@@ -1,16 +1,35 @@
11# SPDX-License-Identifier: AGPL-3.0-or-later
2+"""Yahoo (News)
3+
4+Yahoo News is "English only" and do not offer localized nor language queries.
5+
26"""
3− Yahoo (News)
4−"""
7+
8+# pylint: disable=invalid-name, missing-function-docstring
59
610import re
7−from datetime import datetime, timedelta
811from urllib.parse import urlencode
9−from lxml import html
10−from searx.engines.yahoo import parse_url, language_aliases
11−from searx.engines.yahoo import _fetch_supported_languages, supported_languages_url # NOQA # pylint: disable=unused-import
12+from datetime import datetime, timedelta
1213from dateutil import parser
13−from searx.utils import extract_text, extract_url, match_language
14+from lxml import html
15+
16+from searx import logger
17+from searx.utils import (
18+ eval_xpath_list,
19+ eval_xpath_getindex,
20+ extract_text,
21+)
22+
23+from searx.engines.yahoo import parse_url
24+
25+# pylint: disable=unused-import
26+from searx.engines.yahoo import (
27+ _fetch_supported_languages,
28+ supported_languages_url,
29+)
30+# pylint: enable=unused-import
31+
32+logger = logger.getChild('yahoo_news engine')
1433
1534# about
1635about = {
@@ -22,90 +41,78 @@ about = {
2241 "results": 'HTML',
2342}
2443
25−# engine dependent config
26−categories = ['news']
44+language_support = False
45+time_range_support = False
46+safesearch = False
2747paging = True
48+categories = ['news']
2849
2950# search-url
30−search_url = 'https://news.search.yahoo.com/search?{query}&b={offset}&{lang}=uh3_news_web_gs_1&pz=10&xargs=0&vl=lang_{lang}' # noqa
31−
32−# specific xpath variables
33−results_xpath = '//ol[contains(@class,"searchCenterMiddle")]//li'
34−url_xpath = './/h3/a/@href'
35−title_xpath = './/h3/a'
36−content_xpath = './/div[@class="compText"]'
37−publishedDate_xpath = './/span[contains(@class,"tri")]'
38−suggestion_xpath = '//div[contains(@class,"VerALSOTRY")]//a'
39−
51+search_url = (
52+ 'https://news.search.yahoo.com/search'
53+ '?{query}&b={offset}'
54+ )
55+
56+AGO_RE = re.compile(r'([0-9]+)\s*(year|month|week|day|minute|hour)')
57+AGO_TIMEDELTA = {
58+ 'minute': timedelta(minutes=1),
59+ 'hour': timedelta(hours=1),
60+ 'day': timedelta(days=1),
61+ 'week': timedelta(days=7),
62+ 'month': timedelta(days=30),
63+ 'year': timedelta(days=365),
64+}
4065
41−# do search-request
4266def request(query, params):
4367 offset = (params['pageno'] - 1) * 10 + 1
4468
45− if params['language'] == 'all':
46− language = 'en'
47− else:
48− language = match_language(params['language'], supported_languages, language_aliases).split('-')[0]
49−
50− params['url'] = search_url.format(offset=offset,
51− query=urlencode({'p': query}),
52− lang=language)
53−
54− # TODO required?
55− params['cookies']['sB'] = '"v=1&vm=p&fl=1&vl=lang_{lang}&sh=1&pn=10&rw=new'\
56− .format(lang=language)
69+ params['url'] = search_url.format(
70+ offset = offset,
71+ query = urlencode({'p': query})
72+ )
73+ logger.debug("query_url --> %s", params['url'])
5774 return params
5875
59−
60−def sanitize_url(url):
61− if ".yahoo.com/" in url:
62− return re.sub("\\;\\_ylt\\=.+$", "", url)
63− else:
64− return url
65−
66−
67−# get response from search-request
6876def response(resp):
6977 results = []
70−
7178 dom = html.fromstring(resp.text)
7279
80+
7381 # parse results
74− for result in dom.xpath(results_xpath):
75− urls = result.xpath(url_xpath)
76− if len(urls) != 1:
82+ for result in eval_xpath_list(dom, '//ol[contains(@class,"searchCenterMiddle")]//li'):
83+
84+ url = eval_xpath_getindex(result, './/h4/a/@href', 0, None)
85+ if url is None:
7786 continue
78− url = sanitize_url(parse_url(extract_url(urls, search_url)))
79− title = extract_text(result.xpath(title_xpath)[0])
80− content = extract_text(result.xpath(content_xpath)[0])
81−
82− # parse publishedDate
83− publishedDate = extract_text(result.xpath(publishedDate_xpath)[0])
84−
85− # still useful ?
86− if re.match("^[0-9]+ minute(s|) ago$", publishedDate):
87− publishedDate = datetime.now() - timedelta(minutes=int(re.match(r'\d+', publishedDate).group()))
88− elif re.match("^[0-9]+ days? ago$", publishedDate):
89− publishedDate = datetime.now() - timedelta(days=int(re.match(r'\d+', publishedDate).group()))
90− elif re.match("^[0-9]+ hour(s|), [0-9]+ minute(s|) ago$", publishedDate):
91− timeNumbers = re.findall(r'\d+', publishedDate)
92− publishedDate = datetime.now()\
93− - timedelta(hours=int(timeNumbers[0]))\
94− - timedelta(minutes=int(timeNumbers[1]))
87+ url = parse_url(url)
88+ title = extract_text(result.xpath('.//h4/a'))
89+ content = extract_text(result.xpath('.//p'))
90+ img_src = eval_xpath_getindex(result, './/img/@data-src', 0, None)
91+
92+ item = {
93+ 'url': url,
94+ 'title': title,
95+ 'content': content,
96+ 'img_src' : img_src
97+ }
98+
99+ pub_date = extract_text(result.xpath('.//span[contains(@class,"s-time")]'))
100+ ago = AGO_RE.search(pub_date)
101+ if ago:
102+ number = int(ago.group(1))
103+ delta = AGO_TIMEDELTA[ago.group(2)]
104+ pub_date = datetime.now() - delta * number
95105 else:
96106 try:
97− publishedDate = parser.parse(publishedDate)
98− except:
99− publishedDate = datetime.now()
107+ pub_date = parser.parse(pub_date)
108+ except parser.ParserError:
109+ pub_date = None
100110
101− if publishedDate.year == 1900:
102− publishedDate = publishedDate.replace(year=datetime.now().year)
111+ if pub_date is not None:
112+ item['publishedDate'] = pub_date
113+ results.append(item)
103114
104− # append result
105− results.append({'url': url,
106− 'title': title,
107− 'content': content,
108− 'publishedDate': publishedDate})
115+ for suggestion in eval_xpath_list(dom, '//div[contains(@class,"AlsoTry")]//td'):
116+ results.append({'suggestion': extract_text(suggestion)})
109117
110− # return results
111118 return results