1
0
mirror of https://github.com/searxng/searxng.git synced 2024-11-04 20:30:11 +01:00

Fix bing engine results count (#1387)

This PR fixes the result count from bing which was throwing an (hidden) error and add a validation to avoid reading more results than avalaible.

For example :
If there is 100 results from some search and we try to get results from 120 to 130, Bing will send back the results from 0 to 10 and no error. If we compare results count with the first parameter of the request we can avoid this "invalid" results.
This commit is contained in:
Léo Bourrel 2019-08-05 16:15:40 +02:00 committed by Alexandre Flament
parent 12f891da84
commit 88261e111c
2 changed files with 119 additions and 43 deletions

View File

@ -13,11 +13,15 @@
@todo publishedDate
"""
import re
from lxml import html
from searx import logger, utils
from searx.engines.xpath import extract_text
from searx.url_utils import urlencode
from searx.utils import match_language, gen_useragent
logger = logger.getChild('bing engine')
# engine dependent config
categories = ['general']
paging = True
@ -30,9 +34,13 @@ base_url = 'https://www.bing.com/'
search_string = 'search?{query}&first={offset}'
def _get_offset_from_pageno(pageno):
return (pageno - 1) * 10 + 1
# do search-request
def request(query, params):
offset = (params['pageno'] - 1) * 10 + 1
offset = _get_offset_from_pageno(params.get('pageno', 0))
if params['language'] == 'all':
lang = 'EN'
@ -53,15 +61,9 @@ def request(query, params):
# get response from search-request
def response(resp):
results = []
result_len = 0
dom = html.fromstring(resp.text)
try:
results.append({'number_of_results': int(dom.xpath('//span[@class="sb_count"]/text()')[0]
.split()[0].replace(',', ''))})
except:
pass
# parse results
for result in dom.xpath('//div[@class="sa_cc"]'):
link = result.xpath('.//h3/a')[0]
@ -86,7 +88,24 @@ def response(resp):
'title': title,
'content': content})
# return results
try:
result_len_container = "".join(dom.xpath('//span[@class="sb_count"]/text()'))
result_len_container = utils.to_string(result_len_container)
if "-" in result_len_container:
# Remove the part "from-to" for paginated request ...
result_len_container = result_len_container[result_len_container.find("-") * 2 + 2:]
result_len_container = re.sub('[^0-9]', '', result_len_container)
if len(result_len_container) > 0:
result_len = int(result_len_container)
except Exception as e:
logger.debug('result error :\n%s', e)
pass
if _get_offset_from_pageno(resp.search_params.get("pageno", 0)) > result_len:
return []
results.append({'number_of_results': result_len})
return results

View File

@ -1,3 +1,4 @@
# -*- coding: utf-8 -*-
from collections import defaultdict
import mock
from searx.engines import bing
@ -10,7 +11,7 @@ class TestBingEngine(SearxTestCase):
bing.supported_languages = ['en', 'fr', 'zh-CHS', 'zh-CHT', 'pt-PT', 'pt-BR']
query = u'test_query'
dicto = defaultdict(dict)
dicto['pageno'] = 0
dicto['pageno'] = 1
dicto['language'] = 'fr-FR'
params = bing.request(query.encode('utf-8'), dicto)
self.assertTrue('url' in params)
@ -23,70 +24,126 @@ class TestBingEngine(SearxTestCase):
self.assertTrue('language' in params['url'])
def test_response(self):
dicto = defaultdict(dict)
dicto['pageno'] = 1
dicto['language'] = 'fr-FR'
self.assertRaises(AttributeError, bing.response, None)
self.assertRaises(AttributeError, bing.response, [])
self.assertRaises(AttributeError, bing.response, '')
self.assertRaises(AttributeError, bing.response, '[]')
response = mock.Mock(text='<html></html>')
response.search_params = dicto
self.assertEqual(bing.response(response), [])
response = mock.Mock(text='<html></html>')
response.search_params = dicto
self.assertEqual(bing.response(response), [])
html = """
<div class="sa_cc" u="0|5109|4755453613245655|UAGjXgIrPH5yh-o5oNHRx_3Zta87f_QO">
<div Class="sa_mc">
<div class="sb_tlst">
<h3>
<a href="http://this.should.be.the.link/" h="ID=SERP,5124.1">
<strong>This</strong> should be the title</a>
</h3>
</div>
<div class="sb_meta"><cite><strong>this</strong>.meta.com</cite>
<span class="c_tlbxTrg">
<span class="c_tlbxH" H="BASE:CACHEDPAGEDEFAULT" K="SERP,5125.1">
</span>
</span>
</div>
<p><strong>This</strong> should be the content.</p>
<div>
<div id="b_tween">
<span class="sb_count" data-bm="4">23 900 000 résultats</span>
</div>
<ol id="b_results" role="main">
<div class="sa_cc" u="0|5109|4755453613245655|UAGjXgIrPH5yh-o5oNHRx_3Zta87f_QO">
<div Class="sa_mc">
<div class="sb_tlst">
<h3>
<a href="http://this.should.be.the.link/" h="ID=SERP,5124.1">
<strong>This</strong> should be the title</a>
</h3>
</div>
<div class="sb_meta"><cite><strong>this</strong>.meta.com</cite>
<span class="c_tlbxTrg">
<span class="c_tlbxH" H="BASE:CACHEDPAGEDEFAULT" K="SERP,5125.1">
</span>
</span>
</div>
<p><strong>This</strong> should be the content.</p>
</div>
</div>
</ol>
</div>
"""
response = mock.Mock(text=html)
response.search_params = dicto
results = bing.response(response)
self.assertEqual(type(results), list)
self.assertEqual(len(results), 1)
self.assertEqual(len(results), 2)
self.assertEqual(results[0]['title'], 'This should be the title')
self.assertEqual(results[0]['url'], 'http://this.should.be.the.link/')
self.assertEqual(results[0]['content'], 'This should be the content.')
self.assertEqual(results[-1]['number_of_results'], 23900000)
html = """
<li class="b_algo" u="0|5109|4755453613245655|UAGjXgIrPH5yh-o5oNHRx_3Zta87f_QO">
<div Class="sa_mc">
<div class="sb_tlst">
<h2>
<a href="http://this.should.be.the.link/" h="ID=SERP,5124.1">
<strong>This</strong> should be the title</a>
</h2>
</div>
<div class="sb_meta"><cite><strong>this</strong>.meta.com</cite>
<span class="c_tlbxTrg">
<span class="c_tlbxH" H="BASE:CACHEDPAGEDEFAULT" K="SERP,5125.1">
</span>
</span>
</div>
<p><strong>This</strong> should be the content.</p>
<div>
<div id="b_tween">
<span class="sb_count" data-bm="4">9-18 résultats sur 23 900 000</span>
</div>
</li>
<ol id="b_results" role="main">
<li class="b_algo" u="0|5109|4755453613245655|UAGjXgIrPH5yh-o5oNHRx_3Zta87f_QO">
<div Class="sa_mc">
<div class="sb_tlst">
<h2>
<a href="http://this.should.be.the.link/" h="ID=SERP,5124.1">
<strong>This</strong> should be the title</a>
</h2>
</div>
<div class="sb_meta"><cite><strong>this</strong>.meta.com</cite>
<span class="c_tlbxTrg">
<span class="c_tlbxH" H="BASE:CACHEDPAGEDEFAULT" K="SERP,5125.1">
</span>
</span>
</div>
<p><strong>This</strong> should be the content.</p>
</div>
</li>
</ol>
</div>
"""
dicto['pageno'] = 2
response = mock.Mock(text=html)
response.search_params = dicto
results = bing.response(response)
self.assertEqual(type(results), list)
self.assertEqual(len(results), 1)
self.assertEqual(len(results), 2)
self.assertEqual(results[0]['title'], 'This should be the title')
self.assertEqual(results[0]['url'], 'http://this.should.be.the.link/')
self.assertEqual(results[0]['content'], 'This should be the content.')
self.assertEqual(results[-1]['number_of_results'], 23900000)
html = """
<div>
<div id="b_tween">
<span class="sb_count" data-bm="4">23 900 000 résultats</span>
</div>
<ol id="b_results" role="main">
<li class="b_algo" u="0|5109|4755453613245655|UAGjXgIrPH5yh-o5oNHRx_3Zta87f_QO">
<div Class="sa_mc">
<div class="sb_tlst">
<h2>
<a href="http://this.should.be.the.link/" h="ID=SERP,5124.1">
<strong>This</strong> should be the title</a>
</h2>
</div>
<div class="sb_meta"><cite><strong>this</strong>.meta.com</cite>
<span class="c_tlbxTrg">
<span class="c_tlbxH" H="BASE:CACHEDPAGEDEFAULT" K="SERP,5125.1">
</span>
</span>
</div>
<p><strong>This</strong> should be the content.</p>
</div>
</li>
</ol>
</div>
"""
dicto['pageno'] = 33900000
response = mock.Mock(text=html)
response.search_params = dicto
results = bing.response(response)
self.assertEqual(bing.response(response), [])
def test_fetch_supported_languages(self):
html = """<html></html>"""