git.iko.soy/searx
on master help sign in
implement query parser and use it inside autocompletion
510aba5Thomas Pointhuber
searx/query.pynew file+125
compare with master
unifiedsplit
@@ -0,0 +1,125 @@
1+#!/usr/bin/env python
2+
3+'''
4+searx is free software: you can redistribute it and/or modify
5+it under the terms of the GNU Affero General Public License as published by
6+the Free Software Foundation, either version 3 of the License, or
7+(at your option) any later version.
8+
9+searx is distributed in the hope that it will be useful,
10+but WITHOUT ANY WARRANTY; without even the implied warranty of
11+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12+GNU Affero General Public License for more details.
13+
14+You should have received a copy of the GNU Affero General Public License
15+along with searx. If not, see < http://www.gnu.org/licenses/ >.
16+
17+(C) 2014 by Thomas Pointhuber, <thomas.pointhuber@gmx.at>
18+'''
19+
20+from searx.languages import language_codes
21+from searx.engines import (
22+ categories, engines, engine_shortcuts
23+)
24+import string
25+import re
26+
27+
28+class Query(object):
29+ """parse query"""
30+
31+ def __init__(self, query, blocked_engines):
32+ self.query = query
33+ self.blocked_engines = []
34+
35+ if blocked_engines:
36+ self.blocked_engines = blocked_engines
37+
38+ self.query_parts = []
39+ self.engines = []
40+ self.languages = []
41+
42+ def parse_query(self):
43+ self.query_parts = []
44+
45+ # split query, including whitespaces
46+ raw_query_parts = re.split(r'(\s+)', self.query)
47+
48+ parse_next = True
49+
50+ for query_part in raw_query_parts:
51+ if not parse_next:
52+ self.query_parts[-1] += query_part
53+ continue
54+
55+ parse_next = False
56+
57+ # part does only contain spaces, skip
58+ if query_part.isspace():
59+ parse_next = True
60+ self.query_parts.append(query_part)
61+ continue
62+
63+ # this force a language
64+ if query_part[0] == ':':
65+ lang = query_part[1:].lower()
66+
67+ # check if any language-code is equal with declared language-codes
68+ for lc in language_codes:
69+ lang_id, lang_name, country = map(str.lower, lc)
70+
71+ # if correct language-code is found, set it as new search-language
72+ if lang == lang_id\
73+ or lang_id.startswith(lang)\
74+ or lang == lang_name\
75+ or lang == country:
76+ parse_next = True
77+ self.languages.append(lang)
78+ break
79+
80+ # this force a engine or category
81+ if query_part[0] == '!':
82+ prefix = query_part[1:].replace('_', ' ')
83+
84+ # check if prefix is equal with engine shortcut
85+ if prefix in engine_shortcuts\
86+ and not engine_shortcuts[prefix] in self.blocked_engines:
87+ parse_next = True
88+ self.engines.append({'category': 'none',
89+ 'name': engine_shortcuts[prefix]})
90+
91+ # check if prefix is equal with engine name
92+ elif prefix in engines\
93+ and not prefix in self.blocked_engines:
94+ parse_next = True
95+ self.engines.append({'category': 'none',
96+ 'name': prefix})
97+
98+ # check if prefix is equal with categorie name
99+ elif prefix in categories:
100+ # using all engines for that search, which are declared under that categorie name
101+ parse_next = True
102+ self.engines.extend({'category': prefix,
103+ 'name': engine.name}
104+ for engine in categories[prefix]
105+ if not engine in self.blocked_engines)
106+
107+ # append query part to query_part list
108+ self.query_parts.append(query_part)
109+
110+ def changeSearchQuery(self, search_query):
111+ if len(self.query_parts):
112+ self.query_parts[-1] = search_query
113+ else:
114+ self.query_parts.append(search_query)
115+
116+ def getSearchQuery(self):
117+ if len(self.query_parts):
118+ return self.query_parts[-1]
119+ else:
120+ return ''
121+
122+ def getFullQuery(self):
123+ # get full querry including whitespaces
124+ return string.join(self.query_parts, '')
125+
searx/webapp.py+27−3
@@ -47,6 +47,7 @@ from searx.utils import (
47from searx.https_rewrite import https_rules
47from searx.https_rewrite import https_rules
48from searx.languages import language_codes
48from searx.languages import language_codes
49from searx.search import Search
49from searx.search import Search
50+from searx.query import Query
50from searx.autocomplete import backends as autocomplete_backends
51from searx.autocomplete import backends as autocomplete_backends
51
52
52
53
@@ -308,23 +309,46 @@ def autocompleter():
308 """Return autocompleter results"""
309 """Return autocompleter results"""
309 request_data = {}
310 request_data = {}
310
311
312+ # select request method
311 if request.method == 'POST':
313 if request.method == 'POST':
312 request_data = request.form
314 request_data = request.form
313 else:
315 else:
314 request_data = request.args
316 request_data = request.args
315
317
318+ # set blocked engines
319+ if request.cookies.get('blocked_engines'):
320+ blocked_engines = request.cookies['blocked_engines'].split(',') # noqa
321+ else:
322+ blocked_engines = []
323+
324+ # parse query
316− query = request_data.get('q', '').encode('utf-8')
325+ query = Query(request_data.get('q', '').encode('utf-8'), blocked_engines)
326+ query.parse_query()
317
327
328+ # check if search query is set
318− if not query:
329+ if not query.getSearchQuery():
319 return
330 return
320
331
332+ # run autocompleter
321 completer = autocomplete_backends.get(request.cookies.get('autocomplete'))
333 completer = autocomplete_backends.get(request.cookies.get('autocomplete'))
322
334
335+ # check if valid autocompleter is selected
323 if not completer:
336 if not completer:
324 return
337 return
325
338
339+ # run autocompletion
326− results = completer(query)
340+ raw_results = completer(query.getSearchQuery())
341+
342+ # parse results (write :language and !engine back to result string)
343+ results = []
344+ for result in raw_results:
345+ result_query = query
346+ result_query.changeSearchQuery(result)
347+
348+ # add parsed result
349+ results.append(result_query.getFullQuery())
327
350
351+ # return autocompleter results
328 if request_data.get('format') == 'x-suggestions':
352 if request_data.get('format') == 'x-suggestions':
329 return Response(json.dumps([query, results]),
353 return Response(json.dumps([query, results]),
330 mimetype='application/json')
354 mimetype='application/json')