Ponysearch/searx/engines/xpath.py
a01200356 c3daa08537 [enh] Add onions category with Ahmia, Not Evil and Torch
Xpath engine and results template changed to account for the fact that
archive.org doesn't cache .onions, though some onion engines migth have
their own cache.

Disabled by default. Can be enabled by setting the SOCKS proxies to
wherever Tor is listening and setting using_tor_proxy as True.

Requires Tor and updating packages.

To avoid manually adding the timeout on each engine, you can set
extra_proxy_timeout to account for Tor's (or whatever proxy used) extra
time.
2020-10-25 17:59:05 -07:00

88 lines
3 KiB
Python

from lxml import html
from urllib.parse import urlencode
from searx.utils import extract_text, extract_url, eval_xpath
search_url = None
url_xpath = None
content_xpath = None
title_xpath = None
thumbnail_xpath = False
paging = False
suggestion_xpath = ''
results_xpath = ''
cached_xpath = ''
cached_url = ''
# parameters for engines with paging support
#
# number of results on each page
# (only needed if the site requires not a page number, but an offset)
page_size = 1
# number of the first page (usually 0 or 1)
first_page_num = 1
def request(query, params):
query = urlencode({'q': query})[2:]
fp = {'query': query}
if paging and search_url.find('{pageno}') >= 0:
fp['pageno'] = (params['pageno'] - 1) * page_size + first_page_num
params['url'] = search_url.format(**fp)
params['query'] = query
return params
def response(resp):
results = []
dom = html.fromstring(resp.text)
is_onion = True if 'onions' in categories else False
if results_xpath:
for result in eval_xpath(dom, results_xpath):
url = extract_url(eval_xpath(result, url_xpath), search_url)
title = extract_text(eval_xpath(result, title_xpath))
content = extract_text(eval_xpath(result, content_xpath))
tmp_result = {'url': url, 'title': title, 'content': content}
# add thumbnail if available
if thumbnail_xpath:
thumbnail_xpath_result = eval_xpath(result, thumbnail_xpath)
if len(thumbnail_xpath_result) > 0:
tmp_result['img_src'] = extract_url(thumbnail_xpath_result, search_url)
# add alternative cached url if available
if cached_xpath:
tmp_result['cached_url'] = cached_url + extract_text(result.xpath(cached_xpath))
if is_onion:
tmp_result['is_onion'] = True
results.append(tmp_result)
else:
if cached_xpath:
for url, title, content, cached in zip(
(extract_url(x, search_url) for
x in dom.xpath(url_xpath)),
map(extract_text, dom.xpath(title_xpath)),
map(extract_text, dom.xpath(content_xpath)),
map(extract_text, dom.xpath(cached_xpath))
):
results.append({'url': url, 'title': title, 'content': content,
'cached_url': cached_url + cached, 'is_onion': is_onion})
else:
for url, title, content in zip(
(extract_url(x, search_url) for
x in dom.xpath(url_xpath)),
map(extract_text, dom.xpath(title_xpath)),
map(extract_text, dom.xpath(content_xpath))
):
results.append({'url': url, 'title': title, 'content': content, 'is_onion': is_onion})
if not suggestion_xpath:
return results
for suggestion in eval_xpath(dom, suggestion_xpath):
results.append({'suggestion': extract_text(suggestion)})
return results