2013-10-26 00:22:20 +00:00
|
|
|
from lxml import html
|
2014-01-05 13:06:52 +00:00
|
|
|
from urllib import urlencode, unquote
|
2013-10-26 00:22:20 +00:00
|
|
|
from urlparse import urlparse, urljoin
|
|
|
|
from lxml.etree import _ElementStringResult
|
2014-01-23 10:08:08 +00:00
|
|
|
from searx.utils import html_to_text
|
2013-10-26 00:22:20 +00:00
|
|
|
|
2014-01-20 01:31:20 +00:00
|
|
|
search_url = None
|
|
|
|
url_xpath = None
|
2013-10-26 00:22:20 +00:00
|
|
|
content_xpath = None
|
2014-01-20 01:31:20 +00:00
|
|
|
title_xpath = None
|
2013-11-13 18:33:09 +00:00
|
|
|
suggestion_xpath = ''
|
2013-10-26 11:45:43 +00:00
|
|
|
results_xpath = ''
|
2013-10-26 00:22:20 +00:00
|
|
|
|
2014-01-20 01:31:20 +00:00
|
|
|
|
2014-01-05 13:06:52 +00:00
|
|
|
'''
|
|
|
|
if xpath_results is list, extract the text from each result and concat the list
|
2014-01-20 01:31:20 +00:00
|
|
|
if xpath_results is a xml element, extract all the text node from it
|
|
|
|
( text_content() method from lxml )
|
2014-01-05 13:06:52 +00:00
|
|
|
if xpath_results is a string element, then it's already done
|
|
|
|
'''
|
2014-01-20 01:31:20 +00:00
|
|
|
|
|
|
|
|
2014-01-05 13:06:52 +00:00
|
|
|
def extract_text(xpath_results):
|
2013-10-26 00:22:20 +00:00
|
|
|
if type(xpath_results) == list:
|
2014-01-05 13:06:52 +00:00
|
|
|
# it's list of result : concat everything using recursive call
|
2013-10-26 00:22:20 +00:00
|
|
|
if not len(xpath_results):
|
|
|
|
raise Exception('Empty url resultset')
|
2014-01-05 13:06:52 +00:00
|
|
|
result = ''
|
|
|
|
for e in xpath_results:
|
|
|
|
result = result + extract_text(e)
|
|
|
|
return result
|
|
|
|
elif type(xpath_results) == _ElementStringResult:
|
|
|
|
# it's a string
|
|
|
|
return ''.join(xpath_results)
|
2013-10-26 00:22:20 +00:00
|
|
|
else:
|
2014-01-05 13:06:52 +00:00
|
|
|
# it's a element
|
2014-01-23 10:08:08 +00:00
|
|
|
return html_to_text(xpath_results.text_content())
|
2014-01-05 13:06:52 +00:00
|
|
|
|
|
|
|
|
2014-01-30 01:32:58 +00:00
|
|
|
def extract_url(xpath_results, search_url=search_url):
|
2014-01-05 13:06:52 +00:00
|
|
|
url = extract_text(xpath_results)
|
|
|
|
|
|
|
|
if url.startswith('//'):
|
|
|
|
# add http or https to this kind of url //example.com/
|
|
|
|
parsed_search_url = urlparse(search_url)
|
|
|
|
url = parsed_search_url.scheme+url
|
|
|
|
elif url.startswith('/'):
|
|
|
|
# fix relative url to the search engine
|
|
|
|
url = urljoin(search_url, url)
|
|
|
|
|
|
|
|
# normalize url
|
|
|
|
url = normalize_url(url)
|
|
|
|
|
|
|
|
return url
|
|
|
|
|
|
|
|
|
|
|
|
def normalize_url(url):
|
2013-10-26 00:22:20 +00:00
|
|
|
parsed_url = urlparse(url)
|
2014-01-05 13:06:52 +00:00
|
|
|
|
|
|
|
# add a / at this end of the url if there is no path
|
2013-10-26 00:22:20 +00:00
|
|
|
if not parsed_url.netloc:
|
|
|
|
raise Exception('Cannot parse url')
|
2013-10-27 11:01:03 +00:00
|
|
|
if not parsed_url.path:
|
|
|
|
url += '/'
|
2014-01-05 13:06:52 +00:00
|
|
|
|
|
|
|
# FIXME : hack for yahoo
|
2014-01-20 01:31:20 +00:00
|
|
|
if parsed_url.hostname == 'search.yahoo.com'\
|
|
|
|
and parsed_url.path.startswith('/r'):
|
2014-01-05 13:06:52 +00:00
|
|
|
p = parsed_url.path
|
|
|
|
mark = p.find('/**')
|
|
|
|
if mark != -1:
|
|
|
|
return unquote(p[mark+3:]).decode('utf-8')
|
|
|
|
|
2013-10-26 00:22:20 +00:00
|
|
|
return url
|
|
|
|
|
2014-01-05 13:06:52 +00:00
|
|
|
|
2013-10-26 00:22:20 +00:00
|
|
|
def request(query, params):
|
|
|
|
query = urlencode({'q': query})[2:]
|
|
|
|
params['url'] = search_url.format(query=query)
|
|
|
|
params['query'] = query
|
|
|
|
return params
|
|
|
|
|
|
|
|
|
|
|
|
def response(resp):
|
|
|
|
results = []
|
|
|
|
dom = html.fromstring(resp.text)
|
2013-10-26 11:45:43 +00:00
|
|
|
if results_xpath:
|
|
|
|
for result in dom.xpath(results_xpath):
|
|
|
|
url = extract_url(result.xpath(url_xpath))
|
2014-01-20 01:31:20 +00:00
|
|
|
title = extract_text(result.xpath(title_xpath)[0])
|
2014-01-05 13:06:52 +00:00
|
|
|
content = extract_text(result.xpath(content_xpath)[0])
|
2013-10-26 11:45:43 +00:00
|
|
|
results.append({'url': url, 'title': title, 'content': content})
|
|
|
|
else:
|
2014-01-12 17:48:38 +00:00
|
|
|
for url, title, content in zip(
|
2014-01-20 01:31:20 +00:00
|
|
|
map(extract_url, dom.xpath(url_xpath)),
|
|
|
|
map(extract_text, dom.xpath(title_xpath)),
|
|
|
|
map(extract_text, dom.xpath(content_xpath))
|
|
|
|
):
|
2013-10-26 11:45:43 +00:00
|
|
|
results.append({'url': url, 'title': title, 'content': content})
|
|
|
|
|
2013-11-13 18:33:09 +00:00
|
|
|
if not suggestion_xpath:
|
|
|
|
return results
|
|
|
|
for suggestion in dom.xpath(suggestion_xpath):
|
2014-01-05 13:06:52 +00:00
|
|
|
results.append({'suggestion': extract_text(suggestion)})
|
2013-10-26 00:22:20 +00:00
|
|
|
return results
|