From f0b6136f1184a8411a7566c85673af4ba3a21588 Mon Sep 17 00:00:00 2001 From: Alex Vong Date: Mon, 4 Feb 2019 10:00:03 +0800 Subject: [PATCH 1/3] [thisav] Fix video extraction. --- youtube_dl/extractor/thisav.py | 17 +++++++++++------ 1 file changed, 11 insertions(+), 6 deletions(-) diff --git a/youtube_dl/extractor/thisav.py b/youtube_dl/extractor/thisav.py index dc3dd03c8..505ba809e 100644 --- a/youtube_dl/extractor/thisav.py +++ b/youtube_dl/extractor/thisav.py @@ -4,7 +4,7 @@ from __future__ import unicode_literals import re from .common import InfoExtractor -from ..utils import remove_end +from ..utils import urljoin, xpath_text class ThisAVIE(InfoExtractor): @@ -38,11 +38,16 @@ class ThisAVIE(InfoExtractor): video_id = mobj.group('id') webpage = self._download_webpage(url, video_id) - title = remove_end(self._html_search_regex( - r'([^<]+)', webpage, 'title'), - ' - 視頻 - ThisAV.com-世界第一中文成人娛樂網站') - video_url = self._html_search_regex( - r"addVariable\('file','([^']+)'\);", webpage, 'video url', default=None) + title = self._html_search_regex(r'(?:([^<]+)-\s*視頻\s*-\s*ThisAV.com\s*-\s*世界第一中文成人娛樂網站\s*|([^<]+))', webpage, 'title') + + mpd_url = self._html_search_regex( + r'"(https?://[^"]+\.mpd)"', webpage, 'MPD URL') + mpd_root = self._download_xml(mpd_url, video_id) + mpd_ns = self._search_regex( + r'\{([^}]+)\}', mpd_root.tag, 'MPD namespace', default=None) + base_url = xpath_text(mpd_root, self._xpath_ns('.//BaseURL', mpd_ns)) + video_url = urljoin(mpd_url, base_url) + if video_url: info_dict = { 'formats': [{ From f4bd1878458ad82c42e26b55b3c169d9542fae42 Mon Sep 17 00:00:00 2001 From: Alex Vong Date: Mon, 4 Feb 2019 10:04:18 +0800 Subject: [PATCH 2/3] [thisav] Fix uploader extraction. --- youtube_dl/extractor/thisav.py | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/youtube_dl/extractor/thisav.py b/youtube_dl/extractor/thisav.py index 505ba809e..591df8fb7 100644 --- a/youtube_dl/extractor/thisav.py +++ b/youtube_dl/extractor/thisav.py @@ -61,11 +61,12 @@ class ThisAVIE(InfoExtractor): else: info_dict = self._extract_jwplayer_data( webpage, video_id, require_title=False) + uploader = self._html_search_regex( - r': ([^<]+)', + r': ([^<]+)', webpage, 'uploader name', fatal=False) uploader_id = self._html_search_regex( - r': (?:[^<]+)', + r': (?:[^<]+)', webpage, 'uploader id', fatal=False) info_dict.update({ From 790b0ec3b300a8c79ffd8dd34b07411ca011953f Mon Sep 17 00:00:00 2001 From: Alex Vong Date: Mon, 4 Feb 2019 10:05:39 +0800 Subject: [PATCH 3/3] [thisav] Fix tests. --- youtube_dl/extractor/thisav.py | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/youtube_dl/extractor/thisav.py b/youtube_dl/extractor/thisav.py index 591df8fb7..c1ad4a724 100644 --- a/youtube_dl/extractor/thisav.py +++ b/youtube_dl/extractor/thisav.py @@ -9,20 +9,20 @@ from ..utils import urljoin, xpath_text class ThisAVIE(InfoExtractor): _VALID_URL = r'https?://(?:www\.)?thisav\.com/video/(?P[0-9]+)/.*' + + # all videos are now html5 videos _TESTS = [{ - # jwplayer - 'url': 'http://www.thisav.com/video/47734/%98%26sup1%3B%83%9E%83%82---just-fit.html', - 'md5': '0480f1ef3932d901f0e0e719f188f19b', + 'url': 'https://www.thisav.com/video/47734/%98%26sup1%3B%83%9E%83%82---just-fit.html', + 'md5': 'adad2d2e989b4524698b4fb856d6719f', 'info_dict': { 'id': '47734', - 'ext': 'flv', + 'ext': 'mp4', 'title': '高樹マリア - Just fit', 'uploader': 'dj7970', 'uploader_id': 'dj7970' } }, { - # html5 media - 'url': 'http://www.thisav.com/video/242352/nerdy-18yo-big-ass-tattoos-and-glasses.html', + 'url': 'https://www.thisav.com/video/242352/nerdy-18yo-big-ass-tattoos-and-glasses.html', 'md5': 'ba90c076bd0f80203679e5b60bf523ee', 'info_dict': { 'id': '242352',