searxng/searx/engines/__init__.py

120 lines
4.5 KiB
Python
Raw Normal View History

2013-10-14 21:09:13 +00:00
2013-10-16 22:32:32 +00:00
'''
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>
'''
2013-10-14 21:09:13 +00:00
from os.path import realpath, dirname, splitext, join
from os import listdir
from imp import load_source
2013-10-15 16:19:06 +00:00
import grequests
2013-10-15 21:20:26 +00:00
from itertools import izip_longest, chain
from operator import itemgetter
2013-10-19 15:36:44 +00:00
from urlparse import urlparse
2013-10-19 17:01:06 +00:00
from searx import settings
2013-10-14 21:09:13 +00:00
engine_dir = dirname(realpath(__file__))
2013-10-15 17:11:43 +00:00
engines = {}
2013-10-14 21:09:13 +00:00
2013-10-17 19:06:28 +00:00
categories = {'general': []}
2013-10-14 21:09:13 +00:00
for filename in listdir(engine_dir):
if filename.startswith('_') or not filename.endswith('.py'):
continue
2013-10-19 17:01:06 +00:00
modname = splitext(filename)[0]
if modname in settings.blacklist:
continue
2013-10-14 21:09:13 +00:00
filepath = join(engine_dir, filename)
2013-10-15 16:19:06 +00:00
engine = load_source(modname, filepath)
2013-10-17 19:06:28 +00:00
engine.name = modname
2013-10-15 16:19:06 +00:00
if not hasattr(engine, 'request') or not hasattr(engine, 'response'):
continue
2013-10-15 17:11:43 +00:00
engines[modname] = engine
2013-10-17 19:06:28 +00:00
if not hasattr(engine, 'categories'):
categories['general'].append(engine)
else:
for category_name in engine.categories:
categories.setdefault(category_name, []).append(engine)
2013-10-15 16:19:06 +00:00
def default_request_params():
2013-10-19 20:34:46 +00:00
return {'method': 'GET', 'headers': {}, 'data': {}, 'url': '', 'cookies': {}}
2013-10-15 16:19:06 +00:00
2013-10-15 17:11:43 +00:00
def make_callback(engine_name, results, callback):
2013-10-15 16:19:06 +00:00
def process_callback(response, **kwargs):
2013-10-15 21:20:26 +00:00
cb_res = []
2013-10-15 17:11:43 +00:00
for result in callback(response):
result['engine'] = engine_name
2013-10-15 21:20:26 +00:00
cb_res.append(result)
results[engine_name] = cb_res
2013-10-15 16:19:06 +00:00
return process_callback
2013-10-15 20:18:08 +00:00
def search(query, request, selected_engines):
2013-10-15 16:19:06 +00:00
global engines
requests = []
2013-10-15 21:20:26 +00:00
results = {}
2013-10-15 16:19:06 +00:00
user_agent = request.headers.get('User-Agent', '')
2013-10-15 17:11:43 +00:00
for ename, engine in engines.items():
2013-10-15 20:18:08 +00:00
if ename not in selected_engines:
continue
2013-10-15 16:19:06 +00:00
headers = default_request_params()
headers['User-Agent'] = user_agent
request_params = engine.request(query, headers)
2013-10-15 17:11:43 +00:00
callback = make_callback(ename, results, engine.response)
2013-10-15 16:19:06 +00:00
if request_params['method'] == 'GET':
req = grequests.get(request_params['url']
,headers=headers
,hooks=dict(response=callback)
2013-10-19 20:34:46 +00:00
,cookies = request_params['cookies']
2013-10-15 16:19:06 +00:00
)
else:
req = grequests.post(request_params['url']
,data=request_params['data']
,headers=headers
,hooks=dict(response=callback)
2013-10-19 20:34:46 +00:00
,cookies = request_params['cookies']
2013-10-15 16:19:06 +00:00
)
requests.append(req)
grequests.map(requests)
flat_res = list(filter(None, chain(*izip_longest(*results.values()))))
flat_len = len(flat_res)
results = []
# deduplication + scoring
for i,res in enumerate(flat_res):
2013-10-19 15:36:44 +00:00
res['parsed_url'] = urlparse(res['url'])
2013-10-19 17:04:46 +00:00
score = (flat_len - i)*settings.weights.get(res['engine'], 1)
duplicated = False
for new_res in results:
2013-10-19 15:36:44 +00:00
if res['parsed_url'].netloc == new_res['parsed_url'].netloc and\
2013-10-19 18:11:56 +00:00
res['parsed_url'].path == new_res['parsed_url'].path and\
2013-10-19 18:44:35 +00:00
res['parsed_url'].query == new_res['parsed_url'].query:
duplicated = new_res
break
if duplicated:
2013-10-18 06:24:58 +00:00
if len(res.get('content', '')) > len(duplicated.get('content', '')):
duplicated['content'] = res['content']
duplicated['score'] += score
duplicated['engine'] += ', '+res['engine']
2013-10-19 15:36:44 +00:00
if duplicated['parsed_url'].scheme == 'https':
continue
elif res['parsed_url'].scheme == 'https':
duplicated['parsed_url'].scheme == 'https'
duplicated['url'] = duplicated['parsed_url'].geturl()
else:
res['score'] = score
results.append(res)
return sorted(results, key=itemgetter('score'), reverse=True)