youtube-dl/youtube_dl/extractor/generic.py

232 lines
9.6 KiB
Python
Raw Normal View History

# encoding: utf-8
2013-06-23 20:31:45 +02:00
import os
import re
from .common import InfoExtractor
from ..utils import (
compat_urllib_error,
compat_urllib_parse,
compat_urllib_request,
2013-08-28 12:47:27 +02:00
compat_urlparse,
2013-06-23 20:31:45 +02:00
ExtractorError,
smuggle_url,
unescapeHTML,
2013-06-23 20:31:45 +02:00
)
from .brightcove import BrightcoveIE
2013-06-23 20:31:45 +02:00
2013-06-23 20:31:45 +02:00
class GenericIE(InfoExtractor):
IE_DESC = u'Generic downloader that works on some sites'
2013-06-23 20:31:45 +02:00
_VALID_URL = r'.*'
IE_NAME = u'generic'
_TESTS = [
{
u'url': u'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
u'file': u'13601338388002.mp4',
2013-10-27 14:21:37 +01:00
u'md5': u'6e15c93721d7ec9e9ca3fdbf07982cfd',
u'info_dict': {
u"uploader": u"www.hodiho.fr",
u"title": u"R\u00e9gis plante sa Jeep"
}
},
# embedded vimeo video
{
u'url': u'http://skillsmatter.com/podcast/home/move-semanticsperfect-forwarding-and-rvalue-references',
u'file': u'22444065.mp4',
u'md5': u'2903896e23df39722c33f015af0666e2',
u'info_dict': {
u'title': u'ACCU 2011: Move Semantics,Perfect Forwarding, and Rvalue references- Scott Meyers- 13/04/2011',
u"uploader_id": u"skillsmatter",
u"uploader": u"Skills Matter",
}
},
# bandcamp page with custom domain
{
u'url': u'http://bronyrock.com/track/the-pony-mash',
u'file': u'3235767654.mp3',
u'info_dict': {
u'title': u'The Pony Mash',
u'uploader': u'M_Pallante',
},
u'skip': u'There is a limit of 200 free downloads / month for the test song',
},
]
2013-06-23 20:31:45 +02:00
def report_download_webpage(self, video_id):
"""Report webpage download."""
if not self._downloader.params.get('test', False):
self._downloader.report_warning(u'Falling back on generic information extractor.')
super(GenericIE, self).report_download_webpage(video_id)
def report_following_redirect(self, new_url):
"""Report information extraction."""
self._downloader.to_screen(u'[redirect] Following redirect to %s' % new_url)
def _test_redirect(self, url):
"""Check if it is a redirect, like url shorteners, in case return the new url."""
class HeadRequest(compat_urllib_request.Request):
def get_method(self):
return "HEAD"
class HEADRedirectHandler(compat_urllib_request.HTTPRedirectHandler):
"""
Subclass the HTTPRedirectHandler to make it use our
HeadRequest also on the redirected URL
"""
def redirect_request(self, req, fp, code, msg, headers, newurl):
if code in (301, 302, 303, 307):
newurl = newurl.replace(' ', '%20')
newheaders = dict((k,v) for k,v in req.headers.items()
if k.lower() not in ("content-length", "content-type"))
return HeadRequest(newurl,
headers=newheaders,
origin_req_host=req.get_origin_req_host(),
unverifiable=True)
else:
raise compat_urllib_error.HTTPError(req.get_full_url(), code, msg, headers, fp)
class HTTPMethodFallback(compat_urllib_request.BaseHandler):
"""
Fallback to GET if HEAD is not allowed (405 HTTP error)
"""
def http_error_405(self, req, fp, code, msg, headers):
fp.read()
fp.close()
newheaders = dict((k,v) for k,v in req.headers.items()
if k.lower() not in ("content-length", "content-type"))
return self.parent.open(compat_urllib_request.Request(req.get_full_url(),
headers=newheaders,
origin_req_host=req.get_origin_req_host(),
unverifiable=True))
# Build our opener
opener = compat_urllib_request.OpenerDirector()
for handler in [compat_urllib_request.HTTPHandler, compat_urllib_request.HTTPDefaultErrorHandler,
HTTPMethodFallback, HEADRedirectHandler,
compat_urllib_request.HTTPErrorProcessor, compat_urllib_request.HTTPSHandler]:
opener.add_handler(handler())
response = opener.open(HeadRequest(url))
if response is None:
raise ExtractorError(u'Invalid URL protocol')
new_url = response.geturl()
if url == new_url:
return False
self.report_following_redirect(new_url)
return new_url
def _real_extract(self, url):
parsed_url = compat_urlparse.urlparse(url)
if not parsed_url.scheme:
self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
return self.url_result('http://' + url)
try:
new_url = self._test_redirect(url)
if new_url:
return [self.url_result(new_url)]
except compat_urllib_error.HTTPError:
# This may be a stupid server that doesn't like HEAD, our UA, or so
pass
2013-06-23 20:31:45 +02:00
video_id = url.split('/')[-1]
try:
webpage = self._download_webpage(url, video_id)
except ValueError:
# since this is the last-resort InfoExtractor, if
# this error is thrown, it'll be thrown here
2013-10-06 16:03:18 +02:00
raise ExtractorError(u'Failed to download URL: %s' % url)
2013-06-23 20:31:45 +02:00
self.report_extraction(video_id)
2013-08-26 21:29:31 +02:00
# Look for BrightCove:
m_brightcove = re.search(r'<object[^>]+?class=([\'"])[^>]*?BrightcoveExperience.*?\1.+?</object>', webpage, re.DOTALL)
if m_brightcove is not None:
self.to_screen(u'Brightcove video detected.')
bc_url = BrightcoveIE._build_brighcove_url(m_brightcove.group())
return self.url_result(bc_url, 'Brightcove')
# Look for embedded Vimeo player
mobj = re.search(
r'<iframe[^>]+?src="(https?://player.vimeo.com/video/.+?)"', webpage)
if mobj:
player_url = unescapeHTML(mobj.group(1))
surl = smuggle_url(player_url, {'Referer': url})
return self.url_result(surl, 'Vimeo')
# Look for embedded YouTube player
mobj = re.search(
r'<iframe[^>]+?src="(https?://(?:www\.)?youtube.com/embed/.+?)"', webpage)
if mobj:
surl = unescapeHTML(mobj.group(1))
return self.url_result(surl, 'Youtube')
# Look for Bandcamp pages with custom domain
mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
if mobj is not None:
burl = unescapeHTML(mobj.group(1))
return self.url_result(burl, 'Bandcamp')
2013-06-23 20:31:45 +02:00
# Start with something easy: JW Player in SWFObject
mobj = re.search(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
if mobj is None:
# Broaden the search a little bit
mobj = re.search(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
if mobj is None:
# Broaden the search a little bit: JWPlayer JS loader
mobj = re.search(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http[^\'"&]*)', webpage)
2013-06-23 20:31:45 +02:00
if mobj is None:
# Try to find twitter cards info
mobj = re.search(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
if mobj is None:
# We look for Open Graph info:
# We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
m_video_type = re.search(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
# We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
if m_video_type is not None:
mobj = re.search(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)
2013-08-21 04:32:22 +02:00
if mobj is None:
# HTML5 video
mobj = re.search(r'<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage, flags=re.DOTALL)
2013-06-23 20:31:45 +02:00
if mobj is None:
2013-10-06 16:03:18 +02:00
raise ExtractorError(u'Unsupported URL: %s' % url)
2013-06-23 20:31:45 +02:00
# It's possible that one of the regexes
# matched, but returned an empty group:
if mobj.group(1) is None:
2013-10-06 16:03:18 +02:00
raise ExtractorError(u'Did not find a valid video URL at %s' % url)
2013-06-23 20:31:45 +02:00
video_url = mobj.group(1)
2013-08-28 12:47:27 +02:00
video_url = compat_urlparse.urljoin(url, video_url)
video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
2013-06-23 20:31:45 +02:00
# here's a fun little line of code for you:
video_extension = os.path.splitext(video_id)[1][1:]
video_id = os.path.splitext(video_id)[0]
# it's tempting to parse this further, but you would
# have to take into account all the variations like
# Video Title - Site Name
# Site Name | Video Title
# Video Title - Tagline | Site Name
# and so on and so forth; it's just not practical
video_title = self._html_search_regex(r'<title>(.*)</title>',
webpage, u'video title', default=u'video', flags=re.DOTALL)
2013-06-23 20:31:45 +02:00
# video uploader is domain name
video_uploader = self._search_regex(r'(?:https?://)?([^/]*)/.*',
url, u'video uploader')
return [{
'id': video_id,
'url': video_url,
'uploader': video_uploader,
'upload_date': None,
'title': video_title,
'ext': video_extension,
}]