@@ -1,16 +1,35 @@
1# SPDX-License-Identifier: AGPL-3.0-or-later
1# SPDX-License-Identifier: AGPL-3.0-or-later
4+Yahoo News is "English only" and do not offer localized nor language queries.
4−"""
8+# pylint: disable=invalid-name, missing-function-docstring
7−from datetime import datetime, timedelta
8from urllib.parse import urlencode
11from urllib.parse import urlencode
9−from lxml import html
12+from datetime import datetime, timedelta
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
12from dateutil import parser
13from dateutil import parser
16+from searx import logger
13−from searx.utils import extract_text, extract_url, match_language
17+from searx.utils import (
23+from searx.engines.yahoo import parse_url
25+# pylint: disable=unused-import
26+from searx.engines.yahoo import (
27+ _fetch_supported_languages,
28+ supported_languages_url,
30+# pylint: enable=unused-import
32+logger = logger.getChild('yahoo_news engine')
@@ -22,90 +41,78 @@ about = {
22 "results": 'HTML',
41 "results": 'HTML',
25−# engine dependent config
44+language_support = False
26−categories = ['news']
45+time_range_support = False
27paging = True
47paging = True
29# search-url
50# 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
52+ 'https://news.search.yahoo.com/search'
31−
53+ '?{query}&b={offset}'
32−# specific xpath variables
54+ )
33−results_xpath = '//ol[contains(@class,"searchCenterMiddle")]//li'
55+
34−url_xpath = './/h3/a/@href'
56+AGO_RE = re.compile(r'([0-9]+)\s*(year|month|week|day|minute|hour)')
35−title_xpath = './/h3/a'
57+AGO_TIMEDELTA = {
36−content_xpath = './/div[@class="compText"]'
58+ 'minute': timedelta(minutes=1),
37−publishedDate_xpath = './/span[contains(@class,"tri")]'
59+ 'hour': timedelta(hours=1),
38−suggestion_xpath = '//div[contains(@class,"VerALSOTRY")]//a'
60+ 'day': timedelta(days=1),
39−
61+ 'week': timedelta(days=7),
62+ 'month': timedelta(days=30),
63+ 'year': timedelta(days=365),
42def request(query, params):
66def request(query, params):
43 offset = (params['pageno'] - 1) * 10 + 1
67 offset = (params['pageno'] - 1) * 10 + 1
45− if params['language'] == 'all':
48− language = match_language(params['language'], supported_languages, language_aliases).split('-')[0]
50− params['url'] = search_url.format(offset=offset,
69+ params['url'] = search_url.format(
51− query=urlencode({'p': query}),
71+ query = urlencode({'p': query})
53−
73+ logger.debug("query_url --> %s", params['url'])
55− params['cookies']['sB'] = '"v=1&vm=p&fl=1&vl=lang_{lang}&sh=1&pn=10&rw=new'\
56− .format(lang=language)
57 return params
74 return params
60−def sanitize_url(url):
61− if ".yahoo.com/" in url:
62− return re.sub("\\;\\_ylt\\=.+$", "", url)
67−# get response from search-request
68def response(resp):
76def response(resp):
69 results = []
77 results = []
71 dom = html.fromstring(resp.text)
78 dom = html.fromstring(resp.text)
73 # parse results
81 # parse results
74− for result in dom.xpath(results_xpath):
82+ for result in eval_xpath_list(dom, '//ol[contains(@class,"searchCenterMiddle")]//li'):
75− urls = result.xpath(url_xpath)
83+
76− if len(urls) != 1:
84+ url = eval_xpath_getindex(result, './/h4/a/@href', 0, None)
78− url = sanitize_url(parse_url(extract_url(urls, search_url)))
87+ url = parse_url(url)
79− title = extract_text(result.xpath(title_xpath)[0])
88+ title = extract_text(result.xpath('.//h4/a'))
80− content = extract_text(result.xpath(content_xpath)[0])
89+ content = extract_text(result.xpath('.//p'))
81−
90+ img_src = eval_xpath_getindex(result, './/img/@data-src', 0, None)
82− # parse publishedDate
91+
83− publishedDate = extract_text(result.xpath(publishedDate_xpath)[0])
99+ pub_date = extract_text(result.xpath('.//span[contains(@class,"s-time")]'))
84−
100+ ago = AGO_RE.search(pub_date)
85− # still useful ?
101+ if ago:
86− if re.match("^[0-9]+ minute(s|) ago$", publishedDate):
102+ number = int(ago.group(1))
87− publishedDate = datetime.now() - timedelta(minutes=int(re.match(r'\d+', publishedDate).group()))
103+ delta = AGO_TIMEDELTA[ago.group(2)]
88− elif re.match("^[0-9]+ days? ago$", publishedDate):
104+ pub_date = datetime.now() - delta * number
89− publishedDate = datetime.now() - timedelta(days=int(re.match(r'\d+', publishedDate).group()))
90− elif re.match("^[0-9]+ hour(s|), [0-9]+ minute(s|) ago$", publishedDate):
91− timeNumbers = re.findall(r'\d+', publishedDate)
92− publishedDate = datetime.now()\
93− - timedelta(hours=int(timeNumbers[0]))\
94− - timedelta(minutes=int(timeNumbers[1]))
97− publishedDate = parser.parse(publishedDate)
107+ pub_date = parser.parse(pub_date)
98− except:
108+ except parser.ParserError:
99− publishedDate = datetime.now()
109+ pub_date = None
101− if publishedDate.year == 1900:
111+ if pub_date is not None:
102− publishedDate = publishedDate.replace(year=datetime.now().year)
112+ item['publishedDate'] = pub_date
113+ results.append(item)
104− # append result
115+ for suggestion in eval_xpath_list(dom, '//div[contains(@class,"AlsoTry")]//td'):
105− results.append({'url': url,
116+ results.append({'suggestion': extract_text(suggestion)})
108− 'publishedDate': publishedDate})
111 return results
118 return results