on master help sign in
[enh] introduce private engines

This PR adds a new setting to engines named `tokens`. It expects a list of tokens which lets searx validate if the request should be accepted or not.

commit99435381a84072b110c32004b2fb778af9b96f77 authorNoémi Ványi <sitbackandwait@gmail.com> dateSat 1 Feb 2020, 11:01 (7 years ago) parentf9c7a67
10 files changed+161−28
searx/__init__.py+1 searx/engines/__init__.py+7−2 searx/engines/dummy-offline.py+12 searx/engines/genius.py+1 searx/preferences.py+44 searx/query.py+3−1 searx/search.py+4−1 searx/templates/oscar/preferences.html+6 searx/webapp.py+12−5 tests/unit/test_search.py+71−19
searx/__init__.py+1
@@ -38,6 +38,7 @@ def check_settings_yml(file_name):
3838 else:
3939 return None
4040
41+
4142# find location of settings.yml
4243if 'SEARX_SETTINGS_PATH' in environ:
4344 # if possible set path to settings using the
searx/engines/__init__.py+7−2
@@ -54,7 +54,8 @@ engine_default_args = {'paging': False,
5454 'suspend_end_time': 0,
5555 'continuous_errors': 0,
5656 'time_range_support': False,
57− 'offline': False}
57+ 'offline': False,
58+ 'tokens': []}
5859
5960
6061def load_engine(engine_data):
@@ -160,7 +161,7 @@ def to_percentage(stats, maxvalue):
160161 return stats
161162
162163
163−def get_engines_stats():
164+def get_engines_stats(preferences):
164165 # TODO refactor
165166 pageloads = []
166167 engine_times = []
@@ -171,8 +172,12 @@ def get_engines_stats():
171172
172173 max_pageload = max_engine_times = max_results = max_score = max_errors = max_score_per_result = 0 # noqa
173174 for engine in engines.values():
175+ if not preferences.validate_token(engine):
176+ continue
177+
174178 if engine.stats['search_count'] == 0:
175179 continue
180+
176181 results_num = \
177182 engine.stats['result_count'] / float(engine.stats['search_count'])
178183
searx/engines/dummy-offline.pynew file+12
@@ -0,0 +1,12 @@
1+"""
2+ Dummy Offline
3+
4+ @results one result
5+ @stable yes
6+"""
7+
8+
9+def search(query, request_params):
10+ return [{
11+ 'result': 'this is what you get',
12+ }]
searx/engines/genius.py+1
@@ -72,6 +72,7 @@ def parse_album(hit):
7272 result.update({'content': 'Released: {}'.format(year)})
7373 return result
7474
75+
7576parse = {'lyric': parse_lyric, 'song': parse_lyric, 'artist': parse_artist, 'album': parse_album}
7677
7778
searx/preferences.py+44
@@ -104,6 +104,31 @@ class MultipleChoiceSetting(EnumStringSetting):
104104 resp.set_cookie(name, ','.join(self.value), max_age=COOKIE_MAX_AGE)
105105
106106
107+class SetSetting(Setting):
108+ def _post_init(self):
109+ if not hasattr(self, 'values'):
110+ self.values = set()
111+
112+ def get_value(self):
113+ return ','.join(self.values)
114+
115+ def parse(self, data):
116+ if data == '':
117+ self.values = set()
118+ return
119+
120+ elements = data.split(',')
121+ for element in elements:
122+ self.values.add(element)
123+
124+ def parse_form(self, data):
125+ elements = data.split(',')
126+ self.values = set(elements)
127+
128+ def save(self, name, resp):
129+ resp.set_cookie(name, ','.join(self.values), max_age=COOKIE_MAX_AGE)
130+
131+
107132class SearchLanguageSetting(EnumStringSetting):
108133 """Available choices may change, so user's value may not be in choices anymore"""
109134
@@ -272,6 +297,7 @@ class Preferences(object):
272297
273298 self.engines = EnginesSetting('engines', choices=engines)
274299 self.plugins = PluginsSetting('plugins', choices=plugins)
300+ self.tokens = SetSetting('tokens')
275301 self.unknown_params = {}
276302
277303 def get_as_url_params(self):
@@ -288,6 +314,8 @@ class Preferences(object):
288314 settings_kv['disabled_plugins'] = ','.join(self.plugins.disabled)
289315 settings_kv['enabled_plugins'] = ','.join(self.plugins.enabled)
290316
317+ settings_kv['tokens'] = ','.join(self.tokens.values)
318+
291319 return urlsafe_b64encode(compress(urlencode(settings_kv).encode('utf-8'))).decode('utf-8')
292320
293321 def parse_encoded_data(self, input_data):
@@ -307,6 +335,8 @@ class Preferences(object):
307335 elif user_setting_name == 'disabled_plugins':
308336 self.plugins.parse_cookie((input_data.get('disabled_plugins', ''),
309337 input_data.get('enabled_plugins', '')))
338+ elif user_setting_name == 'tokens':
339+ self.tokens.parse(user_setting)
310340 elif not any(user_setting_name.startswith(x) for x in [
311341 'enabled_',
312342 'disabled_',
@@ -328,6 +358,8 @@ class Preferences(object):
328358 enabled_categories.append(user_setting_name[len('category_'):])
329359 elif user_setting_name.startswith('plugin_'):
330360 disabled_plugins.append(user_setting_name)
361+ elif user_setting_name == 'tokens':
362+ self.tokens.parse_form(user_setting)
331363 else:
332364 self.unknown_params[user_setting_name] = user_setting
333365 self.key_value_settings['categories'].parse_form(enabled_categories)
@@ -346,6 +378,18 @@ class Preferences(object):
346378 user_setting.save(user_setting_name, resp)
347379 self.engines.save(resp)
348380 self.plugins.save(resp)
381+ self.tokens.save('tokens', resp)
349382 for k, v in self.unknown_params.items():
350383 resp.set_cookie(k, v, max_age=COOKIE_MAX_AGE)
351384 return resp
385+
386+ def validate_token(self, engine):
387+ valid = True
388+ if hasattr(engine, 'tokens') and engine.tokens:
389+ valid = False
390+ for token in self.tokens.values:
391+ if token in engine.tokens:
392+ valid = True
393+ break
394+
395+ return valid
searx/query.py+3−1
@@ -177,7 +177,8 @@ class RawTextQuery(object):
177177class SearchQuery(object):
178178 """container for all the search parameters (query, language, etc...)"""
179179
180− def __init__(self, query, engines, categories, lang, safesearch, pageno, time_range, timeout_limit=None):
180+ def __init__(self, query, engines, categories, lang, safesearch, pageno, time_range,
181+ timeout_limit=None, preferences=None):
181182 self.query = query.encode('utf-8')
182183 self.engines = engines
183184 self.categories = categories
@@ -186,6 +187,7 @@ class SearchQuery(object):
186187 self.pageno = pageno
187188 self.time_range = None if time_range in ('', 'None', None) else time_range
188189 self.timeout_limit = timeout_limit
190+ self.preferences = preferences
189191
190192 def __str__(self):
191193 return str(self.query) + ";" + str(self.engines)
searx/search.py+4−1
@@ -407,7 +407,7 @@ def get_search_query_from_webapp(preferences, form):
407407
408408 return (SearchQuery(query, query_engines, query_categories,
409409 query_lang, query_safesearch, query_pageno,
410− query_time_range, query_timeout),
410+ query_time_range, query_timeout, preferences),
411411 raw_text_query)
412412
413413
@@ -459,6 +459,9 @@ class Search(object):
459459
460460 engine = engines[selected_engine['name']]
461461
462+ if not search_query.preferences.validate_token(engine):
463+ continue
464+
462465 # skip suspended engines
463466 if engine.suspend_end_time >= time():
464467 logger.debug('Engine currently suspended: %s', selected_engine['name'])
searx/templates/oscar/preferences.html+6
@@ -131,6 +131,12 @@
131131 {% endfor %}
132132 </select>
133133 {{ preferences_item_footer(info, label, rtl) }}
134+
135+ {% set label = _('Engine tokens') %}
136+ {% set info = _('Access tokens for private engines') %}
137+ {{ preferences_item_header(info, label, rtl) }}
138+ <input class="form-control" id='tokens' name='tokens' value='{{ preferences.tokens.get_value() }}'/>
139+ {{ preferences_item_footer(info, label, rtl) }}
134140 </div>
135141 </fieldset>
136142 </div>
searx/webapp.py+12−5
@@ -731,8 +731,13 @@ def preferences():
731731 # stats for preferences page
732732 stats = {}
733733
734+ engines_by_category = {}
734735 for c in categories:
736+ engines_by_category[c] = []
735737 for e in categories[c]:
738+ if not request.preferences.validate_token(e):
739+ continue
740+
736741 stats[e.name] = {'time': None,
737742 'warn_timeout': False,
738743 'warn_time': False}
@@ -740,9 +745,11 @@ def preferences():
740745 stats[e.name]['warn_timeout'] = True
741746 stats[e.name]['supports_selected_language'] = _is_selected_language_supported(e, request.preferences)
742747
748+ engines_by_category[c].append(e)
749+
743750 # get first element [0], the engine time,
744751 # and then the second element [1] : the time (the first one is the label)
745− for engine_stat in get_engines_stats()[0][1]:
752+ for engine_stat in get_engines_stats(request.preferences)[0][1]:
746753 stats[engine_stat.get('name')]['time'] = round(engine_stat.get('avg'), 3)
747754 if engine_stat.get('avg') > settings['outgoing']['request_timeout']:
748755 stats[engine_stat.get('name')]['warn_time'] = True
@@ -752,7 +759,7 @@ def preferences():
752759 locales=settings['locales'],
753760 current_locale=get_locale(),
754761 image_proxy=image_proxy,
755− engines_by_category=categories,
762+ engines_by_category=engines_by_category,
756763 stats=stats,
757764 answerers=[{'info': a.self_info(), 'keywords': a.keywords} for a in answerers],
758765 disabled_engines=disabled_engines,
@@ -828,7 +835,7 @@ def image_proxy():
828835@app.route('/stats', methods=['GET'])
829836def stats():
830837 """Render engine statistics page."""
831− stats = get_engines_stats()
838+ stats = get_engines_stats(request.preferences)
832839 return render(
833840 'stats.html',
834841 stats=stats,
@@ -891,7 +898,7 @@ def clear_cookies():
891898@app.route('/config')
892899def config():
893900 return jsonify({'categories': list(categories.keys()),
894− 'engines': [{'name': engine_name,
901+ 'engines': [{'name': name,
895902 'categories': engine.categories,
896903 'shortcut': engine.shortcut,
897904 'enabled': not engine.disabled,
@@ -904,7 +911,7 @@ def config():
904911 'safesearch': engine.safesearch,
905912 'time_range_support': engine.time_range_support,
906913 'timeout': engine.timeout}
907− for engine_name, engine in engines.items()],
914+ for name, engine in engines.items() if request.preferences.validate_token(engine)],
908915 'plugins': [{'name': plugin.name,
909916 'enabled': plugin.default_on}
910917 for plugin in plugins],
tests/unit/test_search.py+71−19
@@ -1,60 +1,112 @@
11# -*- coding: utf-8 -*-
22
33from searx.testing import SearxTestCase
4+from searx.preferences import Preferences
5+from searx.engines import engines
46
5−import searx.preferences
67import searx.search
7−import searx.engines
8+
9+
10+SAFESEARCH = 0
11+PAGENO = 1
12+PUBLIC_ENGINE_NAME = 'general dummy'
13+PRIVATE_ENGINE_NAME = 'general private offline'
14+TEST_ENGINES = [
15+ {
16+ 'name': PUBLIC_ENGINE_NAME,
17+ 'engine': 'dummy',
18+ 'categories': 'general',
19+ 'shortcut': 'gd',
20+ 'timeout': 3.0,
21+ 'tokens': [],
22+ },
23+ {
24+ 'name': PRIVATE_ENGINE_NAME,
25+ 'engine': 'dummy-offline',
26+ 'categories': 'general',
27+ 'shortcut': 'do',
28+ 'timeout': 3.0,
29+ 'offline': True,
30+ 'tokens': ['my-token'],
31+ },
32+]
833
934
1035class SearchTestCase(SearxTestCase):
1136
1237 @classmethod
1338 def setUpClass(cls):
14− searx.engines.initialize_engines([{
15− 'name': 'general dummy',
16− 'engine': 'dummy',
17− 'categories': 'general',
18− 'shortcut': 'gd',
19− 'timeout': 3.0
20− }])
39+ searx.engines.initialize_engines(TEST_ENGINES)
2140
2241 def test_timeout_simple(self):
2342 searx.search.max_request_timeout = None
24− search_query = searx.query.SearchQuery('test', [{'category': 'general', 'name': 'general dummy'}],
25− ['general'], 'en-US', 0, 1, None, None)
43+ search_query = searx.query.SearchQuery('test', [{'category': 'general', 'name': PUBLIC_ENGINE_NAME}],
44+ ['general'], 'en-US', SAFESEARCH, PAGENO, None, None,
45+ preferences=Preferences(['oscar'], ['general'], engines, []))
2646 search = searx.search.Search(search_query)
2747 search.search()
2848 self.assertEquals(search.actual_timeout, 3.0)
2949
3050 def test_timeout_query_above_default_nomax(self):
3151 searx.search.max_request_timeout = None
32− search_query = searx.query.SearchQuery('test', [{'category': 'general', 'name': 'general dummy'}],
33− ['general'], 'en-US', 0, 1, None, 5.0)
52+ search_query = searx.query.SearchQuery('test', [{'category': 'general', 'name': PUBLIC_ENGINE_NAME}],
53+ ['general'], 'en-US', SAFESEARCH, PAGENO, None, 5.0,
54+ preferences=Preferences(['oscar'], ['general'], engines, []))
3455 search = searx.search.Search(search_query)
3556 search.search()
3657 self.assertEquals(search.actual_timeout, 3.0)
3758
3859 def test_timeout_query_below_default_nomax(self):
3960 searx.search.max_request_timeout = None
40− search_query = searx.query.SearchQuery('test', [{'category': 'general', 'name': 'general dummy'}],
41− ['general'], 'en-US', 0, 1, None, 1.0)
61+ search_query = searx.query.SearchQuery('test', [{'category': 'general', 'name': PUBLIC_ENGINE_NAME}],
62+ ['general'], 'en-US', SAFESEARCH, PAGENO, None, 1.0,
63+ preferences=Preferences(['oscar'], ['general'], engines, []))
4264 search = searx.search.Search(search_query)
4365 search.search()
4466 self.assertEquals(search.actual_timeout, 1.0)
4567
4668 def test_timeout_query_below_max(self):
4769 searx.search.max_request_timeout = 10.0
48− search_query = searx.query.SearchQuery('test', [{'category': 'general', 'name': 'general dummy'}],
49− ['general'], 'en-US', 0, 1, None, 5.0)
70+ search_query = searx.query.SearchQuery('test', [{'category': 'general', 'name': PUBLIC_ENGINE_NAME}],
71+ ['general'], 'en-US', SAFESEARCH, PAGENO, None, 5.0,
72+ preferences=Preferences(['oscar'], ['general'], engines, []))
5073 search = searx.search.Search(search_query)
5174 search.search()
5275 self.assertEquals(search.actual_timeout, 5.0)
5376
5477 def test_timeout_query_above_max(self):
5578 searx.search.max_request_timeout = 10.0
56− search_query = searx.query.SearchQuery('test', [{'category': 'general', 'name': 'general dummy'}],
57− ['general'], 'en-US', 0, 1, None, 15.0)
79+ search_query = searx.query.SearchQuery('test', [{'category': 'general', 'name': PUBLIC_ENGINE_NAME}],
80+ ['general'], 'en-US', SAFESEARCH, PAGENO, None, 15.0,
81+ preferences=Preferences(['oscar'], ['general'], engines, []))
5882 search = searx.search.Search(search_query)
5983 search.search()
6084 self.assertEquals(search.actual_timeout, 10.0)
85+
86+ def test_query_private_engine_without_token(self):
87+ search_query = searx.query.SearchQuery('test', [{'category': 'general', 'name': PRIVATE_ENGINE_NAME}],
88+ ['general'], 'en-US', SAFESEARCH, PAGENO, None, 2.0,
89+ preferences=Preferences(['oscar'], ['general'], engines, []))
90+ search = searx.search.Search(search_query)
91+ results = search.search()
92+ self.assertEquals(results.results_length(), 0)
93+
94+ def test_query_private_engine_with_incorrect_token(self):
95+ preferences_with_tokens = Preferences(['oscar'], ['general'], engines, [])
96+ preferences_with_tokens.parse_dict({'tokens': 'bad-token'})
97+ search_query = searx.query.SearchQuery('test', [{'category': 'general', 'name': PRIVATE_ENGINE_NAME}],
98+ ['general'], 'en-US', SAFESEARCH, PAGENO, None, 2.0,
99+ preferences=preferences_with_tokens)
100+ search = searx.search.Search(search_query)
101+ results = search.search()
102+ self.assertEquals(results.results_length(), 0)
103+
104+ def test_query_private_engine_with_correct_token(self):
105+ preferences_with_tokens = Preferences(['oscar'], ['general'], engines, [])
106+ preferences_with_tokens.parse_dict({'tokens': 'my-token'})
107+ search_query = searx.query.SearchQuery('test', [{'category': 'general', 'name': PRIVATE_ENGINE_NAME}],
108+ ['general'], 'en-US', SAFESEARCH, PAGENO, None, 2.0,
109+ preferences=preferences_with_tokens)
110+ search = searx.search.Search(search_query)
111+ results = search.search()
112+ self.assertEquals(results.results_length(), 1)