Merge branch 'master' of github.com:rg3/youtube-dl
[youtube-dl] / youtube_dl / extractor / youtube.py
index 3a3a5a39e4bfcdce1a3bee26fee3e43326d62d90..ac2d5db4e1b9d47d73ea1ed630a8616bbe77f110 100644 (file)
@@ -7,13 +7,13 @@ import itertools
 import json
 import os.path
 import re
 import json
 import os.path
 import re
-import string
 import struct
 import traceback
 import zlib
 
 from .common import InfoExtractor, SearchInfoExtractor
 from .subtitles import SubtitlesInfoExtractor
 import struct
 import traceback
 import zlib
 
 from .common import InfoExtractor, SearchInfoExtractor
 from .subtitles import SubtitlesInfoExtractor
+from ..jsinterp import JSInterpreter
 from ..utils import (
     compat_chr,
     compat_parse_qs,
 from ..utils import (
     compat_chr,
     compat_parse_qs,
@@ -151,6 +151,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
                              )
                          ))
                          |youtu\.be/                                          # just youtu.be/xxxx
                              )
                          ))
                          |youtu\.be/                                          # just youtu.be/xxxx
+                         |https?://(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId=
                          )
                      )?                                                       # all until now is optional -> you can pass the naked ID
                      ([0-9A-Za-z_-]{11})                                      # here is it! the YouTube video ID
                          )
                      )?                                                       # all until now is optional -> you can pass the naked ID
                      ([0-9A-Za-z_-]{11})                                      # here is it! the YouTube video ID
@@ -438,113 +439,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
     def _parse_sig_js(self, jscode):
         funcname = self._search_regex(
             r'signature=([a-zA-Z]+)', jscode,
     def _parse_sig_js(self, jscode):
         funcname = self._search_regex(
             r'signature=([a-zA-Z]+)', jscode,
-            u'Initial JS player signature function name')
-
-        functions = {}
-
-        def argidx(varname):
-            return string.lowercase.index(varname)
-
-        def interpret_statement(stmt, local_vars, allow_recursion=20):
-            if allow_recursion < 0:
-                raise ExtractorError(u'Recursion limit reached')
-
-            if stmt.startswith(u'var '):
-                stmt = stmt[len(u'var '):]
-            ass_m = re.match(r'^(?P<out>[a-z]+)(?:\[(?P<index>[^\]]+)\])?' +
-                             r'=(?P<expr>.*)$', stmt)
-            if ass_m:
-                if ass_m.groupdict().get('index'):
-                    def assign(val):
-                        lvar = local_vars[ass_m.group('out')]
-                        idx = interpret_expression(ass_m.group('index'),
-                                                   local_vars, allow_recursion)
-                        assert isinstance(idx, int)
-                        lvar[idx] = val
-                        return val
-                    expr = ass_m.group('expr')
-                else:
-                    def assign(val):
-                        local_vars[ass_m.group('out')] = val
-                        return val
-                    expr = ass_m.group('expr')
-            elif stmt.startswith(u'return '):
-                assign = lambda v: v
-                expr = stmt[len(u'return '):]
-            else:
-                raise ExtractorError(
-                    u'Cannot determine left side of statement in %r' % stmt)
-
-            v = interpret_expression(expr, local_vars, allow_recursion)
-            return assign(v)
-
-        def interpret_expression(expr, local_vars, allow_recursion):
-            if expr.isdigit():
-                return int(expr)
-
-            if expr.isalpha():
-                return local_vars[expr]
-
-            m = re.match(r'^(?P<in>[a-z]+)\.(?P<member>.*)$', expr)
-            if m:
-                member = m.group('member')
-                val = local_vars[m.group('in')]
-                if member == 'split("")':
-                    return list(val)
-                if member == 'join("")':
-                    return u''.join(val)
-                if member == 'length':
-                    return len(val)
-                if member == 'reverse()':
-                    return val[::-1]
-                slice_m = re.match(r'slice\((?P<idx>.*)\)', member)
-                if slice_m:
-                    idx = interpret_expression(
-                        slice_m.group('idx'), local_vars, allow_recursion-1)
-                    return val[idx:]
-
-            m = re.match(
-                r'^(?P<in>[a-z]+)\[(?P<idx>.+)\]$', expr)
-            if m:
-                val = local_vars[m.group('in')]
-                idx = interpret_expression(m.group('idx'), local_vars,
-                                           allow_recursion-1)
-                return val[idx]
-
-            m = re.match(r'^(?P<a>.+?)(?P<op>[%])(?P<b>.+?)$', expr)
-            if m:
-                a = interpret_expression(m.group('a'),
-                                         local_vars, allow_recursion)
-                b = interpret_expression(m.group('b'),
-                                         local_vars, allow_recursion)
-                return a % b
-
-            m = re.match(
-                r'^(?P<func>[a-zA-Z$]+)\((?P<args>[a-z0-9,]+)\)$', expr)
-            if m:
-                fname = m.group('func')
-                if fname not in functions:
-                    functions[fname] = extract_function(fname)
-                argvals = [int(v) if v.isdigit() else local_vars[v]
-                           for v in m.group('args').split(',')]
-                return functions[fname](argvals)
-            raise ExtractorError(u'Unsupported JS expression %r' % expr)
-
-        def extract_function(funcname):
-            func_m = re.search(
-                r'function ' + re.escape(funcname) +
-                r'\((?P<args>[a-z,]+)\){(?P<code>[^}]+)}',
-                jscode)
-            argnames = func_m.group('args').split(',')
-
-            def resf(args):
-                local_vars = dict(zip(argnames, args))
-                for stmt in func_m.group('code').split(';'):
-                    res = interpret_statement(stmt, local_vars)
-                return res
-            return resf
-
-        initial_function = extract_function(funcname)
+             u'Initial JS player signature function name')
+
+        jsi = JSInterpreter(jscode)
+        initial_function = jsi.extract_function(funcname)
         return lambda s: initial_function([s])
 
     def _parse_sig_swf(self, file_contents):
         return lambda s: initial_function([s])
 
     def _parse_sig_swf(self, file_contents):
@@ -1549,12 +1447,15 @@ class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
                 break
 
             more = self._download_json(
                 break
 
             more = self._download_json(
-                'https://youtube.com/%s' % mobj.group('more'), playlist_id, 'Downloading page #%s' % page_num)
+                'https://youtube.com/%s' % mobj.group('more'), playlist_id,
+                'Downloading page #%s' % page_num,
+                transform_source=uppercase_escape)
             content_html = more['content_html']
             more_widget_html = more['load_more_widget_html']
 
         playlist_title = self._html_search_regex(
             content_html = more['content_html']
             more_widget_html = more['load_more_widget_html']
 
         playlist_title = self._html_search_regex(
-                r'<h1 class="pl-header-title">\s*(.*?)\s*</h1>', page, u'title')
+            r'(?s)<h1 class="pl-header-title[^"]*">\s*(.*?)\s*</h1>',
+            page, u'title')
 
         url_results = self._ids_to_results(ids)
         return self.playlist_result(url_results, playlist_id, playlist_title)
 
         url_results = self._ids_to_results(ids)
         return self.playlist_result(url_results, playlist_id, playlist_title)
@@ -1712,7 +1613,7 @@ class YoutubeUserIE(InfoExtractor):
 
 class YoutubeSearchIE(SearchInfoExtractor):
     IE_DESC = u'YouTube.com searches'
 
 class YoutubeSearchIE(SearchInfoExtractor):
     IE_DESC = u'YouTube.com searches'
-    _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
+    _API_URL = u'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
     _MAX_RESULTS = 1000
     IE_NAME = u'youtube:search'
     _SEARCH_KEY = 'ytsearch'
     _MAX_RESULTS = 1000
     IE_NAME = u'youtube:search'
     _SEARCH_KEY = 'ytsearch'
@@ -1723,9 +1624,12 @@ class YoutubeSearchIE(SearchInfoExtractor):
         video_ids = []
         pagenum = 0
         limit = n
         video_ids = []
         pagenum = 0
         limit = n
+        PAGE_SIZE = 50
 
 
-        while (50 * pagenum) < limit:
-            result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1)
+        while (PAGE_SIZE * pagenum) < limit:
+            result_url = self._API_URL % (
+                compat_urllib_parse.quote_plus(query.encode('utf-8')),
+                (PAGE_SIZE * pagenum) + 1)
             data_json = self._download_webpage(
                 result_url, video_id=u'query "%s"' % query,
                 note=u'Downloading page %s' % (pagenum + 1),
             data_json = self._download_webpage(
                 result_url, video_id=u'query "%s"' % query,
                 note=u'Downloading page %s' % (pagenum + 1),
@@ -1836,11 +1740,10 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
         feed_entries = []
         paging = 0
         for i in itertools.count(1):
         feed_entries = []
         paging = 0
         for i in itertools.count(1):
-            info = self._download_webpage(self._FEED_TEMPLATE % paging,
+            info = self._download_json(self._FEED_TEMPLATE % paging,
                                           u'%s feed' % self._FEED_NAME,
                                           u'Downloading page %s' % i)
                                           u'%s feed' % self._FEED_NAME,
                                           u'Downloading page %s' % i)
-            info = json.loads(info)
-            feed_html = info['feed_html']
+            feed_html = info.get('feed_html') or info.get('content_html')
             m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
             ids = orderedSet(m.group(1) for m in m_ids)
             feed_entries.extend(
             m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
             ids = orderedSet(m.group(1) for m in m_ids)
             feed_entries.extend(
@@ -1852,7 +1755,7 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
         return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
 
 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
         return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
 
 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
-    IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)'
+    IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)'
     _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
     _FEED_NAME = 'subscriptions'
     _PLAYLIST_TITLE = u'Youtube Subscriptions'
     _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
     _FEED_NAME = 'subscriptions'
     _PLAYLIST_TITLE = u'Youtube Subscriptions'