From 45024183aea169dc898902388f782485de02cbac Mon Sep 17 00:00:00 2001 From: Mattias Wadman Date: Fri, 3 Feb 2017 05:10:13 +0100 Subject: [PATCH] [infoq] Add audio only format if available (#11565) * [infoq] Add audio only format if available Refactor cookie code into a function. Renamed formats to http_video, http_audio, rtmp_video Renamed extract functions to video instead of videos as they return one or no video. * [infoq] Rename to _extract_cookies as it more than one * [infoq] Remove redundant determine_ext * [infoq] Add comment about hardcoded URL * [infoq] Use _hidden_inputs instead of messy regex * [infoq] Probe if audio URL is valid Make it possible to pass headers to _is_valid_url * [infoq] Add audio only test --- youtube_dl/extractor/common.py | 4 +-- youtube_dl/extractor/infoq.py | 63 ++++++++++++++++++++++++++++------ 2 files changed, 55 insertions(+), 12 deletions(-) diff --git a/youtube_dl/extractor/common.py b/youtube_dl/extractor/common.py index 5a15a9536..2c8ec1417 100644 --- a/youtube_dl/extractor/common.py +++ b/youtube_dl/extractor/common.py @@ -1025,13 +1025,13 @@ def _remove_duplicate_formats(formats): unique_formats.append(f) formats[:] = unique_formats - def _is_valid_url(self, url, video_id, item='video'): + def _is_valid_url(self, url, video_id, item='video', headers={}): url = self._proto_relative_url(url, scheme='http:') # For now assume non HTTP(S) URLs always valid if not (url.startswith('http://') or url.startswith('https://')): return True try: - self._request_webpage(url, video_id, 'Checking %s URL' % item) + self._request_webpage(url, video_id, 'Checking %s URL' % item, headers=headers) return True except ExtractorError as e: if isinstance(e.cause, compat_urllib_error.URLError): diff --git a/youtube_dl/extractor/infoq.py b/youtube_dl/extractor/infoq.py index cca0b8a93..9fb71e8ef 100644 --- a/youtube_dl/extractor/infoq.py +++ b/youtube_dl/extractor/infoq.py @@ -4,7 +4,10 @@ import base64 -from ..compat import compat_urllib_parse_unquote +from ..compat import ( + compat_urllib_parse_unquote, + compat_urlparse, +) from ..utils import determine_ext from .bokecc import BokeCCBaseIE @@ -33,9 +36,21 @@ class InfoQIE(BokeCCBaseIE): 'ext': 'flv', 'description': 'md5:308d981fb28fa42f49f9568322c683ff', }, + }, { + 'url': 'https://www.infoq.com/presentations/Simple-Made-Easy', + 'md5': '0e34642d4d9ef44bf86f66f6399672db', + 'info_dict': { + 'id': 'Simple-Made-Easy', + 'title': 'Simple Made Easy', + 'ext': 'mp3', + 'description': 'md5:3e0e213a8bbd074796ef89ea35ada25b', + }, + 'params': { + 'format': 'bestaudio', + }, }] - def _extract_rtmp_videos(self, webpage): + def _extract_rtmp_video(self, webpage): # The server URL is hardcoded video_url = 'rtmpe://video.infoq.com/cfx/st/' @@ -47,28 +62,53 @@ def _extract_rtmp_videos(self, webpage): playpath = 'mp4:' + real_id return [{ - 'format_id': 'rtmp', + 'format_id': 'rtmp_video', 'url': video_url, 'ext': determine_ext(playpath), 'play_path': playpath, }] - def _extract_http_videos(self, webpage): - http_video_url = self._search_regex(r'P\.s\s*=\s*\'([^\']+)\'', webpage, 'video URL') - + def _extract_cookies(self, webpage): policy = self._search_regex(r'InfoQConstants.scp\s*=\s*\'([^\']+)\'', webpage, 'policy') signature = self._search_regex(r'InfoQConstants.scs\s*=\s*\'([^\']+)\'', webpage, 'signature') key_pair_id = self._search_regex(r'InfoQConstants.sck\s*=\s*\'([^\']+)\'', webpage, 'key-pair-id') + return 'CloudFront-Policy=%s; CloudFront-Signature=%s; CloudFront-Key-Pair-Id=%s' % ( + policy, signature, key_pair_id) + def _extract_http_video(self, webpage): + http_video_url = self._search_regex(r'P\.s\s*=\s*\'([^\']+)\'', webpage, 'video URL') return [{ - 'format_id': 'http', + 'format_id': 'http_video', 'url': http_video_url, 'http_headers': { - 'Cookie': 'CloudFront-Policy=%s; CloudFront-Signature=%s; CloudFront-Key-Pair-Id=%s' % ( - policy, signature, key_pair_id), + 'Cookie': self._extract_cookies(webpage) }, }] + def _extract_http_audio(self, webpage, video_id): + fields = self._hidden_inputs(webpage) + http_audio_url = fields['filename'] + if http_audio_url is None: + return [] + + cookies_header = {'Cookie': self._extract_cookies(webpage)} + + # base URL is found in the Location header in the response returned by + # GET https://www.infoq.com/mp3download.action?filename=... when logged in. + http_audio_url = compat_urlparse.urljoin('http://res.infoq.com/downloads/mp3downloads/', http_audio_url) + + # audio file seem to be missing some times even if there is a download link + # so probe URL to make sure + if not self._is_valid_url(http_audio_url, video_id, headers=cookies_header): + return [] + + return [{ + 'format_id': 'http_audio', + 'url': http_audio_url, + 'vcodec': 'none', + 'http_headers': cookies_header, + }] + def _real_extract(self, url): video_id = self._match_id(url) webpage = self._download_webpage(url, video_id) @@ -80,7 +120,10 @@ def _real_extract(self, url): # for China videos, HTTP video URL exists but always fails with 403 formats = self._extract_bokecc_formats(webpage, video_id) else: - formats = self._extract_rtmp_videos(webpage) + self._extract_http_videos(webpage) + formats = ( + self._extract_rtmp_video(webpage) + + self._extract_http_video(webpage) + + self._extract_http_audio(webpage, video_id)) self._sort_formats(formats)