kamp/searxng
1
0
Fork 0
mirror of https://github.com/searxng/searxng.git synced 2024-11-14 16:50:11 +01:00
Code
f17e54a396
searxng/tests/unit/engines/test_google.py

247 lines
9.7 KiB
Python
Raw Normal View History

Google's unit test
2015-02-11 17:16:52 +01:00
# -*- coding: utf-8 -*-
from collections import defaultdict
import mock
import lxml
from searx.engines import google
from searx.testing import SearxTestCase
class TestGoogleEngine(SearxTestCase):
[enh] google engine : avoid some "sorry google" by adding another cookie : NID. This cookie is specific by hostname. This allow to send request to google.* (according to the search language). Before this commit, request in other languages than english was sent to www.google.com which was redirected to www.google.* The PREF is still use on the www.google.com domain.
2015-05-30 17:41:40 +02:00
def mock_response(self, text):
[fix] prevent google engine to redirect nid/pref cookies are also removed
2015-12-22 20:05:42 +01:00
response = mock.Mock(text=text, url='https://www.google.com/search?q=test&start=0&gbv=1&gws_rd=cr')
[enh] google engine : avoid some "sorry google" by adding another cookie : NID. This cookie is specific by hostname. This allow to send request to google.* (according to the search language). Before this commit, request in other languages than english was sent to www.google.com which was redirected to www.google.* The PREF is still use on the www.google.com domain.
2015-05-30 17:41:40 +02:00
response.search_params = mock.Mock()
response.search_params.get = mock.Mock(return_value='www.google.com')
return response
Google's unit test
2015-02-11 17:16:52 +01:00
def test_request(self):
refactor engine's search language handling Add match_language function in utils to match any user given language code with a list of engine's supported languages. Also add language_aliases dict on each engine to translate standard language codes into the custom codes used by the engine.
2018-03-01 05:30:48 +01:00
google.supported_languages = ['en', 'fr', 'zh-CN']
Google's unit test
2015-02-11 17:16:52 +01:00
query = 'test_query'
dicto = defaultdict(dict)
dicto['pageno'] = 1
Add language support for more engines.
2016-10-30 03:04:01 +01:00
dicto['language'] = 'fr-FR'
[fix] time range detection
2016-07-26 00:22:05 +02:00
dicto['time_range'] = ''
Google's unit test
2015-02-11 17:16:52 +01:00
params = google.request(query, dicto)
self.assertIn('url', params)
self.assertIn(query, params['url'])
[enh] google engine : avoid some "sorry google" by adding another cookie : NID. This cookie is specific by hostname. This allow to send request to google.* (according to the search language). Before this commit, request in other languages than english was sent to www.google.com which was redirected to www.google.* The PREF is still use on the www.google.com domain.
2015-05-30 17:41:40 +02:00
self.assertIn('google.fr', params['url'])
[fix] force English results in Google when using en-US
2018-04-19 05:55:37 +02:00
self.assertIn('fr', params['url'])
Google's unit test
2015-02-11 17:16:52 +01:00
self.assertIn('fr', params['headers']['Accept-Language'])
remove 'all' option from search languages
2017-07-20 22:47:20 +02:00
dicto['language'] = 'en-US'
Google's unit test
2015-02-11 17:16:52 +01:00
params = google.request(query, dicto)
[fix] force English results in Google when using en-US
2018-04-19 05:55:37 +02:00
self.assertIn('google.com', params['url'])
self.assertIn('en', params['url'])
Google's unit test
2015-02-11 17:16:52 +01:00
self.assertIn('en', params['headers']['Accept-Language'])
refactor engine's search language handling Add match_language function in utils to match any user given language code with a list of engine's supported languages. Also add language_aliases dict on each engine to translate standard language codes into the custom codes used by the engine.
2018-03-01 05:30:48 +01:00
dicto['language'] = 'zh'
params = google.request(query, dicto)
self.assertIn('google.com', params['url'])
[fix] force English results in Google when using en-US
2018-04-19 05:55:37 +02:00
self.assertIn('zh-CN', params['url'])
refactor engine's search language handling Add match_language function in utils to match any user given language code with a list of engine's supported languages. Also add language_aliases dict on each engine to translate standard language codes into the custom codes used by the engine.
2018-03-01 05:30:48 +01:00
self.assertIn('zh-CN', params['headers']['Accept-Language'])
Google's unit test
2015-02-11 17:16:52 +01:00
def test_response(self):
self.assertRaises(AttributeError, google.response, None)
self.assertRaises(AttributeError, google.response, [])
self.assertRaises(AttributeError, google.response, '')
self.assertRaises(AttributeError, google.response, '[]')
[enh] google engine : avoid some "sorry google" by adding another cookie : NID. This cookie is specific by hostname. This allow to send request to google.* (according to the search language). Before this commit, request in other languages than english was sent to www.google.com which was redirected to www.google.* The PREF is still use on the www.google.com domain.
2015-05-30 17:41:40 +02:00
response = self.mock_response('<html></html>')
Google's unit test
2015-02-11 17:16:52 +01:00
self.assertEqual(google.response(response), [])
html = """
[fix] google markup change - closes #489
2016-01-10 18:49:50 +01:00
<div class="g">
Google's unit test
2015-02-11 17:16:52 +01:00
<h3 class="r">
<a href="http://this.should.be.the.link/">
<b>This</b> is <b>the</b> title
</a>
</h3>
<div class="s">
<div class="kv" style="margin-bottom:2px">
<cite>
<b>test</b>.psychologies.com/
</cite>
<div class="_nBb">‎
<div style="display:inline" onclick="google.sham(this);" aria-expanded="false"
aria-haspopup="true" tabindex="0" data-ved="0CBUQ7B0wAA">
<span class="_O0">
</span>
</div>
<div style="display:none" class="am-dropdown-menu" role="menu" tabindex="-1">
<ul>
<li class="_Ykb">
<a class="_Zkb" href="http://www.google.fr/url?url=http://webcache.googleusercontent
.com/search%3Fcache:R1Z_4pGXjuIJ:http://test.psychologies.com/">
En cache
</a>
</li>
<li class="_Ykb">
<a class="_Zkb" href="/search?safe=off&amp;q=related:test.psy.com/">
Pages similaires
</a>
</li>
</ul>
</div>
</div>
</div>
<span class="st">
This should be the content.
</span>
<br>
<div class="osl">‎
<a href="http://www.google.fr/url?url=http://test.psychologies.com/tests/">
Test Personnalité
</a> - ‎
<a href="http://www.google.fr/url?url=http://test.psychologies.com/test/">
Tests - Moi
</a> - ‎
<a href="http://www.google.fr/url?url=http://test.psychologies.com/test/tests-couple">
Test Couple
</a>
- ‎
<a href="http://www.google.fr/url?url=http://test.psychologies.com/tests/tests-amour">
Test Amour
</a>
</div>
</div>
[fix] google markup change - closes #489
2016-01-10 18:49:50 +01:00
</div>
<div class="g">
Google's unit test
2015-02-11 17:16:52 +01:00
<h3 class="r">
<a href="http://www.google.com/images?q=toto">
<b>This</b>
</a>
</h3>
[fix] google markup change - closes #489
2016-01-10 18:49:50 +01:00
</div>
<div class="g">
Google's unit test
2015-02-11 17:16:52 +01:00
<h3 class="r">
<a href="http://www.google.com/search?q=toto">
<b>This</b> is
</a>
</h3>
[fix] google markup change - closes #489
2016-01-10 18:49:50 +01:00
</div>
<div class="g">
Google's unit test
2015-02-11 17:16:52 +01:00
<h3 class="r">
<a href="€">
<b>This</b> is <b>the</b>
</a>
</h3>
[fix] google markup change - closes #489
2016-01-10 18:49:50 +01:00
</div>
<div class="g">
Google's unit test
2015-02-11 17:16:52 +01:00
<h3 class="r">
<a href="/url?q=url">
<b>This</b> is <b>the</b>
</a>
</h3>
[fix] google markup change - closes #489
2016-01-10 18:49:50 +01:00
</div>
Google's unit test
2015-02-11 17:16:52 +01:00
<p class="_Bmc" style="margin:3px 8px">
<a href="/search?num=20&amp;safe=off&amp;q=t&amp;revid=1754833769&amp;sa=X&amp;ei=-&amp;ved=">
suggestion <b>title</b>
</a>
</p>
"""
[enh] google engine : avoid some "sorry google" by adding another cookie : NID. This cookie is specific by hostname. This allow to send request to google.* (according to the search language). Before this commit, request in other languages than english was sent to www.google.com which was redirected to www.google.* The PREF is still use on the www.google.com domain.
2015-05-30 17:41:40 +02:00
response = self.mock_response(html)
Google's unit test
2015-02-11 17:16:52 +01:00
results = google.response(response)
self.assertEqual(type(results), list)
self.assertEqual(len(results), 2)
self.assertEqual(results[0]['title'], 'This is 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]['suggestion'], 'suggestion title')
html = """
<li class="b_algo" u="0|5109|4755453613245655|UAGjXgIrPH5yh-o5oNHRx_3Zta87f_QO">
</li>
"""
[enh] google engine : avoid some "sorry google" by adding another cookie : NID. This cookie is specific by hostname. This allow to send request to google.* (according to the search language). Before this commit, request in other languages than english was sent to www.google.com which was redirected to www.google.* The PREF is still use on the www.google.com domain.
2015-05-30 17:41:40 +02:00
response = self.mock_response(html)
Google's unit test
2015-02-11 17:16:52 +01:00
results = google.response(response)
self.assertEqual(type(results), list)
self.assertEqual(len(results), 0)
[enh] google engine : avoid some "sorry google" by adding another cookie : NID. This cookie is specific by hostname. This allow to send request to google.* (according to the search language). Before this commit, request in other languages than english was sent to www.google.com which was redirected to www.google.* The PREF is still use on the www.google.com domain.
2015-05-30 17:41:40 +02:00
response = mock.Mock(text='<html></html>', url='https://sorry.google.com')
response.search_params = mock.Mock()
response.search_params.get = mock.Mock(return_value='www.google.com')
self.assertRaises(RuntimeWarning, google.response, response)
response = mock.Mock(text='<html></html>', url='https://www.google.com/sorry/IndexRedirect')
response.search_params = mock.Mock()
response.search_params.get = mock.Mock(return_value='www.google.com')
self.assertRaises(RuntimeWarning, google.response, response)
Google's unit test
2015-02-11 17:16:52 +01:00
def test_parse_images(self):
html = """
<li>
<div>
<a href="http://www.google.com/url?q=http://this.is.the.url/">
<img style="margin:3px 0;margin-right:6px;padding:0" height="90"
src="https://this.is.the.image/image.jpg" width="60" align="middle" alt="" border="0">
</a>
</div>
</li>
"""
dom = lxml.html.fromstring(html)
[enh] google engine : avoid some "sorry google" by adding another cookie : NID. This cookie is specific by hostname. This allow to send request to google.* (according to the search language). Before this commit, request in other languages than english was sent to www.google.com which was redirected to www.google.* The PREF is still use on the www.google.com domain.
2015-05-30 17:41:40 +02:00
results = google.parse_images(dom, 'www.google.com')
Google's unit test
2015-02-11 17:16:52 +01:00
self.assertEqual(type(results), list)
self.assertEqual(len(results), 1)
self.assertEqual(results[0]['url'], 'http://this.is.the.url/')
self.assertEqual(results[0]['title'], '')
self.assertEqual(results[0]['content'], '')
self.assertEqual(results[0]['img_src'], 'https://this.is.the.image/image.jpg')
tests for _fetch_supported_languages in engines and refactor method to make it testable without making requests
2016-12-15 07:34:43 +01:00
def test_fetch_supported_languages(self):
html = """<html></html>"""
response = mock.Mock(text=html)
languages = google._fetch_supported_languages(response)
self.assertEqual(type(languages), dict)
self.assertEqual(len(languages), 0)
html = u"""
<html>
<body>
<table>
<tbody>
<tr>
<td>
<font>
<label>
<span id="ten">English</span>
</label>
</font>
</td>
<td>
<font>
<label>
<span id="tzh-CN">中文 (简体)</span>
</label>
<label>
<span id="tzh-TW">中文 (繁體)</span>
</label>
</font>
</td>
</tr>
</tbody>
</table>
</body>
</html>
"""
response = mock.Mock(text=html)
languages = google._fetch_supported_languages(response)
self.assertEqual(type(languages), dict)
self.assertEqual(len(languages), 3)
self.assertIn('en', languages)
self.assertIn('zh-CN', languages)
self.assertIn('zh-TW', languages)
self.assertEquals(type(languages['en']), dict)
self.assertEquals(type(languages['zh-CN']), dict)
self.assertEquals(type(languages['zh-TW']), dict)
self.assertIn('name', languages['en'])
self.assertIn('name', languages['zh-CN'])
self.assertIn('name', languages['zh-TW'])
self.assertEquals(languages['en']['name'], 'English')
self.assertEquals(languages['zh-CN']['name'], u'中文 (简体)')
self.assertEquals(languages['zh-TW']['name'], u'中文 (繁體)')
Copy Permalink