diff --git a/test/tests.json b/test/tests.json index cdf3600447..97654edd92 100644 --- a/test/tests.json +++ b/test/tests.json @@ -70,5 +70,22 @@ "url": "http://www.twitch.tv/thegamedevhub/b/296128360", "file": "296128360.flv", "md5": "ecaa8a790c22a40770901460af191c9a" + }, + { + "name": "MyVideo", + "url": "http://www.myvideo.de/watch/8229274/bowling_fail_or_win", + "file": "8229274.flv", + "md5": "2d2753e8130479ba2cb7e0a37002053e" + }, + { + "name": "Escapist", + "url": "http://www.escapistmagazine.com/videos/view/the-escapist-presents/6618-Breaking-Down-Baldurs-Gate", + "file": "6618-Breaking-Down-Baldurs-Gate.flv", + "md5": "c6793dbda81388f4264c1ba18684a74d" + }, + { + "name": "GooglePlus", + "url": "https://plus.google.com/u/0/108897254135232129896/posts/ZButuJc6CtH", + "file": "ZButuJc6CtH.flv" } ] diff --git a/youtube_dl/InfoExtractors.py b/youtube_dl/InfoExtractors.py index 9766dea54f..b53071c433 100755 --- a/youtube_dl/InfoExtractors.py +++ b/youtube_dl/InfoExtractors.py @@ -799,6 +799,7 @@ def _real_extract(self, url): class YahooIE(InfoExtractor): """Information extractor for video.yahoo.com.""" + _WORKING = False # _VALID_URL matches all Yahoo! Video URLs # _VPAGE_URL matches only the extractable '/watch/' URLs _VALID_URL = r'(?:http://)?(?:[a-z]+\.)?video\.yahoo\.com/(?:watch|network)/([0-9]+)(?:/|\?v=)([0-9]+)(?:[#\?].*)?' @@ -1170,7 +1171,7 @@ def extractPlus7Stream(self, url): 'url': compat_urllib_parse.unquote(info.get('url')), 'uploader': u'arte.tv', 'upload_date': info.get('date'), - 'title': info.get('title'), + 'title': info.get('title').decode('utf-8'), 'ext': u'mp4', 'format': u'NA', 'player_url': None, @@ -1495,6 +1496,8 @@ def _download_n_results(self, query, n): class YahooSearchIE(InfoExtractor): """Information Extractor for Yahoo! Video search queries.""" + + _WORKING = False _VALID_URL = r'yvsearch(\d+|all)?:[\s\S]+' _TEMPLATE_URL = 'http://video.yahoo.com/search/?p=%s&o=%s' _VIDEO_INDICATOR = r'href="http://video\.yahoo\.com/watch/([0-9]+/[0-9]+)"' @@ -2274,7 +2277,7 @@ def _real_extract(self,url): request = compat_urllib_request.Request('http://www.myvideo.de/watch/%s' % video_id) try: self.report_download_webpage(video_id) - webpage = compat_urllib_request.urlopen(request).read() + webpage = compat_urllib_request.urlopen(request).read().decode('utf-8') except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: self._downloader.trouble(u'ERROR: Unable to retrieve video webpage: %s' % compat_str(err)) return @@ -2556,7 +2559,9 @@ def _real_extract(self, url): self.report_config_download(showName) try: - configJSON = compat_urllib_request.urlopen(configUrl).read() + configJSON = compat_urllib_request.urlopen(configUrl) + m = re.match(r'text/html; charset="?([^"]+)"?', configJSON.headers['Content-Type']) + configJSON = configJSON.read().decode(m.group(1) if m else 'utf-8') except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: self._downloader.trouble(u'ERROR: unable to download configuration: ' + compat_str(err)) return @@ -2876,6 +2881,8 @@ def _real_extract(self, url): class MixcloudIE(InfoExtractor): """Information extractor for www.mixcloud.com""" + + _WORKING = False # New API, but it seems good http://www.mixcloud.com/developers/documentation/ _VALID_URL = r'^(?:https?://)?(?:www\.)?mixcloud\.com/([\w\d-]+)/([\w\d-]+)' IE_NAME = u'mixcloud' @@ -3378,7 +3385,7 @@ def _real_extract(self, url): class GooglePlusIE(InfoExtractor): """Information extractor for plus.google.com.""" - _VALID_URL = r'(?:https://)?plus\.google\.com/(?:\w+/)*?(\d+)/posts/(\w+)' + _VALID_URL = r'(?:https://)?plus\.google\.com/(?:[^/]+/)*?posts/(\w+)' IE_NAME = u'plus.google' def __init__(self, downloader=None): @@ -3386,7 +3393,7 @@ def __init__(self, downloader=None): def report_extract_entry(self, url): """Report downloading extry""" - self._downloader.to_screen(u'[plus.google] Downloading entry: %s' % url.decode('utf-8')) + self._downloader.to_screen(u'[plus.google] Downloading entry: %s' % url) def report_date(self, upload_date): """Report downloading extry""" @@ -3394,15 +3401,15 @@ def report_date(self, upload_date): def report_uploader(self, uploader): """Report downloading extry""" - self._downloader.to_screen(u'[plus.google] Uploader: %s' % uploader.decode('utf-8')) + self._downloader.to_screen(u'[plus.google] Uploader: %s' % uploader) def report_title(self, video_title): """Report downloading extry""" - self._downloader.to_screen(u'[plus.google] Title: %s' % video_title.decode('utf-8')) + self._downloader.to_screen(u'[plus.google] Title: %s' % video_title) def report_extract_vid_page(self, video_page): """Report information extraction.""" - self._downloader.to_screen(u'[plus.google] Extracting video page: %s' % video_page.decode('utf-8')) + self._downloader.to_screen(u'[plus.google] Extracting video page: %s' % video_page) def _real_extract(self, url): # Extract id from URL @@ -3412,7 +3419,7 @@ def _real_extract(self, url): return post_url = mobj.group(0) - video_id = mobj.group(2) + video_id = mobj.group(1) video_extension = 'flv' @@ -3420,7 +3427,7 @@ def _real_extract(self, url): self.report_extract_entry(post_url) request = compat_urllib_request.Request(post_url) try: - webpage = compat_urllib_request.urlopen(request).read() + webpage = compat_urllib_request.urlopen(request).read().decode('utf-8') except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: self._downloader.trouble(u'ERROR: Unable to retrieve entry webpage: %s' % compat_str(err)) return @@ -3462,7 +3469,7 @@ def _real_extract(self, url): video_page = mobj.group(1) request = compat_urllib_request.Request(video_page) try: - webpage = compat_urllib_request.urlopen(request).read() + webpage = compat_urllib_request.urlopen(request).read().decode('utf-8') except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: self._downloader.trouble(u'ERROR: Unable to retrieve video webpage: %s' % compat_str(err)) return @@ -3484,16 +3491,19 @@ def _real_extract(self, url): # Only get the url. The resolution part in the tuple has no use anymore video_url = video_url[-1] # Treat escaped \u0026 style hex - video_url = unicode(video_url, "unicode_escape") + try: + video_url = video_url.decode("unicode_escape") + except AttributeError: # Python 3 + video_url = bytes(video_url, 'ascii').decode('unicode-escape') return [{ - 'id': video_id.decode('utf-8'), + 'id': video_id, 'url': video_url, - 'uploader': uploader.decode('utf-8'), - 'upload_date': upload_date.decode('utf-8'), - 'title': video_title.decode('utf-8'), - 'ext': video_extension.decode('utf-8'), + 'uploader': uploader, + 'upload_date': upload_date, + 'title': video_title, + 'ext': video_extension, }] class NBAIE(InfoExtractor):