searxng/searx/engines/duckduckgo.py

123 lines
3.9 KiB
Python
Raw Normal View History

"""
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
"""
2013-10-23 21:55:37 +00:00
from urllib import urlencode
2014-03-21 15:33:17 +00:00
from lxml.html import fromstring
2015-02-02 16:55:39 +00:00
from searx.engines.xpath import extract_text
2013-10-14 21:09:13 +00:00
# engine dependent config
categories = ['general']
paging = True
language_support = True
supported_languages = ["es-AR", "en-AU", "de-AT", "fr-BE", "nl-BE", "pt-BR", "bg-BG", "en-CA", "fr-CA", "ca-CT",
"es-CL", "zh-CN", "es-CO", "hr-HR", "cs-CZ", "da-DK", "et-EE", "fi-FI", "fr-FR", "de-DE",
"el-GR", "tzh-HK", "hu-HU", "en-IN", "id-ID", "en-ID", "en-IE", "he-IL", "it-IT", "jp-JP",
"kr-KR", "es-XL", "lv-LV", "lt-LT", "ms-MY", "en-MY", "es-MX", "nl-NL", "en-NZ", "no-NO",
"es-PE", "en-PH", "tl-PH", "pl-PL", "pt-PT", "ro-RO", "ru-RU", "ar-XA", "en-XA", "en-SG",
"sk-SK", "sl-SL", "en-ZA", "es-ES", "ca-ES", "sv-SE", "de-CH", "fr-CH", "it-CH", "tzh-TW",
"th-TH", "tr-TR", "uk-UA", "en-UK", "en-US", "es-US", "vi-VN"]
2016-07-18 14:15:37 +00:00
time_range_support = True
2013-10-14 21:09:13 +00:00
# search-url
url = 'https://duckduckgo.com/html?{query}&s={offset}'
2016-07-18 14:15:37 +00:00
time_range_url = '&df={range}'
time_range_dict = {'day': 'd',
'week': 'w',
'month': 'm'}
# 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"]'
2014-03-29 15:38:45 +00:00
# do search-request
2013-10-14 21:09:13 +00:00
def request(query, params):
if params['time_range'] and params['time_range'] not in time_range_dict:
return params
2014-01-30 00:03:19 +00:00
offset = (params['pageno'] - 1) * 30
# custom fixes for languages
if params['language'] == 'all':
locale = None
elif params['language'][:2] == 'ja':
locale = 'jp-jp'
elif params['language'] == 'zh-TW':
locale = 'tw-tzh'
elif params['language'] == 'zh-HK':
locale = 'hk-tzh'
elif params['language'][-2:] == 'SA':
locale = 'xa' + params['language'].split('-')[0]
elif params['language'][-2:] == 'GB':
locale = 'uk' + params['language'].split('-')[0]
elif params['language'] == 'es-419':
locale = 'xl-es'
else:
locale = params['language'].split('-')
if len(locale) == 2:
# country code goes first
locale = locale[1].lower() + '-' + locale[0].lower()
else:
# tries to get a country code from language
locale = locale[0].lower()
2016-10-30 02:04:01 +00:00
for lc in supported_languages:
lc = lc.split('-')
if locale == lc[0]:
locale = lc[1].lower() + '-' + lc[0].lower()
break
if locale:
params['url'] = url.format(
query=urlencode({'q': query, 'kl': locale}), offset=offset)
else:
params['url'] = url.format(
query=urlencode({'q': query}), offset=offset)
2016-07-25 22:22:05 +00:00
if params['time_range'] in time_range_dict:
2016-07-18 14:15:37 +00:00
params['url'] += time_range_url.format(range=time_range_dict[params['time_range']])
2013-10-14 21:09:13 +00:00
return params
# get response from search-request
2013-10-14 21:09:13 +00:00
def response(resp):
2013-10-15 17:11:43 +00:00
results = []
2014-03-21 15:33:17 +00:00
doc = fromstring(resp.text)
2014-09-02 16:12:42 +00:00
# parse results
2014-03-21 15:33:17 +00:00
for r in doc.xpath(result_xpath):
2014-03-21 17:17:13 +00:00
try:
res_url = r.xpath(url_xpath)[-1]
except:
continue
2014-03-21 15:33:17 +00:00
if not res_url:
2013-10-15 17:11:43 +00:00
continue
2015-02-02 16:55:39 +00:00
title = extract_text(r.xpath(title_xpath))
content = extract_text(r.xpath(content_xpath))
# append result
2014-03-21 15:33:17 +00:00
results.append({'title': title,
'content': content,
2015-09-07 21:13:04 +00:00
'url': res_url})
2014-03-21 15:33:17 +00:00
# return results
2013-10-15 17:11:43 +00:00
return results