=%d' % page_num, display_id)['data']))
return entries
class AdobeTVShowIE(AdobeTVPlaylistBaseIE):
_VALID_URL = r'https?://tv\.adobe\.com/(?:(?P<language>fr|de|es|jp)/)?show/(?P<id>[^/]+)'
_TEST = {
'url': 'http://tv.adobe.com/show/the-complete-picture-with-julieanne-kost',
'info_dict': {
'id': '36',
'title': 'The Complete Picture with Julieanne Kost',
'description': 'md5:fa50867102dcd1aa0ddf2ab039311b27',
},
'playlist_mincount': 136,
}
def _get_element_url(self, element_data):
return element_data['urls'][0]
def _real_extract(self, url):
language, show_urlname = re.match(self._VALID_URL, url).groups()
if not language:
language = 'en'
query = 'language=%s&show_urlname=%s' % (language, show_urlname)
show_data = self._download_json(self._API_BASE_URL + 'show/get/?%s' % query, show_urlname)['data'][0]
return self.playlist_result(
self._extract_playlist_entries(self._API_BASE_URL + 'episode/?%s' % query, show_urlname),
compat_str(show_data['id']),
show_data['show_name'],
show_data['show_description'])
class AdobeTVChannelIE(AdobeTVPlaylistBaseIE):
_VALID_URL = r'https?://tv\.adobe\.com/(?:(?P<language>fr|de|es|jp)/)?channel/(?P<id>[^/]+)(?:/(?P<category_urlname>[^/]+))?'
_TEST = {
'url': 'http://tv.adobe.com/channel/development',
'info_dict': {
'id': 'development',
},
'playlist_mincount': 96,
}
def _get_element_url(self, element_data):
return element_data['url']
def _real_extract(self, url):
language, channel_urlname, category_urlname = re.match(self._VALID_URL, url).groups()
if not language:
language = 'en'
query = 'language=%s&channel_urlname=%s' % (language, channel_urlname)
if category_urlname:
query += '&category_urlname=%s' % category_urlname
return self.playlist_result(
self._extract_playlist_entries(self._API_BASE_URL + 'show/?%s' % query, channel_urlname),
channel_urlname)
class AdobeTVVideoIE(InfoExtractor):
_VALID_URL = r'https?://video\.tv\.adobe\.com/v/(?P<id>\d+)'
@ -91,28 +156,25 @@ class AdobeTVVideoIE(InfoExtractor):
def _real_extract(self, url):
video_id = self._match_id(url)
webpage = self._download_webpage(url, video_id)
player_params = self._parse_json(self._search_regex(
r'var\s+bridge\s*=\s*([^;]+);', webpage, 'player parameters'),
video_id)
video_data = self._download_json(url + '?format=json', video_id)
formats = [{
'format_id': '%s-%s' % (determine_ext(source['src']), source.get('height')),
'url': source['src'],
'width': source.get('width'),
'height': source.get('height'),
'tbr': source.get('bitrate'),
} for source in player_params['sources']]
'width': int_or_none(source.get('width')),
'height': int_or_none(source.get('height')),
'tbr': int_or_none(source.get('bitrate')),
} for source in video_data['sources']]
self._sort_formats(formats)
# For both metadata and downloaded files the duration varies among
# formats. I just pick the max one
duration = max(filter(None, [
float_or_none(source.get('duration'), scale=1000)
for source in player_params['sources']]))
for source in video_data['sources']]))
subtitles = {}
for translation in player_params.get('translations', []):
for translation in video_data.get('translations', []):
lang_id = translation.get('language_w3c') or ISO639Utils.long2short(translation['language_medium'])
if lang_id not in subtitles:
subtitles[lang_id] = []
@ -124,8 +186,9 @@ class AdobeTVVideoIE(InfoExtractor):
return {
'id': video_id,
'formats': formats,
'title': player_params['title'],
'description': self._og_search_description(webpage),
'title': video_data['title'],
'description': video_data.get('description'),
'thumbnail': video_data['video'].get('poster'),
'duration': duration,
'subtitles': subtitles,
}