git.iko.soy/searx
on master help sign in
[fix] engine - yahoo: rewrite and fix issues
4d36aeeMarkus Heiser
searx/data/engines_languages.json+2−2
compare with master
unifiedsplit
@@ -28582,7 +28582,7 @@
28582 "sv",
28582 "sv",
28583 "th",
28583 "th",
28584 "tr",
28584 "tr",
28585− "zh-CHS",
28585+ "zh_chs",
28586− "zh-CHT"
28586+ "zh_cht"
28587 ]
28587 ]
28588}
28588}
searx/engines/yahoo.py+93−91
@@ -3,9 +3,18 @@
3 Yahoo (Web)
3 Yahoo (Web)
4"""
4"""
5
5
6−from urllib.parse import unquote, urlencode
6+from urllib.parse import (
7+ unquote,
8+ urlencode,
9+)
7from lxml import html
10from lxml import html
11+
8−from searx.utils import extract_text, extract_url, match_language, eval_xpath
12+from searx.utils import (
13+ eval_xpath_getindex,
14+ eval_xpath_list,
15+ extract_text,
16+ match_language,
17+)
9
18
10# about
19# about
11about = {
20about = {
@@ -21,29 +30,75 @@ about = {
21categories = ['general']
30categories = ['general']
22paging = True
31paging = True
23time_range_support = True
32time_range_support = True
33+supported_languages_url = 'https://search.yahoo.com/preferences/languages'
34+"""Supported languages are read from Yahoo preference page."""
35+
36+time_range_dict = {
37+ 'day': ('1d', 'd'),
38+ 'week': ('1w', 'w'),
39+ 'month': ('1m', 'm'),
40+}
41+
42+language_aliases = {
43+ 'zh-HK': 'zh_chs',
44+ 'zh-CN': 'zh_chs', # dead since 2015 / routed to hk.search.yahoo.com
45+ 'zh-TW': 'zh_cht',
46+}
47+
48+lang2domain = {
49+ 'zh_chs': 'hk.search.yahoo.com',
50+ 'zh_cht': 'tw.search.yahoo.com',
51+ 'en': 'search.yahoo.com',
52+
53+ 'bg': 'search.yahoo.com',
54+ 'cs': 'search.yahoo.com',
55+ 'da': 'search.yahoo.com',
56+ 'el': 'search.yahoo.com',
57+ 'et': 'search.yahoo.com',
58+ 'he': 'search.yahoo.com',
59+ 'hr': 'search.yahoo.com',
60+ 'ja': 'search.yahoo.com',
61+ 'ko': 'search.yahoo.com',
62+ 'sk': 'search.yahoo.com',
63+ 'sl': 'search.yahoo.com',
64+
65+}
24
66
25−# search-url
26−base_url = 'https://search.yahoo.com/'
27−search_url = 'search?{query}&b={offset}&fl=1&vl=lang_{lang}'
28−search_url_with_time = 'search?{query}&b={offset}&fl=1&vl=lang_{lang}&age={age}&btf={btf}&fr2=time'
29
67
30−supported_languages_url = 'https://search.yahoo.com/web/advanced'
68+def _get_language(params):
31
69
32−# specific xpath variables
70+ lang = language_aliases.get(params['language'])
33−results_xpath = "//div[contains(concat(' ', normalize-space(@class), ' '), ' Sr ')]"
71+ if lang is None:
34−url_xpath = './/h3/a/@href'
72+ lang = match_language(
35−title_xpath = './/h3/a'
73+ params['language'], supported_languages, language_aliases
36−content_xpath = './/div[contains(@class, "compText")]'
74+ )
37−suggestion_xpath = "//div[contains(concat(' ', normalize-space(@class), ' '), ' AlsoTry ')]//a"
75+ lang = lang.split('-')[0]
76+ return lang
38
77
39−time_range_dict = {'day': ['1d', 'd'],
40− 'week': ['1w', 'w'],
41− 'month': ['1m', 'm']}
42
78
43−language_aliases = {'zh-CN': 'zh-CHS', 'zh-TW': 'zh-CHT', 'zh-HK': 'zh-CHT'}
79+def request(query, params):
80+ """build request"""
81+ offset = (params['pageno'] - 1) * 7 + 1
82+ lang = _get_language(params)
83+ age, btf = time_range_dict.get(params['time_range'], ('', ''))
84+
85+ args = urlencode({
86+ 'p': query,
87+ 'ei': 'UTF-8',
88+ 'fl': 1,
89+ 'vl': 'lang_' + lang,
90+ 'btf': btf,
91+ 'fr2': 'time',
92+ 'age': age,
93+ 'b': offset,
94+ 'xargs': 0,
95+ })
96+
97+ domain = lang2domain.get(lang, '%s.search.yahoo.com' % lang)
98+ params['url'] = 'https://%s/search?%s' % (domain, args)
99+ return params
44
100
45
101
46−# remove yahoo-specific tracking-url
47def parse_url(url_string):
102def parse_url(url_string):
48 endings = ['/RS', '/RK']
103 endings = ['/RS', '/RK']
49 endpositions = []
104 endpositions = []
@@ -61,86 +116,37 @@ def parse_url(url_string):
61 return unquote(url_string[start:end])
116 return unquote(url_string[start:end])
62
117
63
118
64−def _get_url(query, offset, language, time_range):
65− if time_range in time_range_dict:
66− return base_url + search_url_with_time.format(offset=offset,
67− query=urlencode({'p': query}),
68− lang=language,
69− age=time_range_dict[time_range][0],
70− btf=time_range_dict[time_range][1])
71− return base_url + search_url.format(offset=offset,
72− query=urlencode({'p': query}),
73− lang=language)
74−
75−
76−def _get_language(params):
77− if params['language'] == 'all':
78− return 'en'
79−
80− language = match_language(params['language'], supported_languages, language_aliases)
81− if language not in language_aliases.values():
82− language = language.split('-')[0]
83− language = language.replace('-', '_').lower()
84−
85− return language
86−
87−
88−# do search-request
89−def request(query, params):
90− if params['time_range'] and params['time_range'] not in time_range_dict:
91− return params
92−
93− offset = (params['pageno'] - 1) * 10 + 1
94− language = _get_language(params)
95−
96− params['url'] = _get_url(query, offset, language, params['time_range'])
97−
98− # TODO required?
99− params['cookies']['sB'] = 'fl=1&vl=lang_{lang}&sh=1&rw=new&v=1'\
100− .format(lang=language)
101−
102− return params
103−
104−
105−# get response from search-request
106def response(resp):
119def response(resp):
107 results = []
120 results = []
108−
109 dom = html.fromstring(resp.text)
121 dom = html.fromstring(resp.text)
110
122
111− try:
123+ for result in eval_xpath_list(dom, '//div[contains(@class,"algo-sr")]'):
112− results_num = int(eval_xpath(dom, '//div[@class="compPagination"]/span[last()]/text()')[0]
124+ url = eval_xpath_getindex(result, './/h3/a/@href', 0, default=None)
113− .split()[0].replace(',', ''))
125+ if url is None:
114− results.append({'number_of_results': results_num})
115− except:
116− pass
117−
118− # parse results
119− for result in eval_xpath(dom, results_xpath):
120− try:
121− url = parse_url(extract_url(eval_xpath(result, url_xpath), search_url))
122− title = extract_text(eval_xpath(result, title_xpath)[0])
123− except:
124 continue
126 continue
127+ url = parse_url(url)
125
128
126− content = extract_text(eval_xpath(result, content_xpath)[0])
129+ title = eval_xpath_getindex(result, './/h3/a', 0, default=None)
130+ if title is None:
131+ continue
132+ offset = len(extract_text(title.xpath('span')))
133+ title = extract_text(title)[offset:]
134+
135+ content = eval_xpath_getindex(
136+ result, './/div[contains(@class, "compText")]', 0, default=''
137+ )
138+ if content:
139+ content = extract_text(content)
127
140
128 # append result
141 # append result
129 results.append({'url': url,
142 results.append({'url': url,
130 'title': title,
143 'title': title,
131 'content': content})
144 'content': content})
132
145
133− # if no suggestion found, return results
146+ for suggestion in eval_xpath_list(dom, '//div[contains(@class, "AlsoTry")]'):
134− suggestions = eval_xpath(dom, suggestion_xpath)
135− if not suggestions:
136− return results
137−
138− # parse suggestion
139− for suggestion in suggestions:
140 # append suggestion
147 # append suggestion
141 results.append({'suggestion': extract_text(suggestion)})
148 results.append({'suggestion': extract_text(suggestion)})
142
149
143− # return results
144 return results
150 return results
145
151
146
152
@@ -148,13 +154,9 @@ def response(resp):
148def _fetch_supported_languages(resp):
154def _fetch_supported_languages(resp):
149 supported_languages = []
155 supported_languages = []
150 dom = html.fromstring(resp.text)
156 dom = html.fromstring(resp.text)
151− options = eval_xpath(dom, '//div[@id="yschlang"]/span/label/input')
157+ offset = len('lang_')
152− for option in options:
158+
153− code_parts = eval_xpath(option, './@value')[0][5:].split('_')
159+ for val in eval_xpath_list(dom, '//div[contains(@class, "lang-item")]/input/@value'):
154− if len(code_parts) == 2:
155− code = code_parts[0] + '-' + code_parts[1].upper()
156− else:
157− code = code_parts[0]
158− supported_languages.append(code)
160+ supported_languages.append(val[offset:])
159
161
160 return supported_languages
162 return supported_languages