searxng/searx/search.py

433 lines
15 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>
'''
import gc
import threading
from thread import start_new_thread
2014-12-14 00:18:01 +00:00
from time import time
from uuid import uuid4
import requests.exceptions
import searx.poolrequests as requests_lib
from searx.engines import (
2014-10-19 10:41:04 +00:00
categories, engines
)
from searx.answerers import ask
from searx.utils import gen_useragent
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
from searx.query import RawTextQuery, SearchQuery
from searx.results import ResultContainer
2015-01-09 03:13:05 +00:00
from searx import logger
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
from searx.plugins import plugins
from searx.languages import language_codes
from searx.exceptions import SearxParameterException
2014-07-07 11:59:27 +00:00
2015-01-09 03:13:05 +00:00
logger = logger.getChild('search')
2014-07-07 11:59:27 +00:00
number_of_searches = 0
language_code_set = set(l[0].lower() for l in language_codes)
language_code_set.add('all')
2014-07-07 11:59:27 +00:00
def send_http_request(engine, request_params, start_time, timeout_limit):
# for page_load_time stats
time_before_request = time()
# create dictionary which contain all
# informations about the request
request_args = dict(
headers=request_params['headers'],
cookies=request_params['cookies'],
timeout=timeout_limit,
verify=request_params['verify']
)
# specific type of request (GET or POST)
if request_params['method'] == 'GET':
req = requests_lib.get
else:
req = requests_lib.post
request_args['data'] = request_params['data']
# send the request
response = req(request_params['url'], **request_args)
# is there a timeout (no parsing in this case)
timeout_overhead = 0.2 # seconds
time_after_request = time()
search_duration = time_after_request - start_time
if search_duration > timeout_limit + timeout_overhead:
raise requests.exceptions.Timeout(response=response)
with threading.RLock():
# no error : reset the suspend variables
engine.continuous_errors = 0
engine.suspend_end_time = 0
# update stats with current page-load-time
# only the HTTP request
engine.stats['page_load_time'] += time_after_request - time_before_request
engine.stats['page_load_count'] += 1
# everything is ok : return the response
return response
def search_one_request(engine, query, request_params, start_time, timeout_limit):
# update request parameters dependent on
# search-engine (contained in engines folder)
engine.request(query, request_params)
# ignoring empty urls
if request_params['url'] is None:
return []
if not request_params['url']:
return []
# send request
response = send_http_request(engine, request_params, start_time, timeout_limit)
# parse the response
response.search_params = request_params
return engine.response(response)
def search_one_request_safe(engine_name, query, request_params, result_container, start_time, timeout_limit):
engine = engines[engine_name]
try:
# send requests and parse the results
search_results = search_one_request(engine, query, request_params, start_time, timeout_limit)
# add results
result_container.extend(engine_name, search_results)
# update engine time when there is no exception
with threading.RLock():
engine.stats['engine_time'] += time() - start_time
engine.stats['engine_time_count'] += 1
return True
except Exception as e:
engine.stats['errors'] += 1
search_duration = time() - start_time
requests_exception = False
if (issubclass(e.__class__, requests.exceptions.Timeout)):
# requests timeout (connect or read)
logger.error("engine {0} : HTTP requests timeout"
"(search duration : {1} s, timeout: {2} s) : {3}"
.format(engine_name, search_duration, timeout_limit, e.__class__.__name__))
requests_exception = True
elif (issubclass(e.__class__, requests.exceptions.RequestException)):
# other requests exception
logger.exception("engine {0} : requests exception"
"(search duration : {1} s, timeout: {2} s) : {3}"
.format(engine_name, search_duration, timeout_limit, e))
requests_exception = True
else:
# others errors
logger.exception('engine {0} : exception : {1}'.format(engine_name, e))
# update continuous_errors / suspend_end_time
if requests_exception:
with threading.RLock():
engine.continuous_errors += 1
engine.suspend_end_time = time() + min(60, engine.continuous_errors)
#
return False
def search_multiple_requests(requests, result_container, start_time, timeout_limit):
search_id = uuid4().__str__()
for engine_name, query, request_params in requests:
th = threading.Thread(
target=search_one_request_safe,
args=(engine_name, query, request_params, result_container, start_time, timeout_limit),
name=search_id,
)
2014-12-19 12:59:41 +00:00
th._engine_name = engine_name
th.start()
for th in threading.enumerate():
if th.name == search_id:
remaining_time = max(0.0, timeout_limit - (time() - start_time))
2014-12-14 00:18:01 +00:00
th.join(remaining_time)
if th.isAlive():
2015-01-09 03:13:05 +00:00
logger.warning('engine timeout: {0}'.format(th._engine_name))
2014-12-14 00:18:01 +00:00
# get default reqest parameter
2014-07-07 11:59:27 +00:00
def default_request_params():
return {
'method': 'GET',
'headers': {},
'data': {},
'url': '',
'cookies': {},
'verify': True
}
2014-07-07 11:59:27 +00:00
def get_search_query_from_webapp(preferences, form):
# no text for the query ?
if not form.get('q'):
raise SearxParameterException('q', '')
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
# set blocked engines
disabled_engines = preferences.engines.get_disabled()
# parse query, if tags are set, which change
# the serch engine or search-language
raw_text_query = RawTextQuery(form['q'], disabled_engines)
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
raw_text_query.parse_query()
# set query
query = raw_text_query.getSearchQuery()
# get and check page number
pageno_param = form.get('pageno', '1')
if not pageno_param.isdigit() or int(pageno_param) < 1:
raise SearxParameterException('pageno', pageno_param)
query_pageno = int(pageno_param)
# get language
2016-12-14 03:36:40 +00:00
# set specific language if set on request, query or preferences
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
# TODO support search with multible languages
if len(raw_text_query.languages):
query_lang = raw_text_query.languages[-1]
2016-12-14 03:36:40 +00:00
elif 'language' in form:
query_lang = form.get('language')
else:
query_lang = preferences.get_value('language')
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
# check language
2017-01-21 19:21:32 +00:00
if query_lang.lower() not in language_code_set:
raise SearxParameterException('language', query_lang)
# get safesearch
if 'safesearch' in form:
query_safesearch = form.get('safesearch')
# first check safesearch
if not query_safesearch.isdigit():
raise SearxParameterException('safesearch', query_safesearch)
query_safesearch = int(query_safesearch)
else:
query_safesearch = preferences.get_value('safesearch')
# safesearch : second check
if query_safesearch < 0 or query_safesearch > 2:
raise SearxParameterException('safesearch', query_safesearch)
# get time_range
query_time_range = form.get('time_range')
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
# check time_range
2017-01-21 19:21:32 +00:00
if query_time_range not in ('None', None, '', 'day', 'week', 'month', 'year'):
raise SearxParameterException('time_range', query_time_range)
# query_engines
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
query_engines = raw_text_query.engines
# query_categories
query_categories = []
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
# if engines are calculated from query,
# set categories by using that informations
if query_engines and raw_text_query.specific:
query_categories = list(set(engine['category']
for engine in query_engines))
# otherwise, using defined categories to
# calculate which engines should be used
else:
# set categories/engines
load_default_categories = True
for pd_name, pd in form.items():
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
if pd_name == 'categories':
query_categories.extend(categ for categ in map(unicode.strip, pd.split(',')) if categ in categories)
elif pd_name == 'engines':
pd_engines = [{'category': engines[engine].categories[0],
'name': engine}
for engine in map(unicode.strip, pd.split(',')) if engine in engines]
if pd_engines:
query_engines.extend(pd_engines)
load_default_categories = False
elif pd_name.startswith('category_'):
category = pd_name[9:]
# if category is not found in list, skip
if category not in categories:
continue
if pd != 'off':
# add category to list
query_categories.append(category)
elif category in query_categories:
# remove category from list if property is set to 'off'
query_categories.remove(category)
if not load_default_categories:
if not query_categories:
query_categories = list(set(engine['category']
for engine in query_engines))
else:
2014-10-19 10:41:04 +00:00
# if no category is specified for this search,
# using user-defined default-configuration which
# (is stored in cookie)
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
if not query_categories:
cookie_categories = preferences.get_value('categories')
for ccateg in cookie_categories:
if ccateg in categories:
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
query_categories.append(ccateg)
2014-10-19 10:41:04 +00:00
# if still no category is specified, using general
# as default-category
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
if not query_categories:
query_categories = ['general']
2014-10-19 10:41:04 +00:00
# using all engines for that search, which are
# declared under the specific categories
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
for categ in query_categories:
query_engines.extend({'category': categ,
'name': engine.name}
for engine in categories[categ]
if (engine.name, categ) not in disabled_engines)
return SearchQuery(query, query_engines, query_categories,
query_lang, query_safesearch, query_pageno, query_time_range)
2014-02-09 00:07:18 +00:00
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
class Search(object):
"""Search information container"""
def __init__(self, search_query):
# init vars
super(Search, self).__init__()
self.search_query = search_query
self.result_container = ResultContainer()
# do search-request
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
def search(self):
2014-07-07 11:59:27 +00:00
global number_of_searches
# start time
start_time = time()
2016-12-09 22:11:45 +00:00
# answeres ?
answerers_results = ask(self.search_query)
if answerers_results:
for results in answerers_results:
self.result_container.extend('answer', results)
return self.result_container
# init vars
2014-07-07 11:59:27 +00:00
requests = []
# increase number of searches
2014-07-07 11:59:27 +00:00
number_of_searches += 1
# set default useragent
2014-10-19 10:41:04 +00:00
# user_agent = request.headers.get('User-Agent', '')
2014-07-07 11:59:27 +00:00
user_agent = gen_useragent()
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
search_query = self.search_query
# max of all selected engine timeout
timeout_limit = 0
# start search-reqest for all selected engines
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
for selected_engine in search_query.engines:
2014-07-07 11:59:27 +00:00
if selected_engine['name'] not in engines:
continue
engine = engines[selected_engine['name']]
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
# skip suspended engines
2016-11-06 01:51:14 +00:00
if engine.suspend_end_time >= time():
logger.debug('Engine currently suspended: %s', selected_engine['name'])
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
continue
# if paging is not supported, skip
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
if search_query.pageno > 1 and not engine.paging:
2014-07-07 11:59:27 +00:00
continue
2014-10-19 10:41:04 +00:00
# if search-language is set and engine does not
# provide language-support, skip
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
if search_query.lang != 'all' and not engine.language_support:
2014-07-07 11:59:27 +00:00
continue
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
# if time_range is not supported, skip
if search_query.time_range and not engine.time_range_support:
2016-07-17 16:42:30 +00:00
continue
# set default request parameters
2014-07-07 11:59:27 +00:00
request_params = default_request_params()
request_params['headers']['User-Agent'] = user_agent
request_params['category'] = selected_engine['category']
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
request_params['pageno'] = search_query.pageno
if hasattr(engine, 'language') and engine.language:
request_params['language'] = engine.language
else:
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
request_params['language'] = search_query.lang
# 0 = None, 1 = Moderate, 2 = Strict
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
request_params['safesearch'] = search_query.safesearch
request_params['time_range'] = search_query.time_range
# append request to list
requests.append((selected_engine['name'], search_query.query.encode('utf-8'), request_params))
2014-07-07 11:59:27 +00:00
# update timeout_limit
timeout_limit = max(timeout_limit, engine.timeout)
2014-07-07 11:59:27 +00:00
if requests:
# send all search-request
search_multiple_requests(requests, self.result_container, start_time, timeout_limit)
start_new_thread(gc.collect, tuple())
2014-09-28 14:51:41 +00:00
# return results, suggestions, answers and infoboxes
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
return self.result_container
class SearchWithPlugins(Search):
"""Similar to the Search class but call the plugins."""
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
def __init__(self, search_query, request):
super(SearchWithPlugins, self).__init__(search_query)
self.request = request
def search(self):
if plugins.call('pre_search', self.request, self):
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
super(SearchWithPlugins, self).search()
plugins.call('post_search', self.request, self)
results = self.result_container.get_ordered_results()
for result in results:
plugins.call('on_result', self.request, self, result)
Clean up the architecture Purposes : - isolate the plugins calls - distinction between parsing the web request and running the search (Search class). To be able to test code easily, to run searx code outside a web server, to filter the search query parameters with plugins more easily, etc... Details : - request.request_data contains request.form or request.args (initialize inside pre_request() function) - Query class is renamed RawTextQuery - SearchQuery class defines all search parameters - get_search_query_from_webapp create a SearchQuery instance (basically the previous Search.__init__ code) - Search class and SearchWithPlugins class takes a SearchQuery instance as class constructor parameter - SearchWithPlugins class inherites from Search class, and run plugins - A dedicated function search_with_plugins executes plugins to have a well define locals() (which is used by the plugins code). - All plugins code is executed inside the try...except block (webapp.py, index function) - advanced_search HTTP parameter value stays in webapp.py (it is only part of UI) - multiple calls to result_container.get_ordered_results() doesn't compute the order multiple time (note : this method was call only once before) - paging value is stored in the result_container class (compute in the extend method) - test about engine.suspend_end_time is done during search method call (instead of __init__) - check that the format parameter value is one of these : html, rss, json, rss (before the html value was assumed but some text formatting wasn't not done)
2016-10-22 11:10:31 +00:00
return self.result_container