searxng/searx/autocomplete.py

198 lines
6.4 KiB
Python
Raw Normal View History

'''
searx is free software: you can redistribute it and/or modify
it under the terms of the GNU Affero General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
searx is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU Affero General Public License for more details.
You should have received a copy of the GNU Affero General Public License
along with searx. If not, see < http://www.gnu.org/licenses/ >.
(C) 2013- by Adam Tauber, <asciimoo@gmail.com>
'''
2014-03-29 15:30:49 +00:00
from lxml import etree
from json import loads
2014-03-29 16:04:33 +00:00
from urllib import urlencode
2015-04-09 22:59:25 +00:00
from searx import settings
from searx.languages import language_codes
from searx.engines import (
categories, engines, engine_shortcuts
)
2015-04-09 22:59:25 +00:00
from searx.poolrequests import get as http_get
def get(*args, **kwargs):
if 'timeout' not in kwargs:
kwargs['timeout'] = settings['outgoing']['request_timeout']
2015-04-09 22:59:25 +00:00
return http_get(*args, **kwargs)
def searx_bang(full_query):
'''check if the searchQuery contain a bang, and create fitting autocompleter results'''
# check if there is a query which can be parsed
if len(full_query.getSearchQuery()) == 0:
return []
results = []
# check if current query stats with !bang
2015-01-19 18:47:32 +00:00
first_char = full_query.getSearchQuery()[0]
if first_char == '!' or first_char == '?':
if len(full_query.getSearchQuery()) == 1:
# show some example queries
# TODO, check if engine is not avaliable
2015-01-19 18:47:32 +00:00
results.append(first_char + "images")
results.append(first_char + "wikipedia")
results.append(first_char + "osm")
else:
engine_query = full_query.getSearchQuery()[1:]
# check if query starts with categorie name
for categorie in categories:
if categorie.startswith(engine_query):
results.append(first_char + '{categorie}'.format(categorie=categorie))
# check if query starts with engine name
for engine in engines:
2015-01-10 18:55:21 +00:00
if engine.startswith(engine_query.replace('_', ' ')):
results.append(first_char + '{engine}'.format(engine=engine.replace(' ', '_')))
# check if query starts with engine shortcut
for engine_shortcut in engine_shortcuts:
if engine_shortcut.startswith(engine_query):
results.append(first_char + '{engine_shortcut}'.format(engine_shortcut=engine_shortcut))
# check if current query stats with :bang
2015-01-19 18:47:32 +00:00
elif first_char == ':':
if len(full_query.getSearchQuery()) == 1:
# show some example queries
results.append(":en")
results.append(":en_us")
results.append(":english")
results.append(":united_kingdom")
else:
engine_query = full_query.getSearchQuery()[1:]
for lc in language_codes:
lang_id, lang_name, country = map(str.lower, lc)
# check if query starts with language-id
if lang_id.startswith(engine_query):
if len(engine_query) <= 2:
results.append(':{lang_id}'.format(lang_id=lang_id.split('_')[0]))
else:
results.append(':{lang_id}'.format(lang_id=lang_id))
# check if query starts with language name
if lang_name.startswith(engine_query):
results.append(':{lang_name}'.format(lang_name=lang_name))
# check if query starts with country
2015-01-10 18:55:21 +00:00
if country.startswith(engine_query.replace('_', ' ')):
results.append(':{country}'.format(country=country.replace(' ', '_')))
# remove duplicates
result_set = set(results)
# remove results which are already contained in the query
for query_part in full_query.query_parts:
if query_part in result_set:
result_set.remove(query_part)
# convert result_set back to list
return list(result_set)
2014-03-29 15:30:49 +00:00
def dbpedia(query, lang):
# dbpedia autocompleter, no HTTPS
autocomplete_url = 'http://lookup.dbpedia.org/api/search.asmx/KeywordSearch?'
2014-03-29 15:30:49 +00:00
2016-01-18 11:47:31 +00:00
response = get(autocomplete_url + urlencode(dict(QueryString=query)))
2014-03-29 15:30:49 +00:00
results = []
if response.ok:
dom = etree.fromstring(response.content)
results = dom.xpath('//a:Result/a:Label//text()',
namespaces={'a': 'http://lookup.dbpedia.org/'})
return results
def duckduckgo(query, lang):
# duckduckgo autocompleter
2014-09-07 21:56:06 +00:00
url = 'https://ac.duckduckgo.com/ac/?{0}&type=list'
resp = loads(get(url.format(urlencode(dict(q=query)))).text)
if len(resp) > 1:
return resp[1]
return []
def google(query, lang):
2014-03-29 15:30:49 +00:00
# google autocompleter
autocomplete_url = 'https://suggestqueries.google.com/complete/search?client=toolbar&'
2014-03-29 15:30:49 +00:00
response = get(autocomplete_url + urlencode(dict(hl=lang, q=query)))
2014-03-29 15:30:49 +00:00
results = []
if response.ok:
2014-03-29 16:04:33 +00:00
dom = etree.fromstring(response.text)
2014-03-29 15:30:49 +00:00
results = dom.xpath('//suggestion/@data')
return results
def startpage(query, lang):
# startpage autocompleter
url = 'https://startpage.com/do/suggest?{query}'
resp = get(url.format(query=urlencode({'query': query}))).text.split('\n')
if len(resp) > 1:
return resp
return []
def qwant(query, lang):
2016-03-02 11:54:06 +00:00
# qwant autocompleter (additional parameter : lang=en_en&count=xxx )
url = 'https://api.qwant.com/api/suggest?{query}'
resp = get(url.format(query=urlencode({'q': query, 'lang': lang})))
2016-03-02 11:54:06 +00:00
results = []
if resp.ok:
data = loads(resp.text)
if data['status'] == 'success':
for item in data['data']['items']:
results.append(item['value'])
return results
def wikipedia(query, lang):
2014-03-29 15:30:49 +00:00
# wikipedia autocompleter
url = 'https://' + lang + '.wikipedia.org/w/api.php?action=opensearch&{0}&limit=10&namespace=0&format=json'
2014-03-29 15:30:49 +00:00
resp = loads(get(url.format(urlencode(dict(search=query)))).text)
if len(resp) > 1:
return resp[1]
return []
2014-03-29 15:30:49 +00:00
backends = {'dbpedia': dbpedia,
2014-09-07 21:56:06 +00:00
'duckduckgo': duckduckgo,
2014-03-29 15:30:49 +00:00
'google': google,
'startpage': startpage,
2016-03-02 11:54:06 +00:00
'qwant': qwant,
2014-03-29 15:30:49 +00:00
'wikipedia': wikipedia
}