@@ -1,15 +1,42 @@
11# SPDX-License-Identifier: AGPL-3.0-or-later
2−"""
3− Qwant (Web, Images, News, Social)
2+# lint: pylint
3+"""Qwant (Web, News, Images, Videos)
4+
5+This engine uses the Qwant API (https://api.qwant.com/v3). The API is
6+undocumented but can be reverse engineered by reading the network log of
7+https://www.qwant.com/ queries.
8+
9+This implementation is used by different qwant engines in the settings.yml::
10+
11+ - name: qwant
12+ categories: general
13+ ...
14+ - name: qwant news
15+ categories: news
16+ ...
17+ - name: qwant images
18+ categories: images
19+ ...
20+ - name: qwant videos
21+ categories: videos
22+ ...
23+
424"""
525
6−from datetime import datetime
26+from datetime import (
27+ datetime,
28+ timedelta,
29+)
730from json import loads
831from urllib.parse import urlencode
9−from searx.utils import html_to_text, match_language
10−from searx.exceptions import SearxEngineAPIException, SearxEngineCaptchaException
32+
33+# from searx import logger
34+from searx.utils import match_language
35+from searx.exceptions import SearxEngineAPIException
1136from searx.network import raise_for_httperror
1237
38+#logger = logger.getChild('qwant')
39+
1340# about
1441about = {
1542 "website": 'https://www.qwant.com/',
@@ -25,98 +52,148 @@ categories = []
2552paging = True
2653supported_languages_url = about['website']
2754
28−category_to_keyword = {'general': 'web',
29− 'images': 'images',
30− 'news': 'news'}
55+category_to_keyword = {
56+ 'general': 'web',
57+ 'news': 'news',
58+ 'images': 'images',
59+ 'videos': 'videos',
60+}
3161
3262# search-url
33−url = 'https://api.qwant.com/api/search/{keyword}?count=10&offset={offset}&f=&{query}&t={keyword}&uiv=4'
34−
63+url = 'https://api.qwant.com/v3/search/{keyword}?q={query}&count={count}&offset={offset}'
3564
36−# do search-request
3765def request(query, params):
38− offset = (params['pageno'] - 1) * 10
39−
40− if categories[0] and categories[0] in category_to_keyword:
41−
42− params['url'] = url.format(keyword=category_to_keyword[categories[0]],
43− query=urlencode({'q': query}),
44− offset=offset)
66+ """Qwant search request"""
67+ keyword = category_to_keyword[categories[0]]
68+ count = 10 # web: count must be equal to 10
69+
70+ if keyword == 'images':
71+ count = 50
72+ offset = (params['pageno'] - 1) * count
73+ # count + offset must be lower than 250
74+ offset = min(offset, 199)
4575 else:
46− params['url'] = url.format(keyword='web',
47− query=urlencode({'q': query}),
48− offset=offset)
76+ offset = (params['pageno'] - 1) * count
77+ # count + offset must be lower than 50
78+ offset = min(offset, 40)
79+
80+ params['url'] = url.format(
81+ keyword = keyword,
82+ query = urlencode({'q': query}),
83+ offset = offset,
84+ count = count,
85+ )
4986
5087 # add language tag
5188 if params['language'] != 'all':
52− language = match_language(params['language'], supported_languages, language_aliases)
53− params['url'] += '&locale=' + language.replace('-', '_').lower()
89+ language = match_language(
90+ params['language'],
91+ # pylint: disable=undefined-variable
92+ supported_languages,
93+ language_aliases,
94+ )
95+ params['url'] += '&locale=' + language.replace('-', '_')
5496
55− params['headers']['User-Agent'] = 'Mozilla/5.0 (X11; Linux x86_64; rv:69.0) Gecko/20100101 Firefox/69.0'
5697 params['raise_for_httperror'] = False
5798 return params
5899
59100
60−# get response from search-request
61101def response(resp):
62− results = []
102+ """Get response from Qwant's search request"""
63103
64− # According to https://www.qwant.com/js/app.js
65− if resp.status_code == 429:
66− raise SearxEngineCaptchaException()
67−
68− # raise for other errors
69− raise_for_httperror(resp)
104+ keyword = category_to_keyword[categories[0]]
105+ results = []
70106
71107 # load JSON result
72108 search_results = loads(resp.text)
109+ data = search_results.get('data', {})
73110
74111 # check for an API error
75112 if search_results.get('status') != 'success':
76− raise SearxEngineAPIException('API error ' + str(search_results.get('error', '')))
113+ msg = ",".join(data.get('message', ['unknown', ]))
114+ raise SearxEngineAPIException('API error::' + msg)
115+
116+ # raise for other errors
117+ raise_for_httperror(resp)
118+
119+ if keyword == 'web':
120+ # The WEB query contains a list named 'mainline'. This list can contain
121+ # different result types (e.g. mainline[0]['type'] returns type of the
122+ # result items in mainline[0]['items']
123+ mainline = data.get('result', {}).get('items', {}).get('mainline', {})
124+ else:
125+ # Queries on News, Images and Videos do not have a list named 'mainline'
126+ # in the response. The result items are directly in the list
127+ # result['items'].
128+ mainline = data.get('result', {}).get('items', [])
129+ mainline = [
130+ {'type' : keyword, 'items' : mainline },
131+ ]
77132
78133 # return empty array if there are no results
79− if 'data' not in search_results:
134+ if not mainline:
80135 return []
81136
82− data = search_results.get('data', {})
137+ for row in mainline:
138+
139+ mainline_type = row.get('type', 'web')
140+ if mainline_type == 'ads':
141+ # ignore adds
142+ continue
143+
144+ mainline_items = row.get('items', [])
145+ for item in mainline_items:
146+
147+ title = item['title']
148+ res_url = item['url']
149+
150+ if mainline_type == 'web':
151+ content = item['desc']
152+ results.append({
153+ 'title': title,
154+ 'url': res_url,
155+ 'content': content,
156+ })
83157
84− res = data.get('result', {})
85−
86− # parse results
87− for result in res.get('items', {}):
88−
89− title = html_to_text(result['title'])
90− res_url = result['url']
91− content = html_to_text(result['desc'])
92−
93− if category_to_keyword.get(categories[0], '') == 'web':
94− results.append({'title': title,
95− 'content': content,
96− 'url': res_url})
97−
98− elif category_to_keyword.get(categories[0], '') == 'images':
99− thumbnail_src = result['thumbnail']
100− img_src = result['media']
101− results.append({'template': 'images.html',
102− 'url': res_url,
103− 'title': title,
104− 'content': '',
105− 'thumbnail_src': thumbnail_src,
106− 'img_src': img_src})
107−
108− elif category_to_keyword.get(categories[0], '') == 'news':
109− published_date = datetime.fromtimestamp(result['date'], None)
110− media = result.get('media', [])
111− if len(media) > 0:
112− img_src = media[0].get('pict', {}).get('url', None)
113− else:
158+ elif mainline_type == 'news':
159+ pub_date = datetime.fromtimestamp(item['date'], None)
160+ news_media = item.get('media', [])
114161 img_src = None
115− results.append({'url': res_url,
116− 'title': title,
117− 'publishedDate': published_date,
118− 'content': content,
119− 'img_src': img_src})
162+ if news_media:
163+ img_src = news_media[0].get('pict', {}).get('url', None)
164+ results.append({
165+ 'title': title,
166+ 'url': res_url,
167+ 'publishedDate': pub_date,
168+ 'img_src': img_src,
169+ })
170+
171+ elif mainline_type == 'images':
172+ thumbnail = item['thumbnail']
173+ img_src = item['media']
174+ results.append({
175+ 'title': title,
176+ 'url': res_url,
177+ 'template': 'images.html',
178+ 'thumbnail_src': thumbnail,
179+ 'img_src': img_src,
180+ })
181+
182+ elif mainline_type == 'videos':
183+ content = item['desc']
184+ length = timedelta(seconds=item['duration'])
185+ pub_date = datetime.fromtimestamp(item['date'])
186+ thumbnail = item['thumbnail']
187+
188+ results.append({
189+ 'title': title,
190+ 'url': res_url,
191+ 'content': content,
192+ 'publishedDate': pub_date,
193+ 'thumbnail': thumbnail,
194+ 'template': 'videos.html',
195+ 'length': length,
196+ })
120197
121198 return results
122199