[steam] fix extraction(fixes #14067)

youtube_dl/extractor/steam.py

@@ -4,8 +4,10 @@
 
 from .common import InfoExtractor
 from ..utils import (
+    extract_attributes,
     ExtractorError,
-    unescapeHTML,
+    get_element_by_class,
+    js_to_json,
 )
 
 
@@ -25,35 +27,39 @@ class SteamIE(InfoExtractor):
         'url': 'http://store.steampowered.com/video/105600/',
         'playlist': [
             {
-                'md5': 'f870007cee7065d7c76b88f0a45ecc07',
+                'md5': '6a294ee0c4b1f47f5bb76a65e31e3592',
                 'info_dict': {
-                    'id': '81300',
+                    'id': '2040428',
-                    'ext': 'flv',
+                    'ext': 'mp4',
-                    'title': 'Terraria 1.1 Trailer',
+                    'title': 'Terraria 1.3 Trailer',
                     'playlist_index': 1,
                 }
             },
             {
-                'md5': '61aaf31a5c5c3041afb58fb83cbb5751',
+                'md5': '911672b20064ca3263fa89650ba5a7aa',
                 'info_dict': {
-                    'id': '80859',
+                    'id': '2029566',
-                    'ext': 'flv',
+                    'ext': 'mp4',
-                    'title': 'Terraria Trailer',
+                    'title': 'Terraria 1.2 Trailer',
                     'playlist_index': 2,
                 }
             }
         ],
+        'info_dict': {
+            'id': '105600',
+            'title': 'Terraria',
+        },
         'params': {
             'playlistend': 2,
         }
     }, {
         'url': 'http://steamcommunity.com/sharedfiles/filedetails/?id=242472205',
         'info_dict': {
-            'id': 'WB5DvDOOvAY',
+            'id': 'X8kpJBlzD2E',
             'ext': 'mp4',
-            'upload_date': '20140329',
+            'upload_date': '20140617',
-            'title': 'FRONTIERS - Final Greenlight Trailer',
+            'title': 'FRONTIERS - Trapping',
-            'description': 'md5:dc96a773669d0ca1b36c13c1f30250d9',
+            'description': 'md5:bf6f7f773def614054089e5769c12a6e',
             'uploader': 'AAD Productions',
             'uploader_id': 'AtomicAgeDogGames',
         }
@@ -76,48 +82,65 @@ def _real_extract(self, url):
             self.report_age_confirmation()
             webpage = self._download_webpage(videourl, playlist_id)
 
+        flash_vars = self._parse_json(self._search_regex(
+            r'(?s)rgMovieFlashvars\s*=\s*({.+?});', webpage,
+            'flash vars'), playlist_id, js_to_json)
+
+        playlist_title = None
+        entries = []
         if fileID:
-            playlist_title = self._html_search_regex(
+            playlist_title = get_element_by_class('workshopItemTitle', webpage)
-                r'<div class="workshopItemTitle">(.+)</div>', webpage, 'title')
+            for movie in flash_vars.values():
-            mweb = re.finditer(r'''(?x)
+                if not movie:
-                'movie_(?P<videoID>[0-9]+)':\s*\{\s*
+                    continue
-                YOUTUBE_VIDEO_ID:\s*"(?P<youtube_id>[^"]+)",
+                youtube_id = movie.get('YOUTUBE_VIDEO_ID')
-                ''', webpage)
+                if not youtube_id:
-            videos = [{
+                    continue
-                '_type': 'url',
+                entries.append({
-                'url': vid.group('youtube_id'),
+                    '_type': 'url',
-                'ie_key': 'Youtube',
+                    'url': youtube_id,
-            } for vid in mweb]
+                    'ie_key': 'Youtube',
-        else:
-            playlist_title = self._html_search_regex(
-                r'<h2 class="pageheader">(.*?)</h2>', webpage, 'game title')
-
-            mweb = re.finditer(r'''(?x)
-                'movie_(?P<videoID>[0-9]+)':\s*\{\s*
-                FILENAME:\s*"(?P<videoURL>[\w:/\.\?=]+)"
-                (,\s*MOVIE_NAME:\s*\"(?P<videoName>[\w:/\.\?=\+-]+)\")?\s*\},
-                ''', webpage)
-            titles = re.finditer(
-                r'<span class="title">(?P<videoName>.+?)</span>', webpage)
-            thumbs = re.finditer(
-                r'<img class="movie_thumb" src="(?P<thumbnail>.+?)">', webpage)
-            videos = []
-
-            for vid, vtitle, thumb in zip(mweb, titles, thumbs):
-                video_id = vid.group('videoID')
-                title = vtitle.group('videoName')
-                video_url = vid.group('videoURL')
-                video_thumb = thumb.group('thumbnail')
-                if not video_url:
-                    raise ExtractorError('Cannot find video url for %s' % video_id)
-                videos.append({
-                    'id': video_id,
-                    'url': video_url,
-                    'ext': 'flv',
-                    'title': unescapeHTML(title),
-                    'thumbnail': video_thumb
                 })
-        if not videos:
+        else:
+            playlist_title = get_element_by_class('apphub_AppName', webpage)
+            for movie_id, movie in flash_vars.items():
+                if not movie:
+                    continue
+                video_id = self._search_regex(r'movie_(\d+)', movie_id, 'video id', fatal=False)
+                title = movie.get('MOVIE_NAME')
+                if not title or not video_id:
+                    continue
+                entry = {
+                    'id': video_id,
+                    'title': title.replace('+', ' '),
+                }
+                formats = []
+                flv_url = movie.get('FILENAME')
+                if flv_url:
+                    formats.append({
+                        'format_id': 'flv',
+                        'url': flv_url,
+                    })
+                highlight_element = self._search_regex(
+                    r'(<div[^>]+id="highlight_movie_%s"[^>]+>)' % video_id,
+                    webpage, 'highlight element', fatal=False)
+                if highlight_element:
+                    highlight_attribs = extract_attributes(highlight_element)
+                    if highlight_attribs:
+                        entry['thumbnail'] = highlight_attribs.get('data-poster')
+                        for quality in ('', '-hd'):
+                            for ext in ('webm', 'mp4'):
+                                video_url = highlight_attribs.get('data-%s%s-source' % (ext, quality))
+                                if video_url:
+                                    formats.append({
+                                        'format_id': ext + quality,
+                                        'url': video_url,
+                                    })
+                if not formats:
+                    continue
+                entry['formats'] = formats
+                entries.append(entry)
+        if not entries:
             raise ExtractorError('Could not find any videos')
 
-        return self.playlist_result(videos, playlist_id, playlist_title)
+        return self.playlist_result(entries, playlist_id, playlist_title)