Ponysearch/searx/engines/ahmia.py

81 lines
2.5 KiB
Python
Raw Permalink Normal View History

# SPDX-License-Identifier: AGPL-3.0-or-later
"""
Ahmia (Onions)
"""
from urllib.parse import urlencode, urlparse, parse_qs
from lxml.html import fromstring
from searx.engines.xpath import extract_url, extract_text, eval_xpath_list, eval_xpath
# about
about = {
"website": 'http://juhanurmihxlp77nkq76byazcldy2hlmovfu2epvl5ankdibsot4csyd.onion',
"wikidata_id": 'Q18693938',
"official_api_documentation": None,
"use_official_api": False,
"require_api_key": False,
"results": 'HTML',
}
# engine config
categories = ['onions']
paging = True
page_size = 10
# search url
search_url = 'http://juhanurmihxlp77nkq76byazcldy2hlmovfu2epvl5ankdibsot4csyd.onion/search/?{query}'
time_range_support = True
time_range_dict = {'day': 1, 'week': 7, 'month': 30}
# xpaths
results_xpath = '//li[@class="result"]'
url_xpath = './h4/a/@href'
title_xpath = './h4/a[1]'
content_xpath = './/p[1]'
correction_xpath = '//*[@id="didYouMean"]//a'
number_of_results_xpath = '//*[@id="totalResults"]'
def request(query, params):
params['url'] = search_url.format(query=urlencode({'q': query}))
if params['time_range'] in time_range_dict:
params['url'] += '&' + urlencode({'d': time_range_dict[params['time_range']]})
return params
def response(resp):
results = []
dom = fromstring(resp.text)
# trim results so there's not way too many at once
first_result_index = page_size * (resp.search_params.get('pageno', 1) - 1)
all_results = eval_xpath_list(dom, results_xpath)
trimmed_results = all_results[first_result_index : first_result_index + page_size]
# get results
for result in trimmed_results:
# remove ahmia url and extract the actual url for the result
raw_url = extract_url(eval_xpath_list(result, url_xpath, min_len=1), search_url)
cleaned_url = parse_qs(urlparse(raw_url).query).get('redirect_url', [''])[0]
title = extract_text(eval_xpath(result, title_xpath))
content = extract_text(eval_xpath(result, content_xpath))
results.append({'url': cleaned_url, 'title': title, 'content': content, 'is_onion': True})
# get spelling corrections
for correction in eval_xpath_list(dom, correction_xpath):
results.append({'correction': extract_text(correction)})
# get number of results
number_of_results = eval_xpath(dom, number_of_results_xpath)
if number_of_results:
try:
results.append({'number_of_results': int(extract_text(number_of_results))})
except: # pylint: disable=bare-except
pass
return results