diff options
author | Sergey M․ <dstftw@gmail.com> | 2019-12-31 23:29:06 +0700 |
---|---|---|
committer | Sergey M․ <dstftw@gmail.com> | 2019-12-31 23:29:06 +0700 |
commit | f41347260c2c2cf723bc2bb8a5c11f67a22175d5 (patch) | |
tree | 03ab47e00b46b1cdef117870fb1e693efa353349 /youtube_dl/extractor | |
parent | 060680874654e77cfd03d150a834b58213379c8c (diff) |
[pornhub] Fix extraction and add support for m3u8 formats (closes #22749, closes #23082)
Diffstat (limited to 'youtube_dl/extractor')
-rw-r--r-- | youtube_dl/extractor/pornhub.py | 47 |
1 files changed, 37 insertions, 10 deletions
diff --git a/youtube_dl/extractor/pornhub.py b/youtube_dl/extractor/pornhub.py index ba0ad7da2..75ed69cde 100644 --- a/youtube_dl/extractor/pornhub.py +++ b/youtube_dl/extractor/pornhub.py @@ -227,12 +227,13 @@ class PornHubIE(PornHubBaseIE): else: thumbnail, duration = [None] * 2 - if not video_urls: - tv_webpage = dl_webpage('tv') - + def extract_js_vars(webpage, pattern, fatal=True): assignments = self._search_regex( - r'(var.+?mediastring.+?)</script>', tv_webpage, - 'encoded url').split(';') + pattern, webpage, 'encoded url', fatal=fatal) + if not assignments: + return {} + + assignments = assignments.split(';') js_vars = {} @@ -254,11 +255,31 @@ class PornHubIE(PornHubBaseIE): assn = re.sub(r'var\s+', '', assn) vname, value = assn.split('=', 1) js_vars[vname] = parse_js_value(value) + return js_vars - video_url = js_vars['mediastring'] - if video_url not in video_urls_set: - video_urls.append((video_url, None)) - video_urls_set.add(video_url) + def add_video_url(video_url): + v_url = url_or_none(video_url) + if not v_url: + return + if v_url in video_urls_set: + return + video_urls.append((v_url, None)) + video_urls_set.add(v_url) + + if not video_urls: + FORMAT_PREFIXES = ('media', 'quality') + js_vars = extract_js_vars( + webpage, r'(var\s+(?:%s)_.+)' % '|'.join(FORMAT_PREFIXES), + fatal=False) + if js_vars: + for key, format_url in js_vars.items(): + if any(key.startswith(p) for p in FORMAT_PREFIXES): + add_video_url(format_url) + + if not video_urls: + js_vars = extract_js_vars( + dl_webpage('tv'), r'(var.+?mediastring.+?)</script>') + add_video_url(js_vars['mediastring']) for mobj in re.finditer( r'<a[^>]+\bclass=["\']downloadBtn\b[^>]+\bhref=(["\'])(?P<url>(?:(?!\1).)+)\1', @@ -276,10 +297,16 @@ class PornHubIE(PornHubBaseIE): r'/(\d{6}/\d{2})/', video_url, 'upload data', default=None) if upload_date: upload_date = upload_date.replace('/', '') - if determine_ext(video_url) == 'mpd': + ext = determine_ext(video_url) + if ext == 'mpd': formats.extend(self._extract_mpd_formats( video_url, video_id, mpd_id='dash', fatal=False)) continue + elif ext == 'm3u8': + formats.extend(self._extract_m3u8_formats( + video_url, video_id, 'mp4', entry_protocol='m3u8_native', + m3u8_id='hls', fatal=False)) + continue tbr = None mobj = re.search(r'(?P<height>\d+)[pP]?_(?P<tbr>\d+)[kK]', video_url) if mobj: |