X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;ds=sidebyside;f=youtube_dl%2Fextractor%2Fpbs.py;h=683c81de3444d09c8abb017c0c7ecc943c21db00;hb=8de28761c41ca0451e3bbe4b9ee236d6651cca44;hp=8f140d62660b896f5a6f819d621a762d13fbdb69;hpb=27aede907436fa58600cd46bb04e7eae6e1e9279;p=youtube-dl
diff --git a/youtube_dl/extractor/pbs.py b/youtube_dl/extractor/pbs.py
index 8f140d626..683c81de3 100644
--- a/youtube_dl/extractor/pbs.py
+++ b/youtube_dl/extractor/pbs.py
@@ -1,9 +1,13 @@
+# coding: utf-8
from __future__ import unicode_literals
import re
from .common import InfoExtractor
from ..utils import (
+ ExtractorError,
+ determine_ext,
+ int_or_none,
unified_strdate,
US_RATINGS,
)
@@ -28,10 +32,13 @@ class PBSIE(InfoExtractor):
'info_dict': {
'id': '2365006249',
'ext': 'mp4',
- 'title': 'A More Perfect Union',
+ 'title': 'Constitution USA with Peter Sagal - A More Perfect Union',
'description': 'md5:ba0c207295339c8d6eced00b7c363c6a',
'duration': 3190,
},
+ 'params': {
+ 'skip_download': True, # requires ffmpeg
+ },
},
{
'url': 'http://www.pbs.org/wgbh/pages/frontline/losing-iraq/',
@@ -39,10 +46,13 @@ class PBSIE(InfoExtractor):
'info_dict': {
'id': '2365297690',
'ext': 'mp4',
- 'title': 'Losing Iraq',
+ 'title': 'FRONTLINE - Losing Iraq',
'description': 'md5:f5bfbefadf421e8bb8647602011caf8e',
'duration': 5050,
},
+ 'params': {
+ 'skip_download': True, # requires ffmpeg
+ }
},
{
'url': 'http://www.pbs.org/newshour/bb/education-jan-june12-cyberschools_02-23/',
@@ -50,7 +60,7 @@ class PBSIE(InfoExtractor):
'info_dict': {
'id': '2201174722',
'ext': 'mp4',
- 'title': 'Cyber Schools Gain Popularity, but Quality Questions Persist',
+ 'title': 'PBS NewsHour - Cyber Schools Gain Popularity, but Quality Questions Persist',
'description': 'md5:5871c15cba347c1b3d28ac47a73c7c28',
'duration': 801,
},
@@ -62,10 +72,13 @@ class PBSIE(InfoExtractor):
'id': '2365297708',
'ext': 'mp4',
'description': 'md5:68d87ef760660eb564455eb30ca464fe',
- 'title': 'Dudamel Conducts Verdi Requiem at the Hollywood Bowl - Full',
+ 'title': 'Great Performances - Dudamel Conducts Verdi Requiem at the Hollywood Bowl - Full',
'duration': 6559,
'thumbnail': 're:^https?://.*\.jpg$',
- }
+ },
+ 'params': {
+ 'skip_download': True, # requires ffmpeg
+ },
},
{
'url': 'http://www.pbs.org/wgbh/nova/earth/killer-typhoon.html',
@@ -75,13 +88,53 @@ class PBSIE(InfoExtractor):
'display_id': 'killer-typhoon',
'ext': 'mp4',
'description': 'md5:c741d14e979fc53228c575894094f157',
- 'title': 'Killer Typhoon',
+ 'title': 'NOVA - Killer Typhoon',
'duration': 3172,
'thumbnail': 're:^https?://.*\.jpg$',
'upload_date': '20140122',
- }
+ 'age_limit': 10,
+ },
+ 'params': {
+ 'skip_download': True, # requires ffmpeg
+ },
+ },
+ {
+ 'url': 'http://www.pbs.org/wgbh/pages/frontline/united-states-of-secrets/',
+ 'info_dict': {
+ 'id': 'united-states-of-secrets',
+ },
+ 'playlist_count': 2,
+ },
+ {
+ 'url': 'http://www.pbs.org/wgbh/americanexperience/films/death/player/',
+ 'info_dict': {
+ 'id': '2280706814',
+ 'display_id': 'player',
+ 'ext': 'mp4',
+ 'title': 'American Experience - Death and the Civil War',
+ 'description': 'American Experience, TVâs most-watched history series, brings to life the compelling stories from our past that inform our understanding of the world today.',
+ 'duration': 6705,
+ 'thumbnail': 're:^https?://.*\.jpg$',
+ },
+ 'params': {
+ 'skip_download': True, # requires ffmpeg
+ },
+ },
+ {
+ 'url': 'http://video.pbs.org/video/2365367186/',
+ 'info_dict': {
+ 'id': '2365367186',
+ 'display_id': '2365367186',
+ 'ext': 'mp4',
+ 'title': 'To Catch A Comet - Full Episode',
+ 'description': 'On November 12, 2014, billions of kilometers from Earth, spacecraft orbiter Rosetta and lander Philae did what no other had dared to attempt \u2014 land on the volatile surface of a comet as it zooms around the sun at 67,000 km/hr. The European Space Agency hopes this mission can help peer into our past and unlock secrets of our origins.',
+ 'duration': 3342,
+ 'thumbnail': 're:^https?://.*\.jpg$',
+ },
+ 'params': {
+ 'skip_download': True, # requires ffmpeg
+ },
}
-
]
def _extract_webpage(self, url):
@@ -96,6 +149,12 @@ class PBSIE(InfoExtractor):
r']+class="videotab[^"]*"[^>]+vid="(\d+)"', webpage)
+ if tabbed_videos:
+ return tabbed_videos, presumptive_id, upload_date
+
MEDIA_ID_REGEXES = [
r"div\s*:\s*'videoembed'\s*,\s*mediaid\s*:\s*'(\d+)'", # frontline video embed
r'class="coveplayerid">([^<]+)<', # coveplayer
@@ -108,7 +167,7 @@ class PBSIE(InfoExtractor):
return media_id, presumptive_id, upload_date
url = self._search_regex(
- r'