From 6d3a7d03e14fcbc704bf30d305fb95c5829e55a6 Mon Sep 17 00:00:00 2001 From: huohuarong Date: Fri, 2 Aug 2013 15:26:11 +0800 Subject: [PATCH 1/5] fix bug: kankan extractor not support http://vod.kankan.com/v/70/70309.shtml --- youtube_dl/extractor/kankan.py | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/youtube_dl/extractor/kankan.py b/youtube_dl/extractor/kankan.py index 8537ba584..445d46501 100644 --- a/youtube_dl/extractor/kankan.py +++ b/youtube_dl/extractor/kankan.py @@ -21,8 +21,10 @@ def _real_extract(self, url): video_id = mobj.group('id') webpage = self._download_webpage(url, video_id) - title = self._search_regex(r'G_TITLE=[\'"](.+?)[\'"]', webpage, u'video title') - gcid = self._search_regex(r'lurl:[\'"]http://.+?/.+?/(.+?)/', webpage, u'gcid') + title = self._search_regex(r'(?:G_TITLE=|G_MOVIE_TITLE = )[\'"](.+?)[\'"]', webpage, u'video title') + surls = re.search(r'surls:\[\'.+?\'\]|lurl:\'.+?\.flv\'', webpage).group(0) + gcids = re.findall(r"http://.+?/.+?/(.+?)/", surls) + gcid = gcids[-1] video_info_page = self._download_webpage('http://p2s.cl.kankan.com/getCdnresource_flv?gcid=%s' % gcid, video_id, u'Downloading video url info') From 6624a2b07dafad4de895b4e84f4595214817518d Mon Sep 17 00:00:00 2001 From: huohuarong Date: Fri, 2 Aug 2013 17:58:46 +0800 Subject: [PATCH 2/5] add an extractor for tv.sohu.com --- youtube_dl/extractor/__init__.py | 1 + youtube_dl/extractor/sohu.py | 97 ++++++++++++++++++++++++++++++++ 2 files changed, 98 insertions(+) create mode 100644 youtube_dl/extractor/sohu.py diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py index c20172a53..3a08d676f 100644 --- a/youtube_dl/extractor/__init__.py +++ b/youtube_dl/extractor/__init__.py @@ -55,6 +55,7 @@ from .ringtv import RingTVIE from .roxwel import RoxwelIE from .sina import SinaIE +from .sohu import SohuIE from .soundcloud import SoundcloudIE, SoundcloudSetIE from .spiegel import SpiegelIE from .stanfordoc import StanfordOpenClassroomIE diff --git a/youtube_dl/extractor/sohu.py b/youtube_dl/extractor/sohu.py new file mode 100644 index 000000000..830814221 --- /dev/null +++ b/youtube_dl/extractor/sohu.py @@ -0,0 +1,97 @@ +# encoding: utf-8 + +import re +import json +import time +import logging +import urllib2 + +from .common import InfoExtractor +from ..utils import compat_urllib_request + + +class SohuIE(InfoExtractor): + _VALID_URL = r'https?://tv\.sohu\.com/\d+?/n(?P\d+)\.shtml.*?' + + _TEST = { + u'url': u'http://tv.sohu.com/20130724/n382479172.shtml#super', + u'file': u'382479172.flv', + u'md5': u'cc84eed6b6fbf0f2f9a8d3cb9da1939b', + u'info_dict': { + u'title': u'The Illest - Far East Movement Riff Raff', + }, + } + + def _clearn_html(self, string): + tags = re.findall(r'<.+?>', string) + for t in tags: + string = string.replace(t, ' ') + for i in range(2): + spaces = re.findall(r'\s+', string) + for s in spaces: + string = string.replace(s, ' ') + string = string.strip() + return string + + def _real_extract(self, url): + mobj = re.match(self._VALID_URL, url) + video_id = mobj.group('id') + webpage = self._download_webpage(url, video_id) + pattern = r'

\n*?(.+?)\n*?

' + compiled = re.compile(pattern, re.DOTALL) + title = self._search_regex(compiled, webpage, u'video title').strip('\t\n') + title = self._clearn_html(title) + pattern = re.compile(r'var vid="(\d+)"') + result = re.search(pattern, webpage) + if not result: + logging.info('[Sohu] could not get vid') + return None + vid = result.group(1) + logging.info('vid: %s' % vid) + base_url_1 = 'http://hot.vrs.sohu.com/vrs_flash.action?vid=' + url_1 = base_url_1 + vid + logging.info('json url: %s' % url_1) + json_1 = json.loads(urllib2.urlopen(url_1).read()) + # get the highest definition video vid and json infomation. + vids = [] + qualities = ('oriVid', 'superVid', 'highVid', 'norVid') + for vid_name in qualities: + vids.append(json_1['data'][vid_name]) + clearest_vid = 0 + for i, v in enumerate(vids): + if v != 0: + clearest_vid = v + logging.info('quality definition: %s' % qualities[i][:-3]) + break + if not clearest_vid: + logging.warning('could not find valid clearest_vid') + return None + if vid != clearest_vid: + url_1 = '%s%d' % (base_url_1, clearest_vid) + logging.info('highest definition json url: %s' % url_1) + json_1 = json.loads(urllib2.urlopen(url_1).read()) + allot = json_1['allot'] + prot = json_1['prot'] + clipsURL = json_1['data']['clipsURL'] + su = json_1['data']['su'] + num_of_parts = json_1['data']['totalBlocks'] + logging.info('Total parts: %d' % num_of_parts) + base_url_3 = 'http://allot/?prot=prot&file=clipsURL[i]&new=su[i]' + files_info = [] + for i in range(num_of_parts): + middle_url = 'http://%s/?prot=%s&file=%s&new=%s' % (allot, prot, clipsURL[i], su[i]) + logging.info('middle url part %d: %s' % (i, middle_url)) + middle_info = urllib2.urlopen(middle_url).read().split('|') + middle_part_1 = middle_info[0] + download_url = '%s%s?key=%s' % (middle_info[0], su[i], middle_info[3]) + + info = { + 'id': '%s_part%02d' % (video_id, i + 1), + 'title': title, + 'url': download_url, + 'ext': 'mp4', + } + files_info.append(info) + time.sleep(1) + + return files_info From 4ec929dc9b55a2588b4a27e64871c5bfa900bf37 Mon Sep 17 00:00:00 2001 From: huohuarong Date: Sat, 3 Aug 2013 10:29:58 +0800 Subject: [PATCH 3/5] use ..utils/clean_html() --- youtube_dl/extractor/sohu.py | 19 ++++++------------- 1 file changed, 6 insertions(+), 13 deletions(-) diff --git a/youtube_dl/extractor/sohu.py b/youtube_dl/extractor/sohu.py index 830814221..cf0ab5478 100644 --- a/youtube_dl/extractor/sohu.py +++ b/youtube_dl/extractor/sohu.py @@ -7,7 +7,7 @@ import urllib2 from .common import InfoExtractor -from ..utils import compat_urllib_request +from ..utils import compat_urllib_request, clean_html class SohuIE(InfoExtractor): @@ -22,16 +22,6 @@ class SohuIE(InfoExtractor): }, } - def _clearn_html(self, string): - tags = re.findall(r'<.+?>', string) - for t in tags: - string = string.replace(t, ' ') - for i in range(2): - spaces = re.findall(r'\s+', string) - for s in spaces: - string = string.replace(s, ' ') - string = string.strip() - return string def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) @@ -40,7 +30,7 @@ def _real_extract(self, url): pattern = r'

\n*?(.+?)\n*?

' compiled = re.compile(pattern, re.DOTALL) title = self._search_regex(compiled, webpage, u'video title').strip('\t\n') - title = self._clearn_html(title) + title = clean_html(title) pattern = re.compile(r'var vid="(\d+)"') result = re.search(pattern, webpage) if not result: @@ -93,5 +83,8 @@ def _real_extract(self, url): } files_info.append(info) time.sleep(1) - + if num_of_parts == 1: + info = files_info[0] + info['id'] = video_id + return info return files_info From b5a6d408181c118bf51382f486a2492643ed74ec Mon Sep 17 00:00:00 2001 From: huohuarong Date: Mon, 5 Aug 2013 22:51:54 +0800 Subject: [PATCH 4/5] fix parse title bug --- youtube_dl/extractor/sohu.py | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/youtube_dl/extractor/sohu.py b/youtube_dl/extractor/sohu.py index cf0ab5478..cd049b6f0 100644 --- a/youtube_dl/extractor/sohu.py +++ b/youtube_dl/extractor/sohu.py @@ -27,10 +27,10 @@ def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) video_id = mobj.group('id') webpage = self._download_webpage(url, video_id) - pattern = r'

\n*?(.+?)\n*?

' + pattern = r'(.+?)' compiled = re.compile(pattern, re.DOTALL) - title = self._search_regex(compiled, webpage, u'video title').strip('\t\n') - title = clean_html(title) + title = self._search_regex(compiled, webpage, u'video title') + title = clean_html(title).split('-')[0].strip() pattern = re.compile(r'var vid="(\d+)"') result = re.search(pattern, webpage) if not result: @@ -41,7 +41,8 @@ def _real_extract(self, url): base_url_1 = 'http://hot.vrs.sohu.com/vrs_flash.action?vid=' url_1 = base_url_1 + vid logging.info('json url: %s' % url_1) - json_1 = json.loads(urllib2.urlopen(url_1).read()) + webpage = self._download_webpage(url_1, vid) + json_1 = json.loads(webpage) # get the highest definition video vid and json infomation. vids = [] qualities = ('oriVid', 'superVid', 'highVid', 'norVid') From d5b00ee6e0ba70fd5d87752e8772fc1c39e4bd59 Mon Sep 17 00:00:00 2001 From: huohuarong Date: Tue, 6 Aug 2013 10:26:57 +0800 Subject: [PATCH 5/5] improve sohu extractor --- youtube_dl/extractor/sohu.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/youtube_dl/extractor/sohu.py b/youtube_dl/extractor/sohu.py index cd049b6f0..24fc3a5d7 100644 --- a/youtube_dl/extractor/sohu.py +++ b/youtube_dl/extractor/sohu.py @@ -31,6 +31,7 @@ def _real_extract(self, url): compiled = re.compile(pattern, re.DOTALL) title = self._search_regex(compiled, webpage, u'video title') title = clean_html(title).split('-')[0].strip() + self.to_screen('Title: %s' % title) pattern = re.compile(r'var vid="(\d+)"') result = re.search(pattern, webpage) if not result: @@ -70,6 +71,7 @@ def _real_extract(self, url): base_url_3 = 'http://allot/?prot=prot&file=clipsURL[i]&new=su[i]' files_info = [] for i in range(num_of_parts): + self.to_screen('Geting json infomation of part %s/%s' % (i + 1, num_of_parts)) middle_url = 'http://%s/?prot=%s&file=%s&new=%s' % (allot, prot, clipsURL[i], su[i]) logging.info('middle url part %d: %s' % (i, middle_url)) middle_info = urllib2.urlopen(middle_url).read().split('|')