on master help sign in
[fix] bing_videos (#1579)

use JSON where possible, compose 'content' using all available data, use correct 'url' (direct to source instead of redirect through bing)

commit204a2cbbf0618d1d8b7d4c52878e43afe4e92444 authorFrank de Lange <github-f@unternet.org> dateSat 27 Jul 2019, 17:49 (7 years ago) parent4b56a4b
2 files changed+41−63
searx/engines/bing_videos.py+28−22 tests/unit/engines/test_bing_videos.py+13−41
searx/engines/bing_videos.py+28−22
@@ -13,7 +13,6 @@
1313from json import loads
1414from lxml import html
1515from searx.engines.bing_images import _fetch_supported_languages, supported_languages_url
16−from searx.engines.xpath import extract_text
1716from searx.url_utils import urlencode
1817from searx.utils import match_language
1918
@@ -22,11 +21,16 @@ categories = ['videos']
2221paging = True
2322safesearch = True
2423time_range_support = True
25−number_of_results = 10
24+number_of_results = 28
2625language_support = True
2726
28−search_url = 'https://www.bing.com/videos/asyncv2?{query}&async=content&'\
29− 'first={offset}&count={number_of_results}&CW=1366&CH=25&FORM=R5VR5'
27+base_url = 'https://www.bing.com/'
28+search_string = 'videos/search'\
29+ '?{query}'\
30+ '&count={count}'\
31+ '&first={first}'\
32+ '&scope=video'\
33+ '&FORM=QBLH'
3034time_range_string = '&qft=+filterui:videoage-lt{interval}'
3135time_range_dict = {'day': '1440',
3236 'week': '10080',
@@ -41,7 +45,12 @@ safesearch_types = {2: 'STRICT',
4145
4246# do search-request
4347def request(query, params):
44− offset = (params['pageno'] - 1) * 10 + 1
48+ offset = ((params['pageno'] - 1) * number_of_results) + 1
49+
50+ search_path = search_string.format(
51+ query=urlencode({'q': query}),
52+ count=number_of_results,
53+ first=offset)
4554
4655 # safesearch cookie
4756 params['cookies']['SRCHHPGUSR'] = \
@@ -52,9 +61,7 @@ def request(query, params):
5261 params['cookies']['_EDGE_S'] = 'mkt=' + language + '&F=1'
5362
5463 # query and paging
55− params['url'] = search_url.format(query=urlencode({'q': query}),
56− offset=offset,
57− number_of_results=number_of_results)
64+ params['url'] = base_url + search_path
5865
5966 # time range
6067 if params['time_range'] in time_range_dict:
@@ -70,19 +77,18 @@ def response(resp):
7077 dom = html.fromstring(resp.text)
7178
7279 for result in dom.xpath('//div[@class="dg_u"]'):
73− url = result.xpath('./div[@class="mc_vtvc"]/a/@href')[0]
74− url = 'https://bing.com' + url
75− title = extract_text(result.xpath('./div/a/div/div[@class="mc_vtvc_title"]/@title'))
76− content = extract_text(result.xpath('./div/a/div/div/div/div/text()'))
77− thumbnail = result.xpath('./div/a/div/div/img/@src')[0]
78−
79− results.append({'url': url,
80− 'title': title,
81− 'content': content,
82− 'thumbnail': thumbnail,
83− 'template': 'videos.html'})
84−
85− if len(results) >= number_of_results:
86− break
80+ try:
81+ metadata = loads(result.xpath('.//div[@class="vrhdata"]/@vrhm')[0])
82+ info = ' - '.join(result.xpath('.//div[@class="mc_vtvc_meta_block"]//span/text()')).strip()
83+ content = '{0} - {1}'.format(metadata['du'], info)
84+ thumbnail = '{0}th?id={1}'.format(base_url, metadata['thid'])
85+ results.append({'url': metadata['murl'],
86+ 'thumbnail': thumbnail,
87+ 'title': metadata.get('vt', ''),
88+ 'content': content,
89+ 'template': 'videos.html'})
90+
91+ except:
92+ continue
8793
8894 return results
tests/unit/engines/test_bing_videos.py+13−41
@@ -29,7 +29,7 @@ class TestBingVideosEngine(SearxTestCase):
2929 dicto['time_range'] = 'day'
3030 dicto['safesearch'] = 2
3131 params = bing_videos.request(query, dicto)
32− self.assertTrue('first=11' in params['url'])
32+ self.assertTrue('first=29' in params['url'])
3333 self.assertTrue('1440' in params['url'])
3434 self.assertIn('SRCHHPGUSR', params['cookies'])
3535 self.assertTrue('STRICT' in params['cookies']['SRCHHPGUSR'])
@@ -48,53 +48,25 @@ class TestBingVideosEngine(SearxTestCase):
4848
4949 html = """
5050 <div class="dg_u">
51− <div id="mc_vtvc_1" class="mc_vtvc">
52− <a class="mc_vtvc_link" href="/video">
53− <div class="mc_vtvc_th">
54− <div class="cico">
55− <img src="thumb_1.jpg" />
56− </div>
57− <div class="mc_vtvc_ban_lo">
58− <div class="vtbc">
59− <div class="mc_bc_w b_smText">
60− <div class="mc_bc pivot bpi_2">
61− <span title="">
62− <span class="mv_vtvc_play cipg "></span>
63− </span>
64− </div>
65− <div class="mc_bc items">10:06</div>
66− </div>
51+ <div>
52+ <a>
53+ <div>
54+ <div>
55+ <div class="mc_vtvc_meta_block">
56+ <div><span>100 views</span><span>1 year ago</span></div><div><span>ExampleTube</span><span>Channel 1<span></div> #noqa
6757 </div>
6858 </div>
69− </div>
70− <div class="mc_vtvc_meta">
71− <div class="mc_vtvc_title" title="Title 1"></div>
72− <div class="mc_vtvc_meta_block_area">
73− <div class="mc_vtvc_meta_block">
74− <div class="mc_vtvc_meta_row">
75− <span>65,696,000+ views</span>
76− <span>1 year ago</span>
77− </div>
78− <div class="mc_vtvc_meta_row mc_vtvc_meta_row_channel">Content 1</div>
79− <div class="mc_vtvc_meta_row"><span>
80− <div class="cico mc_vtvc_src_ico">
81− <div></div>
82− </div>
83− <span>YouTube</span>
84− </span></div>
85− </div>
86− </div>
59+ <div class="vrhdata" vrhm='{"du":"01:11","murl":"https://www.example.com/watch?v=DEADBEEF","thid":"OVP.BINGTHUMB1","vt":"Title 1"}'></div> # noqa
8760 </div>
88− <div class="vrhdata"></div>
89− </a>
90− </div>
61+ </a>
9162 </div>
63+ </div>
9264 """
9365 response = mock.Mock(text=html)
9466 results = bing_videos.response(response)
9567 self.assertEqual(type(results), list)
9668 self.assertEqual(len(results), 1)
9769 self.assertEqual(results[0]['title'], 'Title 1')
98− self.assertEqual(results[0]['url'], 'https://bing.com/video')
99− self.assertEqual(results[0]['content'], 'Content 1')
100− self.assertEqual(results[0]['thumbnail'], 'thumb_1.jpg')
70+ self.assertEqual(results[0]['url'], 'https://www.example.com/watch?v=DEADBEEF')
71+ self.assertEqual(results[0]['content'], '01:11 - 100 views - 1 year ago - ExampleTube - Channel 1')
72+ self.assertEqual(results[0]['thumbnail'], 'https://www.bing.com/th?id=OVP.BINGTHUMB1')