OwlBurst/Ponysearch
1
0
Fork 0
forked from Ponysearch/Ponysearch
Code Pull requests Activity

Merge pull request #523 from a01200356/master

Browse source 
[fix] duckduckgo's xpaths changed
This commit is contained in:
Adam Tauber 2016-03-22 11:12:50 +01:00
commit 24ea39d046
3 changed files with 29 additions and 49 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
.travis.yml
View file

@ -10,7 +10,7 @@ python:
before_install:
- "export DISPLAY=:99.0"
- "sh -e /etc/init.d/xvfb start"
- npm install -g less grunt-cli
- npm install less grunt-cli
- ( cd searx/static/themes/oscar;npm install; cd - )
install:
- ./manage.sh update_dev_packages

8
searx/engines/duckduckgo.py
View file

@ -28,10 +28,10 @@ language_support = True
url = 'https://duckduckgo.com/html?{query}&s={offset}'
# specific xpath variables
result_xpath = '//div[@class="results_links results_links_deep web-result"]' # noqa
url_xpath = './/a[@class="large"]/@href'
title_xpath = './/a[@class="large"]'
content_xpath = './/div[@class="snippet"]'
result_xpath = '//div[@class="result results_links results_links_deep web-result "]' # noqa
url_xpath = './/a[@class="result__a"]/@href'
title_xpath = './/a[@class="result__a"]'
content_xpath = './/a[@class="result__snippet"]'
# do search-request

66
tests/unit/engines/test_duckduckgo.py
View file

@ -32,20 +32,32 @@ class TestDuckduckgoEngine(SearxTestCase):
self.assertEqual(duckduckgo.response(response), [])
html = u"""
<div class="results_links results_links_deep web-result">
<div class="icon_fav" style="display: block;">
<a rel="nofollow" href="https://www.test.com/">
<img width="16" height="16" alt=""
src="/i/www.test.com.ico" style="visibility: visible;" name="i15" />
</a>
<div class="result results_links results_links_deep web-result result--no-result">
<div class="links_main links_deep result__body">
<h2 class="result__title">
</h2>
<div class="no-results">No results</div>
<div class="result__extras">
</div>
<div class="links_main links_deep"> <!-- This is the visible part -->
<a rel="nofollow" class="large" href="http://this.should.be.the.link/ű">
</div>
</div>
"""
response = mock.Mock(text=html)
results = duckduckgo.response(response)
self.assertEqual(duckduckgo.response(response), [])
html = u"""
<div class="result results_links results_links_deep web-result ">
<div class="links_main links_deep result__body">
<h2 class="result__title">
<a rel="nofollow" class="result__a" href="http://this.should.be.the.link/ű">
This <b>is</b> <b>the</b> title
</a>
<div class="snippet"><b>This</b> should be the content.</div>
<div class="url">
http://this.should.be.the.link/
</h2>
<a class="result__snippet" href="http://this.should.be.the.link/ű">
<b>This</b> should be the content.
</a>
<div class="result__extras">
</div>
</div>
</div>
@ -57,35 +69,3 @@ class TestDuckduckgoEngine(SearxTestCase):
self.assertEqual(results[0]['title'], 'This is the title')
self.assertEqual(results[0]['url'], u'http://this.should.be.the.link/ű')
self.assertEqual(results[0]['content'], 'This should be the content.')
html = """
<div class="results_links results_links_deep web-result">
<div class="icon_fav" style="display: block;">
</div>
<div class="links_main links_deep"> <!-- This is the visible part -->
<div class="snippet"><b>This</b> should be the content.</div>
<div class="url">
http://this.should.be.the.link/
</div>
</div>
</div>
<div class="results_links results_links_deep web-result">
<div class="icon_fav" style="display: block;">
<img width="16" height="16" alt=""
src="/i/www.test.com.ico" style="visibility: visible;" name="i15" />
</div>
<div class="links_main links_deep"> <!-- This is the visible part -->
<a rel="nofollow" class="large" href="">
This <b>is</b> <b>the</b> title
</a>
<div class="snippet"><b>This</b> should be the content.</div>
<div class="url">
http://this.should.be.the.link/
</div>
</div>
</div>
"""
response = mock.Mock(text=html)
results = duckduckgo.response(response)
self.assertEqual(type(results), list)
self.assertEqual(len(results), 0)