From d90e40305bee84f5e3cd4927c729e1d16bbd3dc6 Mon Sep 17 00:00:00 2001 From: remitamine Date: Sat, 17 Oct 2015 17:28:09 +0100 Subject: [PATCH 1/4] [bilibili] fix info extraction --- youtube_dl/extractor/bilibili.py | 151 ++++++++++++------------------- 1 file changed, 57 insertions(+), 94 deletions(-) diff --git a/youtube_dl/extractor/bilibili.py b/youtube_dl/extractor/bilibili.py index ecc17ebeb..8f23a30ab 100644 --- a/youtube_dl/extractor/bilibili.py +++ b/youtube_dl/extractor/bilibili.py @@ -1,15 +1,12 @@ # coding: utf-8 from __future__ import unicode_literals -import re -import itertools import json import xml.etree.ElementTree as ET from .common import InfoExtractor from ..utils import ( int_or_none, - unified_strdate, ExtractorError, ) @@ -21,12 +18,15 @@ class BiliBiliIE(InfoExtractor): 'url': 'http://www.bilibili.tv/video/av1074402/', 'md5': '2c301e4dab317596e837c3e7633e7d86', 'info_dict': { - 'id': '1074402_part1', + 'id': '1554319', 'ext': 'flv', 'title': '【金坷垃】金泡沫', - 'duration': 308, + 'duration': 308313, 'upload_date': '20140420', 'thumbnail': 're:^https?://.+\.jpg', + 'description': 'md5:ce18c2a2d2193f0df2917d270f2e5923', + 'timestamp': 1397983878, + 'uploader': '菊子桑', }, }, { 'url': 'http://www.bilibili.com/video/av1041170/', @@ -34,109 +34,72 @@ class BiliBiliIE(InfoExtractor): 'id': '1041170', 'title': '【BD1080P】刀语【诸神&异域】', }, - 'playlist_count': 9, + 'playlist_count': 12, }] - def _real_extract(self, url): - video_id = self._match_id(url) - webpage = self._download_webpage(url, video_id) + def _extract_video_info(self, cid, view_data, page_num=1, num_pages=1): + title = view_data['title'] - if '(此视频不存在或被删除)' in webpage: - raise ExtractorError( - 'The video does not exist or was deleted', expected=True) - - if '>你没有权限浏览！ 由于版权相关问题 我们不对您所在的地区提供服务<' in webpage: - raise ExtractorError( - 'The video is not available in your region due to copyright reasons', - expected=True) - - video_code = self._search_regex( - r'(?s)
(.*?)
', webpage, 'video code') - - title = self._html_search_meta( - 'media:title', video_code, 'title', fatal=True) - duration_str = self._html_search_meta( - 'duration', video_code, 'duration') - if duration_str is None: - duration = None - else: - duration_mobj = re.match( - r'^T(?:(?P[0-9]+)H)?(?P[0-9]+)M(?P[0-9]+)S$', - duration_str) - duration = ( - int_or_none(duration_mobj.group('hours'), default=0) * 3600 + - int(duration_mobj.group('minutes')) * 60 + - int(duration_mobj.group('seconds'))) - upload_date = unified_strdate(self._html_search_meta( - 'uploadDate', video_code, fatal=False)) - thumbnail = self._html_search_meta( - 'thumbnailUrl', video_code, 'thumbnail', fatal=False) - - cid = self._search_regex(r'cid=(\d+)', webpage, 'cid') - - entries = [] - - lq_page = self._download_webpage( - 'http://interface.bilibili.com/v_cdn_play?appkey=1&cid=%s' % cid, - video_id, - note='Downloading LQ video info' + page = self._download_webpage( + 'http://interface.bilibili.com/v_cdn_play?appkey=8e9fc618fbd41e28&cid=%s' % cid, + cid, + 'Downloading page %d/%d' % (page_num, num_pages) ) try: - err_info = json.loads(lq_page) + err_info = json.loads(page) raise ExtractorError( 'BiliBili said: ' + err_info['error_text'], expected=True) except ValueError: pass - lq_doc = ET.fromstring(lq_page) - lq_durls = lq_doc.findall('./durl') + doc = ET.fromstring(page) + durls = doc.findall('./durl') - hq_doc = self._download_xml( - 'http://interface.bilibili.com/playurl?appkey=1&cid=%s' % cid, - video_id, - note='Downloading HQ video info', - fatal=False, - ) - if hq_doc is not False: - hq_durls = hq_doc.findall('./durl') - assert len(lq_durls) == len(hq_durls) - else: - hq_durls = itertools.repeat(None) - - i = 1 - for lq_durl, hq_durl in zip(lq_durls, hq_durls): - formats = [{ - 'format_id': 'lq', - 'quality': 1, - 'url': lq_durl.find('./url').text, - 'filesize': int_or_none( - lq_durl.find('./size'), get_attr='text'), - }] - if hq_durl is not None: - formats.append({ - 'format_id': 'hq', - 'quality': 2, - 'ext': 'flv', - 'url': hq_durl.find('./url').text, - 'filesize': int_or_none( - hq_durl.find('./size'), get_attr='text'), - }) - self._sort_formats(formats) + entries = [] + for durl in durls: entries.append({ - 'id': '%s_part%d' % (video_id, i), + 'id': '%s_part%s' % (cid, durl.find('./order').text), 'title': title, - 'formats': formats, - 'duration': duration, - 'upload_date': upload_date, - 'thumbnail': thumbnail, + 'url': durl.find('./url').text, + 'filesize': int_or_none(durl.find('./filesize').text), + 'ext': 'flv', + 'duration': int_or_none(durl.find('./length').text) // 1000, }) - i += 1 - - return { - '_type': 'multi_video', - 'entries': entries, - 'id': video_id, - 'title': title + info = { + 'id': cid, + 'title': title, + 'description': view_data.get('description'), + 'thumbnail': view_data.get('pic'), + 'uploader': view_data.get('author'), + 'timestamp': int_or_none(view_data.get('created')), + 'view_count': view_data.get('play'), + 'duration': int_or_none(doc.find('./timelength').text), } + + if len(entries) == 1: + entries[0].update(info) + return entries[0] + else: + info.update({ + '_type': 'multi_video', + 'entries': entries, + }) + return info + + def _real_extract(self, url): + video_id = self._match_id(url) + view_data = self._download_json('http://api.bilibili.com/view?type=json&appkey=8e9fc618fbd41e28&id=%s' % video_id, video_id) + + num_pages = int_or_none(view_data['pages']) + if num_pages > 1: + play_list_title = view_data['title'] + page_list = self._download_json('http://www.bilibili.com/widget/getPageList?aid=%s' % video_id, video_id, 'Downloading page list metadata') + entries = [] + for page in page_list: + view_data['title'] = page['pagename'] + entries.append(self._extract_video_info(str(page['cid']), view_data, page['page'], num_pages)) + return self.playlist_result(entries, video_id, play_list_title, view_data.get('description')) + else: + return self._extract_video_info(str(view_data['cid']), view_data) From 55af2b26e0f169bef2f10a7b5f6ec8e34c6dbb6d Mon Sep 17 00:00:00 2001 From: remitamine Date: Sat, 17 Oct 2015 18:30:51 +0100 Subject: [PATCH 2/4] [bilibili] extract backup url --- youtube_dl/extractor/bilibili.py | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) diff --git a/youtube_dl/extractor/bilibili.py b/youtube_dl/extractor/bilibili.py index 8f23a30ab..85156ce49 100644 --- a/youtube_dl/extractor/bilibili.py +++ b/youtube_dl/extractor/bilibili.py @@ -58,13 +58,21 @@ class BiliBiliIE(InfoExtractor): entries = [] for durl in durls: + formats = [] + backup_url = durl.find('./backup_url') + if backup_url is not None: + formats.append({'url': backup_url.find('./url').text}) + size = durl.find('./filesize|./size') + formats.append({ + 'url': durl.find('./url').text, + 'filesize': int_or_none(size.text) if size else None, + 'ext': 'flv', + }) entries.append({ 'id': '%s_part%s' % (cid, durl.find('./order').text), 'title': title, - 'url': durl.find('./url').text, - 'filesize': int_or_none(durl.find('./filesize').text), - 'ext': 'flv', 'duration': int_or_none(durl.find('./length').text) // 1000, + 'formats': formats, }) info = { From 520e753390aab5845bb257b964ebcd6f818455df Mon Sep 17 00:00:00 2001 From: remitamine Date: Sat, 17 Oct 2015 23:12:58 +0100 Subject: [PATCH 3/4] [bilibili] add support for specefic page extraction --- youtube_dl/extractor/bilibili.py | 45 ++++++++++++++++---------------- 1 file changed, 23 insertions(+), 22 deletions(-) diff --git a/youtube_dl/extractor/bilibili.py b/youtube_dl/extractor/bilibili.py index 85156ce49..e00efb6eb 100644 --- a/youtube_dl/extractor/bilibili.py +++ b/youtube_dl/extractor/bilibili.py @@ -1,18 +1,20 @@ # coding: utf-8 from __future__ import unicode_literals +import re import json import xml.etree.ElementTree as ET from .common import InfoExtractor from ..utils import ( int_or_none, + unescapeHTML, ExtractorError, ) class BiliBiliIE(InfoExtractor): - _VALID_URL = r'http://www\.bilibili\.(?:tv|com)/video/av(?P[0-9]+)/' + _VALID_URL = r'http://www\.bilibili\.(?:tv|com)/video/av(?P\d+)(?:/index_(?P\d+).html)?' _TESTS = [{ 'url': 'http://www.bilibili.tv/video/av1074402/', @@ -33,17 +35,31 @@ class BiliBiliIE(InfoExtractor): 'info_dict': { 'id': '1041170', 'title': '【BD1080P】刀语【诸神&异域】', + 'description': '这是个神奇的故事~每个人不留弹幕不给走哦~切利哦！~', + 'uploader': '枫叶逝去', + 'timestamp': 1396501299, }, - 'playlist_count': 12, + 'playlist_count': 9, }] - def _extract_video_info(self, cid, view_data, page_num=1, num_pages=1): - title = view_data['title'] + def _real_extract(self, url): + mobj = re.match(self._VALID_URL, url) + video_id = mobj.group('id') + page_num = mobj.group('page_num') or '1' + + view_data = self._download_json( + 'http://api.bilibili.com/view?type=json&appkey=8e9fc618fbd41e28&id=%s&page=%s' % (video_id, page_num), + video_id) + if 'error' in view_data: + raise ExtractorError('%s said: %s' % (self.IE_NAME, view_data['error']), expected=True) + + cid = view_data['cid'] + title = unescapeHTML(view_data['title']) page = self._download_webpage( 'http://interface.bilibili.com/v_cdn_play?appkey=8e9fc618fbd41e28&cid=%s' % cid, cid, - 'Downloading page %d/%d' % (page_num, num_pages) + 'Downloading page %s/%s' % (page_num, view_data['pages']) ) try: err_info = json.loads(page) @@ -76,7 +92,7 @@ class BiliBiliIE(InfoExtractor): }) info = { - 'id': cid, + 'id': str(cid), 'title': title, 'description': view_data.get('description'), 'thumbnail': view_data.get('pic'), @@ -92,22 +108,7 @@ class BiliBiliIE(InfoExtractor): else: info.update({ '_type': 'multi_video', + 'id': video_id, 'entries': entries, }) return info - - def _real_extract(self, url): - video_id = self._match_id(url) - view_data = self._download_json('http://api.bilibili.com/view?type=json&appkey=8e9fc618fbd41e28&id=%s' % video_id, video_id) - - num_pages = int_or_none(view_data['pages']) - if num_pages > 1: - play_list_title = view_data['title'] - page_list = self._download_json('http://www.bilibili.com/widget/getPageList?aid=%s' % video_id, video_id, 'Downloading page list metadata') - entries = [] - for page in page_list: - view_data['title'] = page['pagename'] - entries.append(self._extract_video_info(str(page['cid']), view_data, page['page'], num_pages)) - return self.playlist_result(entries, video_id, play_list_title, view_data.get('description')) - else: - return self._extract_video_info(str(view_data['cid']), view_data) From 497f5fd93fe1efd0df8dc58d518c328ed1409457 Mon Sep 17 00:00:00 2001 From: remitamine Date: Wed, 21 Oct 2015 08:24:05 +0100 Subject: [PATCH 4/4] [bilibili] extract multiple backup_urls --- youtube_dl/extractor/bilibili.py | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-) diff --git a/youtube_dl/extractor/bilibili.py b/youtube_dl/extractor/bilibili.py index e00efb6eb..935fcc55c 100644 --- a/youtube_dl/extractor/bilibili.py +++ b/youtube_dl/extractor/bilibili.py @@ -69,21 +69,22 @@ class BiliBiliIE(InfoExtractor): pass doc = ET.fromstring(page) - durls = doc.findall('./durl') entries = [] - for durl in durls: - formats = [] - backup_url = durl.find('./backup_url') - if backup_url is not None: - formats.append({'url': backup_url.find('./url').text}) + for durl in doc.findall('./durl'): size = durl.find('./filesize|./size') - formats.append({ + formats = [{ 'url': durl.find('./url').text, 'filesize': int_or_none(size.text) if size else None, 'ext': 'flv', - }) + }] + backup_urls = durl.find('./backup_url') + if backup_urls is not None: + for backup_url in backup_urls.findall('./url'): + formats.append({'url': backup_url.text}) + formats.reverse() + entries.append({ 'id': '%s_part%s' % (cid, durl.find('./order').text), 'title': title,