|
@@ -0,0 +1,81 @@
|
|
1
|
+## Bing (Images)
|
|
2
|
+#
|
|
3
|
+# @website https://www.bing.com/images
|
|
4
|
+# @provide-api yes (http://datamarket.azure.com/dataset/bing/search), max. 5000 query/month
|
|
5
|
+#
|
|
6
|
+# @using-api no (because of query limit)
|
|
7
|
+# @results HTML (using search portal)
|
|
8
|
+# @stable no (HTML can change)
|
|
9
|
+# @parse url, title, img_src
|
|
10
|
+#
|
|
11
|
+# @todo currently there are up to 35 images receive per page, because bing does not parse count=10. limited response to 10 images
|
|
12
|
+
|
|
13
|
+from urllib import urlencode
|
|
14
|
+from cgi import escape
|
|
15
|
+from lxml import html
|
|
16
|
+from yaml import load
|
|
17
|
+import re
|
|
18
|
+
|
|
19
|
+# engine dependent config
|
|
20
|
+categories = ['images']
|
|
21
|
+paging = True
|
|
22
|
+
|
|
23
|
+# search-url
|
|
24
|
+base_url = 'https://www.bing.com/'
|
|
25
|
+search_string = 'images/search?{query}&count=10&first={offset}'
|
|
26
|
+
|
|
27
|
+# do search-request
|
|
28
|
+def request(query, params):
|
|
29
|
+ offset = (params['pageno'] - 1) * 10 + 1
|
|
30
|
+
|
|
31
|
+ # required for cookie
|
|
32
|
+ language = 'en-US'
|
|
33
|
+
|
|
34
|
+ search_path = search_string.format(
|
|
35
|
+ query=urlencode({'q': query}),
|
|
36
|
+ offset=offset)
|
|
37
|
+
|
|
38
|
+ params['cookies']['SRCHHPGUSR'] = \
|
|
39
|
+ 'NEWWND=0&NRSLT=-1&SRCHLANG=' + language.split('-')[0]
|
|
40
|
+
|
|
41
|
+ params['url'] = base_url + search_path
|
|
42
|
+
|
|
43
|
+ print(params['url'])
|
|
44
|
+
|
|
45
|
+ return params
|
|
46
|
+
|
|
47
|
+
|
|
48
|
+# get response from search-request
|
|
49
|
+def response(resp):
|
|
50
|
+ results = []
|
|
51
|
+
|
|
52
|
+ dom = html.fromstring(resp.content)
|
|
53
|
+
|
|
54
|
+ # init regex for yaml-parsing
|
|
55
|
+ p = re.compile( '({|,)([a-z]+):(")')
|
|
56
|
+
|
|
57
|
+ # parse results
|
|
58
|
+ for result in dom.xpath('//div[@class="dg_u"]'):
|
|
59
|
+ link = result.xpath('./a')[0]
|
|
60
|
+
|
|
61
|
+ # parse yaml-data (it is required to add a space, to make it parsable)
|
|
62
|
+ yaml_data = load(p.sub( r'\1\2: \3', link.attrib.get('m')))
|
|
63
|
+
|
|
64
|
+ title = link.attrib.get('t1')
|
|
65
|
+ #url = 'http://' + link.attrib.get('t3')
|
|
66
|
+ url = yaml_data.get('surl')
|
|
67
|
+ img_src = yaml_data.get('imgurl')
|
|
68
|
+
|
|
69
|
+ # append result
|
|
70
|
+ results.append({'template': 'images.html',
|
|
71
|
+ 'url': url,
|
|
72
|
+ 'title': title,
|
|
73
|
+ 'content': '',
|
|
74
|
+ 'img_src': img_src})
|
|
75
|
+
|
|
76
|
+ # TODO stop parsing if 10 images are found
|
|
77
|
+ if len(results) >= 10:
|
|
78
|
+ break
|
|
79
|
+
|
|
80
|
+ # return results
|
|
81
|
+ return results
|