debian/NEWS: Update with notes about split audio/video downloads.

[youtubedl] / youtube_dl / extractor / generic.py

diff --git a/youtube_dl/extractor/generic.py b/youtube_dl/extractor/generic.py
index 7a877b3bcb4e51dfb68a3cbc2720ce52d17c4be9..69e0a7bd271dd0965f5c1f6f9c3a7cdce7a3da0b 100644 (file)
--- a/youtube_dl/extractor/generic.py
+++ b/youtube_dl/extractor/generic.py
@@ -1,3 +1,5 @@
+# encoding: utf-8
+

 import os
 import re
 
 import os
 import re
 


@@ -6,15 +8,41 @@ from ..utils import (
     compat_urllib_error,
     compat_urllib_parse,
     compat_urllib_request,
     compat_urllib_error,
     compat_urllib_parse,
     compat_urllib_request,

+    compat_urlparse,

 
     ExtractorError,
 
     ExtractorError,

+    smuggle_url,
+    unescapeHTML,

 )
 )

+from .brightcove import BrightcoveIE

 
 

-class GenericIE(InfoExtractor):
-    """Generic last-resort information extractor."""

 
 

+class GenericIE(InfoExtractor):
+    IE_DESC = u'Generic downloader that works on some sites'

     _VALID_URL = r'.*'
     IE_NAME = u'generic'
     _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',
+            u'md5': u'85b90ccc9d73b4acd9138d3af4c27f89',
+            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",
+            }
+        }
+    ]

 
     def report_download_webpage(self, video_id):
         """Report webpage download."""
 
     def report_download_webpage(self, video_id):
         """Report webpage download."""


@@ -83,8 +111,18 @@ class GenericIE(InfoExtractor):
         return new_url
 
     def _real_extract(self, url):
         return new_url
 
     def _real_extract(self, url):

-        new_url = self._test_redirect(url)
-        if new_url: return [self.url_result(new_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

 
         video_id = url.split('/')[-1]
         try:
 
         video_id = url.split('/')[-1]
         try:


@@ -92,9 +130,31 @@ class GenericIE(InfoExtractor):
         except ValueError:
             # since this is the last-resort InfoExtractor, if
             # this error is thrown, it'll be thrown here
         except ValueError:
             # since this is the last-resort InfoExtractor, if
             # this error is thrown, it'll be thrown here

-            raise ExtractorError(u'Invalid URL: %s' % url)
+            raise ExtractorError(u'Failed to download URL: %s' % url)

 
         self.report_extraction(video_id)
 
         self.report_extraction(video_id)

+        # 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')
+

         # Start with something easy: JW Player in SWFObject
         mobj = re.search(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
         if mobj is None:
         # Start with something easy: JW Player in SWFObject
         mobj = re.search(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
         if mobj is None:


@@ -102,7 +162,7 @@ class GenericIE(InfoExtractor):
             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|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)
+            mobj = re.search(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http[^\'"&]*)', webpage)

         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:
             # Try to find twitter cards info
             mobj = re.search(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)


@@ -114,15 +174,19 @@ class GenericIE(InfoExtractor):
             if m_video_type is not None:
                 mobj = re.search(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)
         if mobj is None:
             if m_video_type is not None:
                 mobj = re.search(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)
         if mobj is None:

-            raise ExtractorError(u'Invalid URL: %s' % url)
+            # HTML5 video
+            mobj = re.search(r'<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage, flags=re.DOTALL)
+        if mobj is None:
+            raise ExtractorError(u'Unsupported URL: %s' % url)

 
         # It's possible that one of the regexes
         # matched, but returned an empty group:
         if mobj.group(1) is None:
 
         # It's possible that one of the regexes
         # matched, but returned an empty group:
         if mobj.group(1) is None:

-            raise ExtractorError(u'Invalid URL: %s' % url)
+            raise ExtractorError(u'Did not find a valid video URL at %s' % url)

 
 

-        video_url = compat_urllib_parse.unquote(mobj.group(1))
-        video_id = os.path.basename(video_url)
+        video_url = mobj.group(1)
+        video_url = compat_urlparse.urljoin(url, video_url)
+        video_id = compat_urllib_parse.unquote(os.path.basename(video_url))

 
         # here's a fun little line of code for you:
         video_extension = os.path.splitext(video_id)[1][1:]
 
         # here's a fun little line of code for you:
         video_extension = os.path.splitext(video_id)[1][1:]


@@ -135,7 +199,7 @@ class GenericIE(InfoExtractor):
         #   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>',
         #   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')
+            webpage, u'video title', default=u'video', flags=re.DOTALL)

 
         # video uploader is domain name
         video_uploader = self._search_regex(r'(?:https?://)?([^/]*)/.*',
 
         # video uploader is domain name
         video_uploader = self._search_regex(r'(?:https?://)?([^/]*)/.*',