diff options
Diffstat (limited to 'youtube_dl/extractor/googlesearch.py')
| -rw-r--r-- | youtube_dl/extractor/googlesearch.py | 38 |
1 files changed, 25 insertions, 13 deletions
diff --git a/youtube_dl/extractor/googlesearch.py b/youtube_dl/extractor/googlesearch.py index f9c88e9b5..5c2564270 100644 --- a/youtube_dl/extractor/googlesearch.py +++ b/youtube_dl/extractor/googlesearch.py @@ -1,3 +1,5 @@ +from __future__ import unicode_literals + import itertools import re @@ -8,32 +10,42 @@ from ..utils import ( class GoogleSearchIE(SearchInfoExtractor): - IE_DESC = u'Google Video search' - _MORE_PAGES_INDICATOR = r'id="pnnext" class="pn"' + IE_DESC = 'Google Video search' _MAX_RESULTS = 1000 - IE_NAME = u'video.google:search' + IE_NAME = 'video.google:search' _SEARCH_KEY = 'gvsearch' def _get_n_results(self, query, n): """Get a specified number of results for a query""" + entries = [] res = { '_type': 'playlist', 'id': query, - 'entries': [] + 'title': query, } - for pagenum in itertools.count(1): - result_url = u'http://www.google.com/search?tbm=vid&q=%s&start=%s&hl=en' % (compat_urllib_parse.quote_plus(query), pagenum*10) - webpage = self._download_webpage(result_url, u'gvsearch:' + query, - note='Downloading result page ' + str(pagenum)) + for pagenum in itertools.count(): + result_url = ( + 'http://www.google.com/search?tbm=vid&q=%s&start=%s&hl=en' + % (compat_urllib_parse.quote_plus(query), pagenum * 10)) + + webpage = self._download_webpage( + result_url, 'gvsearch:' + query, + note='Downloading result page ' + str(pagenum + 1)) + + for hit_idx, mobj in enumerate(re.finditer( + r'<h3 class="r"><a href="([^"]+)"', webpage)): + + # Skip playlists + if not re.search(r'id="vidthumb%d"' % (hit_idx + 1), webpage): + continue - for mobj in re.finditer(r'<h3 class="r"><a href="([^"]+)"', webpage): - e = { + entries.append({ '_type': 'url', 'url': mobj.group(1) - } - res['entries'].append(e) + }) - if (pagenum * 10 > n) or not re.search(self._MORE_PAGES_INDICATOR, webpage): + if (len(entries) >= n) or not re.search(r'class="pn" id="pnnext"', webpage): + res['entries'] = entries[:n] return res |