searxngRebrandZaclys/searx/engines/duckduckgo.py
a01200356 751c9a346e [fix] duckduckgo's xpaths changed
test_duckduckgo modified to reflect changes in duckduckgo's html
2016-03-21 20:19:13 -06:00

79 lines
1.9 KiB
Python

"""
DuckDuckGo (Web)
@website https://duckduckgo.com/
@provide-api yes (https://duckduckgo.com/api),
but not all results from search-site
@using-api no
@results HTML (using search portal)
@stable no (HTML can change)
@parse url, title, content
@todo rewrite to api
@todo language support
(the current used site does not support language-change)
"""
from urllib import urlencode
from lxml.html import fromstring
from searx.engines.xpath import extract_text
# engine dependent config
categories = ['general']
paging = True
language_support = True
# search-url
url = 'https://duckduckgo.com/html?{query}&s={offset}'
# specific xpath variables
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
def request(query, params):
offset = (params['pageno'] - 1) * 30
if params['language'] == 'all':
locale = 'en-us'
else:
locale = params['language'].replace('_', '-').lower()
params['url'] = url.format(
query=urlencode({'q': query, 'kl': locale}),
offset=offset)
return params
# get response from search-request
def response(resp):
results = []
doc = fromstring(resp.text)
# parse results
for r in doc.xpath(result_xpath):
try:
res_url = r.xpath(url_xpath)[-1]
except:
continue
if not res_url:
continue
title = extract_text(r.xpath(title_xpath))
content = extract_text(r.xpath(content_xpath))
# append result
results.append({'title': title,
'content': content,
'url': res_url})
# return results
return results