|
|
- from __future__ import unicode_literals
-
- import itertools
- import re
-
- from .common import SearchInfoExtractor
- from ..compat import (
- compat_urllib_parse,
- )
-
-
- class GoogleSearchIE(SearchInfoExtractor):
- IE_DESC = 'Google Video search'
- _MAX_RESULTS = 1000
- IE_NAME = 'video.google:search'
- _SEARCH_KEY = 'gvsearch'
- _TEST = {
- 'url': 'gvsearch15:python language',
- 'info_dict': {
- 'id': 'python language',
- 'title': 'python language',
- },
- 'playlist_count': 15,
- }
-
- def _get_n_results(self, query, n):
- """Get a specified number of results for a query"""
-
- entries = []
- res = {
- '_type': 'playlist',
- 'id': query,
- 'title': query,
- }
-
- 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
-
- entries.append({
- '_type': 'url',
- 'url': mobj.group(1)
- })
-
- if (len(entries) >= n) or not re.search(r'id="pnnext"', webpage):
- res['entries'] = entries[:n]
- return res
|