mirrors/searxng
1
0
Fork 0
mirror of https://github.com/searxng/searxng.git synced 2024-11-29 06:01:02 +00:00
Code Issues Projects Releases Wiki Activity

[mod] Google: reversed engineered & upgrade to data_type: traits_v1

Browse source 
Partial reverse engineering of the Google engines including a improved language
and region handling based on the engine.traits_v1 data.

When ever possible the implementations of the Google engines try to make use of
the async REST APIs.  The get_lang_info() has been generalized to a
get_google_info() function / especially the region handling has been improved by
adding the cr parameter.

searx/data/engine_traits.json
  Add data type "traits_v1" generated by the fetch_traits() functions from:

  - Google (WEB),
  - Google images,
  - Google news,
  - Google scholar and
  - Google videos

  and remove data from obsolete data type "supported_languages".

  A traits.custom type that maps region codes to *supported_domains* is fetched
  from https://www.google.com/supported_domains

searx/autocomplete.py:
  Reversed engineered autocomplete from Google WEB.  Supports Google's languages and
  subdomains.  The old API suggestqueries.google.com/complete has been replaced
  by the async REST API: https://{subdomain}/complete/search?{args}

searx/engines/google.py
  Reverse engineering and extensive testing ..
  - fetch_traits():  Fetch languages & regions from Google properties.
  - always use the async REST API (formally known as 'use_mobile_ui')
  - use *supported_domains* from traits
  - improved the result list by fetching './/div[@data-content-feature]'
    and parsing the type of the various *content features* --> thumbnails are
    added

searx/engines/google_images.py
  Reverse engineering and extensive testing ..
  - fetch_traits():  Fetch languages & regions from Google properties.
  - use *supported_domains* from traits
  - if exists, freshness_date is added to the result
  - issue 1864: result list has been improved a lot (due to the new cr parameter)

searx/engines/google_news.py
  Reverse engineering and extensive testing ..
  - fetch_traits():  Fetch languages & regions from Google properties.
    *supported_domains* is not needed but a ceid list has been added.
  - different region handling compared to Google WEB
  - fixed for various languages & regions (due to the new ceid parameter) /
    avoid CONSENT page
  - Google News do no longer support time range
  - result list has been fixed: XPath of pub_date and pub_origin

searx/engines/google_videos.py
  - fetch_traits():  Fetch languages & regions from Google properties.
  - use *supported_domains* from traits
  - add paging support
  - implement a async request ('asearch': 'arc' & 'async':
    'use_ac:true,_fmt:html')
  - simplified code (thanks to '_fmt:html' request)
  - issue 1359: fixed xpath of video length data

searx/engines/google_scholar.py
  - fetch_traits():  Fetch languages & regions from Google properties.
  - use *supported_domains* from traits
  - request(): include patents & citations
  - response(): fixed CAPTCHA detection (Scholar has its own CATCHA manager)
  - hardening XPath to iterate over results
  - fixed XPath of pub_type (has been change from gs_ct1 to gs_cgt2 class)
  - issue 1769 fixed: new request implementation is no longer incompatible

Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
This commit is contained in:
Markus Heiser 2022-12-04 22:57:22 +01:00
parent c80e82a855
commit 2499899554
11 changed files with 2510 additions and 2174 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

27
docs/src/searx.engines.google.rst
View file

@ -12,15 +12,21 @@ Google Engines
.. _google API: .. _google API:
google API Google API
========== ==========
.. _Query Parameter Definitions: .. _Query Parameter Definitions:
https://developers.google.com/custom-search/docs/xml_results#WebSearch_Query_Parameter_Definitions https://developers.google.com/custom-search/docs/xml_results#WebSearch_Query_Parameter_Definitions
SearXNG's implementation of the Google API is mainly done in
:py:obj:`get_google_info <searx.engines.google.get_google_info>`.
For detailed description of the *REST-full* API see: `Query Parameter For detailed description of the *REST-full* API see: `Query Parameter
Definitions`_. Not all parameters can be appied and some engines are *special* Definitions`_. The linked API documentation can sometimes be helpful during
(e.g. :ref:`google news engine`). reverse engineering. However, we cannot use it in the freely accessible WEB
services; not all parameters can be applied and some engines are more *special*
than other (e.g. :ref:`google news engine`).
.. _google web engine: .. _google web engine:
@ -30,6 +36,13 @@ Google WEB
.. automodule:: searx.engines.google .. automodule:: searx.engines.google
:members: :members:
.. _google autocomplete:
Google Autocomplete
====================
.. autofunction:: searx.autocomplete.google_complete
.. _google images engine: .. _google images engine:
Google Images Google Images
@ -53,3 +66,11 @@ Google News
.. automodule:: searx.engines.google_news .. automodule:: searx.engines.google_news
:members: :members:
.. _google scholar engine:
Google Scholar
==============
.. automodule:: searx.engines.google_scholar
:members:

49
searx/autocomplete.py
View file

@ -5,14 +5,17 @@
""" """
# pylint: disable=use-dict-literal # pylint: disable=use-dict-literal
from json import loads import json
from urllib.parse import urlencode from urllib.parse import urlencode
from lxml import etree import lxml
from httpx import HTTPError from httpx import HTTPError
from searx import settings from searx import settings
from searx.engines import engines from searx.engines import (
engines,
google,
)
from searx.network import get as http_get from searx.network import get as http_get
from searx.exceptions import SearxEngineResponseException from searx.exceptions import SearxEngineResponseException
@ -55,7 +58,7 @@ def dbpedia(query, _lang):
results = [] results = []
if response.ok: if response.ok:
dom = etree.fromstring(response.content) dom = lxml.etree.fromstring(response.content)
results = dom.xpath('//Result/Label//text()') results = dom.xpath('//Result/Label//text()')
return results return results
@ -81,18 +84,32 @@ def duckduckgo(query, sxng_locale):
return ret_val return ret_val
def google(query, lang): def google_complete(query, sxng_locale):
# google autocompleter """Autocomplete from Google. Supports Google's languages and subdomains
autocomplete_url = 'https://suggestqueries.google.com/complete/search?client=toolbar&' (:py:obj:`searx.engines.google.get_google_info`) by using the async REST
API::
response = get(autocomplete_url + urlencode(dict(hl=lang, q=query))) https://{subdomain}/complete/search?{args}
"""
google_info = google.get_google_info({'searxng_locale': sxng_locale}, engines['google'].traits)
url = 'https://{subdomain}/complete/search?{args}'
args = urlencode(
{
'q': query,
'client': 'gws-wiz',
'hl': google_info['params']['hl'],
}
)
results = [] results = []
resp = get(url.format(subdomain=google_info['subdomain'], args=args))
if response.ok: if resp.ok:
dom = etree.fromstring(response.text) json_txt = resp.text[resp.text.find('[') : resp.text.find(']', -3) + 1]
results = dom.xpath('//suggestion/@data') data = json.loads(json_txt)
for item in data[0]:
results.append(lxml.html.fromstring(item[0]).text_content())
return results return results
@ -132,7 +149,7 @@ def swisscows(query, _lang):
# swisscows autocompleter # swisscows autocompleter
url = 'https://swisscows.ch/api/suggest?{query}&itemsCount=5' url = 'https://swisscows.ch/api/suggest?{query}&itemsCount=5'
resp = loads(get(url.format(query=urlencode({'query': query}))).text) resp = json.loads(get(url.format(query=urlencode({'query': query}))).text)
return resp return resp
@ -184,7 +201,7 @@ def yandex(query, _lang):
# yandex autocompleter # yandex autocompleter
url = "https://suggest.yandex.com/suggest-ff.cgi?{0}" url = "https://suggest.yandex.com/suggest-ff.cgi?{0}"
resp = loads(get(url.format(urlencode(dict(part=query)))).text) resp = json.loads(get(url.format(urlencode(dict(part=query)))).text)
if len(resp) > 1: if len(resp) > 1:
return resp[1] return resp[1]
return [] return []
@ -193,7 +210,7 @@ def yandex(query, _lang):
backends = { backends = {
'dbpedia': dbpedia, 'dbpedia': dbpedia,
'duckduckgo': duckduckgo, 'duckduckgo': duckduckgo,
'google': google, 'google': google_complete,
'seznam': seznam, 'seznam': seznam,
'startpage': startpage, 'startpage': startpage,
'swisscows': swisscows, 'swisscows': swisscows,

3613
searx/data/engine_traits.json
View file

File diff suppressed because it is too large Load diff

440
searx/engines/google.py
View file

@ -1,38 +1,40 @@
# SPDX-License-Identifier: AGPL-3.0-or-later # SPDX-License-Identifier: AGPL-3.0-or-later
# lint: pylint # lint: pylint
"""This is the implementation of the google WEB engine. Some of this """This is the implementation of the Google WEB engine. Some of this
implementations are shared by other engines: implementations (manly the :py:obj:`get_google_info`) are shared by other
engines:
- :ref:`google images engine` - :ref:`google images engine`
- :ref:`google news engine` - :ref:`google news engine`
- :ref:`google videos engine` - :ref:`google videos engine`
- :ref:`google scholar engine`
The google WEB engine itself has a special setup option: - :ref:`google autocomplete`
.. code:: yaml
- name: google
...
use_mobile_ui: false
``use_mobile_ui``: (default: ``false``)
Enables to use *mobile endpoint* to bypass the google blocking (see
:issue:`159`). On the mobile UI of Google Search, the button :guilabel:`More
results` is not affected by Google rate limiting and we can still do requests
while actively blocked by the original Google search. By activate
``use_mobile_ui`` this behavior is simulated by adding the parameter
``async=use_ac:true,_fmt:pc`` to the :py:func:`request`.
""" """
from typing import TYPE_CHECKING
import re
from urllib.parse import urlencode from urllib.parse import urlencode
from lxml import html from lxml import html
from searx.utils import match_language, extract_text, eval_xpath, eval_xpath_list, eval_xpath_getindex import babel
import babel.core
import babel.languages
from searx.utils import extract_text, eval_xpath, eval_xpath_list, eval_xpath_getindex
from searx.locales import language_tag, region_tag, get_offical_locales
from searx import network
from searx.exceptions import SearxEngineCaptchaException from searx.exceptions import SearxEngineCaptchaException
from searx.enginelib.traits import EngineTraits from searx.enginelib.traits import EngineTraits
if TYPE_CHECKING:
import logging
logger: logging.Logger
traits: EngineTraits traits: EngineTraits
# about # about
about = { about = {
"website": 'https://www.google.com', "website": 'https://www.google.com',
@ -48,64 +50,6 @@ categories = ['general', 'web']
paging = True paging = True
time_range_support = True time_range_support = True
safesearch = True safesearch = True
send_accept_language_header = True
use_mobile_ui = False
supported_languages_url = 'https://www.google.com/preferences?#languages'
# based on https://en.wikipedia.org/wiki/List_of_Google_domains and tests
google_domains = {
'BG': 'google.bg', # Bulgaria
'CZ': 'google.cz', # Czech Republic
'DE': 'google.de', # Germany
'DK': 'google.dk', # Denmark
'AT': 'google.at', # Austria
'CH': 'google.ch', # Switzerland
'GR': 'google.gr', # Greece
'AU': 'google.com.au', # Australia
'CA': 'google.ca', # Canada
'GB': 'google.co.uk', # United Kingdom
'ID': 'google.co.id', # Indonesia
'IE': 'google.ie', # Ireland
'IN': 'google.co.in', # India
'MY': 'google.com.my', # Malaysia
'NZ': 'google.co.nz', # New Zealand
'PH': 'google.com.ph', # Philippines
'SG': 'google.com.sg', # Singapore
'US': 'google.com', # United States (google.us) redirects to .com
'ZA': 'google.co.za', # South Africa
'AR': 'google.com.ar', # Argentina
'CL': 'google.cl', # Chile
'ES': 'google.es', # Spain
'MX': 'google.com.mx', # Mexico
'EE': 'google.ee', # Estonia
'FI': 'google.fi', # Finland
'BE': 'google.be', # Belgium
'FR': 'google.fr', # France
'IL': 'google.co.il', # Israel
'HR': 'google.hr', # Croatia
'HU': 'google.hu', # Hungary
'IT': 'google.it', # Italy
'JP': 'google.co.jp', # Japan
'KR': 'google.co.kr', # South Korea
'LT': 'google.lt', # Lithuania
'LV': 'google.lv', # Latvia
'NO': 'google.no', # Norway
'NL': 'google.nl', # Netherlands
'PL': 'google.pl', # Poland
'BR': 'google.com.br', # Brazil
'PT': 'google.pt', # Portugal
'RO': 'google.ro', # Romania
'RU': 'google.ru', # Russia
'SK': 'google.sk', # Slovakia
'SI': 'google.si', # Slovenia
'SE': 'google.se', # Sweden
'TH': 'google.co.th', # Thailand
'TR': 'google.com.tr', # Turkey
'UA': 'google.com.ua', # Ukraine
'CN': 'google.com.hk', # There is no google.cn, we use .com.hk for zh-CN
'HK': 'google.com.hk', # Hong Kong
'TW': 'google.com.tw', # Taiwan
}
time_range_dict = {'day': 'd', 'week': 'w', 'month': 'm', 'year': 'y'} time_range_dict = {'day': 'd', 'week': 'w', 'month': 'm', 'year': 'y'}
@ -118,47 +62,50 @@ filter_mapping = {0: 'off', 1: 'medium', 2: 'high'}
results_xpath = './/div[@data-sokoban-container]' results_xpath = './/div[@data-sokoban-container]'
title_xpath = './/a/h3[1]' title_xpath = './/a/h3[1]'
href_xpath = './/a[h3]/@href' href_xpath = './/a[h3]/@href'
content_xpath = './/div[@data-content-feature=1]' content_xpath = './/div[@data-content-feature]'
# google *sections* are no usual *results*, we ignore them # google *sections* are no usual *results*, we ignore them
g_section_with_header = './g-section-with-header' g_section_with_header = './g-section-with-header'
# Suggestions are links placed in a *card-section*, we extract only the text # Suggestions are links placed in a *card-section*, we extract only the text
# from the links not the links itself. # from the links not the links itself.
suggestion_xpath = '//div[contains(@class, "EIaa9b")]//a' suggestion_xpath = '//div[contains(@class, "EIaa9b")]//a'
# UI_ASYNC = 'use_ac:true,_fmt:html' # returns a HTTP 500 when user search for
# # celebrities like '!google natasha allegri'
# # or '!google chris evans'
UI_ASYNC = 'use_ac:true,_fmt:prog'
"""Format of the response from UI's async request."""
def get_lang_info(params, lang_list, custom_aliases, supported_any_language):
"""Composing various language properties for the google engines. def get_google_info(params, eng_traits):
"""Composing various (language) properties for the google engines (:ref:`google
API`).
This function is called by the various google engines (:ref:`google web This function is called by the various google engines (:ref:`google web
engine`, :ref:`google images engine`, :ref:`google news engine` and engine`, :ref:`google images engine`, :ref:`google news engine` and
:ref:`google videos engine`). :ref:`google videos engine`).
:param dict param: request parameters of the engine :param dict param: Request parameters of the engine. At least
a ``searxng_locale`` key should be in the dictionary.
:param list lang_list: list of supported languages of the engine :param eng_traits: Engine's traits fetched from google preferences
:py:obj:`ENGINES_LANGUAGES[engine-name] <searx.data.ENGINES_LANGUAGES>` (:py:obj:`searx.enginelib.traits.EngineTraits`)
:param dict lang_list: custom aliases for non standard language codes
(used when calling :py:func:`searx.utils.match_language`)
:param bool supported_any_language: When a language is not specified, the
language interpretation is left up to Google to decide how the search
results should be delivered. This argument is ``True`` for the google
engine and ``False`` for the other engines (google-images, -news,
-scholar, -videos).
:rtype: dict :rtype: dict
:returns: :returns:
Py-Dictionary with the key/value pairs: Py-Dictionary with the key/value pairs:
language: language:
Return value from :py:func:`searx.utils.match_language` The language code that is used by google (e.g. ``lang_en`` or
``lang_zh-TW``)
country: country:
The country code (e.g. US, AT, CA, FR, DE ..) The country code that is used by google (e.g. ``US`` or ``TW``)
locale:
A instance of :py:obj:`babel.core.Locale` build from the
``searxng_locale`` value.
subdomain: subdomain:
Google subdomain :py:obj:`google_domains` that fits to the country Google subdomain :py:obj:`google_domains` that fits to the country
@ -168,52 +115,67 @@ def get_lang_info(params, lang_list, custom_aliases, supported_any_language):
Py-Dictionary with additional request arguments (can be passed to Py-Dictionary with additional request arguments (can be passed to
:py:func:`urllib.parse.urlencode`). :py:func:`urllib.parse.urlencode`).
- ``hl`` parameter: specifies the interface language of user interface.
- ``lr`` parameter: restricts search results to documents written in
a particular language.
- ``cr`` parameter: restricts search results to documents
originating in a particular country.
- ``ie`` parameter: sets the character encoding scheme that should
be used to interpret the query string ('utf8').
- ``oe`` parameter: sets the character encoding scheme that should
be used to decode the XML result ('utf8').
headers: headers:
Py-Dictionary with additional HTTP headers (can be passed to Py-Dictionary with additional HTTP headers (can be passed to
request's headers) request's headers)
- ``Accept: '*/*``
""" """
ret_val = { ret_val = {
'language': None, 'language': None,
'country': None, 'country': None,
'subdomain': None, 'subdomain': None,
'params': {}, 'params': {},
'headers': {}, 'headers': {},
'cookies': {},
'locale': None,
} }
# language ... sxng_locale = params.get('searxng_locale', 'all')
try:
locale = babel.Locale.parse(sxng_locale, sep='-')
except babel.core.UnknownLocaleError:
locale = None
_lang = params['language'] eng_lang = eng_traits.get_language(sxng_locale, 'lang_en')
_any_language = _lang.lower() == 'all' lang_code = eng_lang.split('_')[-1] # lang_zh-TW --> zh-TW / lang_en --> en
if _any_language: country = eng_traits.get_region(sxng_locale, eng_traits.all_locale)
_lang = 'en-US'
language = match_language(_lang, lang_list, custom_aliases)
ret_val['language'] = language
# country ... # Test zh_hans & zh_hant --> in the topmost links in the result list of list
# TW and HK you should a find wiktionary.org zh_hant link. In the result
# list of zh-CN should not be no hant link instead you should find
# zh.m.wikipedia.org/zh somewhere in the top.
_l = _lang.split('-') # '!go 日 :zh-TW' --> https://zh.m.wiktionary.org/zh-hant/%E6%97%A5
if len(_l) == 2: # '!go 日 :zh-CN' --> https://zh.m.wikipedia.org/zh/%E6%97%A5
country = _l[1]
else: ret_val['language'] = eng_lang
country = _l[0].upper()
if country == 'EN':
country = 'US'
ret_val['country'] = country ret_val['country'] = country
ret_val['locale'] = locale
# subdomain ... ret_val['subdomain'] = eng_traits.custom['supported_domains'].get(country.upper(), 'www.google.com')
ret_val['subdomain'] = 'www.' + google_domains.get(country.upper(), 'google.com')
# params & headers
lang_country = '%s-%s' % (language, country) # (en-US, en-EN, de-DE, de-AU, fr-FR ..)
# hl parameter: # hl parameter:
# https://developers.google.com/custom-search/docs/xml_results#hlsp The # The hl parameter specifies the interface language (host language) of
# Interface Language: # your user interface. To improve the performance and the quality of your
# search results, you are strongly encouraged to set this parameter
# explicitly.
# https://developers.google.com/custom-search/docs/xml_results#hlsp
# The Interface Language:
# https://developers.google.com/custom-search/docs/xml_results_appendices#interfaceLanguages # https://developers.google.com/custom-search/docs/xml_results_appendices#interfaceLanguages
ret_val['params']['hl'] = lang_list.get(lang_country, language) ret_val['params']['hl'] = lang_code
# lr parameter: # lr parameter:
# The lr (language restrict) parameter restricts search results to # The lr (language restrict) parameter restricts search results to
@ -221,22 +183,72 @@ def get_lang_info(params, lang_list, custom_aliases, supported_any_language):
# https://developers.google.com/custom-search/docs/xml_results#lrsp # https://developers.google.com/custom-search/docs/xml_results#lrsp
# Language Collection Values: # Language Collection Values:
# https://developers.google.com/custom-search/docs/xml_results_appendices#languageCollections # https://developers.google.com/custom-search/docs/xml_results_appendices#languageCollections
#
# To select 'all' languages an empty 'lr' value is used.
#
# Different to other google services, Google Schloar supports to select more
# than one language. The languages are seperated by a pipe '|' (logical OR).
# By example: &lr=lang_zh-TW%7Clang_de selects articles written in
# traditional chinese OR german language.
if _any_language and supported_any_language: ret_val['params']['lr'] = eng_lang
if sxng_locale == 'all':
ret_val['params']['lr'] = ''
# interpretation is left up to Google (based on whoogle) # cr parameter:
# # The cr parameter restricts search results to documents originating in a
# - add parameter ``source=lnt`` # particular country.
# - don't use parameter ``lr`` # https://developers.google.com/custom-search/docs/xml_results#crsp
# - don't add a ``Accept-Language`` HTTP header.
ret_val['params']['source'] = 'lnt' ret_val['params']['cr'] = 'country' + country
if sxng_locale == 'all':
ret_val['params']['cr'] = ''
else: # gl parameter: (mandatory by Geeogle News)
# The gl parameter value is a two-letter country code. For WebSearch
# results, the gl parameter boosts search results whose country of origin
# matches the parameter value. See the Country Codes section for a list of
# valid values.
# Specifying a gl parameter value in WebSearch requests should improve the
# relevance of results. This is particularly true for international
# customers and, even more specifically, for customers in English-speaking
# countries other than the United States.
# https://developers.google.com/custom-search/docs/xml_results#glsp
# restricts search results to documents written in a particular ret_val['params']['gl'] = country
# language.
ret_val['params']['lr'] = "lang_" + lang_list.get(lang_country, language) # ie parameter:
# The ie parameter sets the character encoding scheme that should be used
# to interpret the query string. The default ie value is latin1.
# https://developers.google.com/custom-search/docs/xml_results#iesp
ret_val['params']['ie'] = 'utf8'
# oe parameter:
# The oe parameter sets the character encoding scheme that should be used
# to decode the XML result. The default oe value is latin1.
# https://developers.google.com/custom-search/docs/xml_results#oesp
ret_val['params']['oe'] = 'utf8'
# num parameter:
# The num parameter identifies the number of search results to return.
# The default num value is 10, and the maximum value is 20. If you request
# more than 20 results, only 20 results will be returned.
# https://developers.google.com/custom-search/docs/xml_results#numsp
# HINT: seems to have no effect (tested in google WEB & Images)
# ret_val['params']['num'] = 20
# HTTP headers
ret_val['headers']['Accept'] = '*/*'
# Cookies
# - https://github.com/searxng/searxng/pull/1679#issuecomment-1235432746
# - https://github.com/searxng/searxng/issues/1555
ret_val['cookies']['CONSENT'] = "YES+"
return ret_val return ret_val
@ -248,33 +260,34 @@ def detect_google_sorry(resp):
def request(query, params): def request(query, params):
"""Google search request""" """Google search request"""
# pylint: disable=line-too-long
offset = (params['pageno'] - 1) * 10 offset = (params['pageno'] - 1) * 10
google_info = get_google_info(params, traits)
lang_info = get_lang_info(params, supported_languages, language_aliases, True)
additional_parameters = {}
if use_mobile_ui:
additional_parameters = {
'asearch': 'arc',
'async': 'use_ac:true,_fmt:prog',
}
# https://www.google.de/search?q=corona&hl=de&lr=lang_de&start=0&tbs=qdr%3Ad&safe=medium # https://www.google.de/search?q=corona&hl=de&lr=lang_de&start=0&tbs=qdr%3Ad&safe=medium
query_url = ( query_url = (
'https://' 'https://'
+ lang_info['subdomain'] + google_info['subdomain']
+ '/search' + '/search'
+ "?" + "?"
+ urlencode( + urlencode(
{ {
'q': query, 'q': query,
**lang_info['params'], **google_info['params'],
'ie': "utf8",
'oe': "utf8",
'start': offset,
'filter': '0', 'filter': '0',
**additional_parameters, 'start': offset,
# 'vet': '12ahUKEwik3ZbIzfn7AhXMX_EDHbUDBh0QxK8CegQIARAC..i',
# 'ved': '2ahUKEwik3ZbIzfn7AhXMX_EDHbUDBh0Q_skCegQIARAG',
# 'cs' : 1,
# 'sa': 'N',
# 'yv': 3,
# 'prmd': 'vin',
# 'ei': 'GASaY6TxOcy_xc8PtYeY6AE',
# 'sa': 'N',
# 'sstk': 'AcOHfVkD7sWCSAheZi-0tx_09XDO55gTWY0JNq3_V26cNN-c8lfD45aZYPI8s_Bqp8s57AHz5pxchDtAGCA_cikAWSjy9kw3kgg'
# formally known as use_mobile_ui
'asearch': 'arc',
'async': UI_ASYNC,
} }
) )
) )
@ -285,25 +298,45 @@ def request(query, params):
query_url += '&' + urlencode({'safe': filter_mapping[params['safesearch']]}) query_url += '&' + urlencode({'safe': filter_mapping[params['safesearch']]})
params['url'] = query_url params['url'] = query_url
params['cookies']['CONSENT'] = "YES+" params['cookies'] = google_info['cookies']
params['headers'].update(lang_info['headers']) params['headers'].update(google_info['headers'])
if use_mobile_ui:
params['headers']['Accept'] = '*/*'
else:
params['headers']['Accept'] = 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8'
return params return params
# (function(){var s='data:image/jpeg;base64,/9j/4AAQ ...
# ... DX/Ff5XSpSgdU32xSlKDJ//9k\x3d';var ii=['dimg_21'];_setImagesSrc(ii,s);})();
RE_DATA_IMAGE = re.compile(r"'(data:image[^']*)'[^']*ii=\['([^']*)'\];_setImagesSrc")
def _parse_data_images(dom):
data_image_map = {}
for _script in eval_xpath_list(dom, "//script[@nonce]"):
script = _script.text
if not script:
continue
script = RE_DATA_IMAGE.search(script)
if not script:
continue
data_image_map[script.group(2)] = script.group(1).replace(r'\x3d', '=')
logger.debug('data:image objects --> %s', list(data_image_map.keys()))
return data_image_map
def response(resp): def response(resp):
"""Get response from google's search request""" """Get response from google's search request"""
# pylint: disable=too-many-branches, too-many-statements
detect_google_sorry(resp) detect_google_sorry(resp)
results = [] results = []
# convert the text to dom # convert the text to dom
dom = html.fromstring(resp.text) dom = html.fromstring(resp.text)
data_image_map = {}
if '_fmt:html' in UI_ASYNC:
# in this format images are embedded by a bse64 encoded 'data:image'
data_image_map = _parse_data_images(dom)
# results --> answer # results --> answer
answer_list = eval_xpath(dom, '//div[contains(@class, "LGOjhe")]') answer_list = eval_xpath(dom, '//div[contains(@class, "LGOjhe")]')
if answer_list: if answer_list:
@ -312,20 +345,9 @@ def response(resp):
else: else:
logger.debug("did not find 'answer'") logger.debug("did not find 'answer'")
# results --> number_of_results
if not use_mobile_ui:
try:
_txt = eval_xpath_getindex(dom, '//div[@id="result-stats"]//text()', 0)
_digit = ''.join([n for n in _txt if n.isdigit()])
number_of_results = int(_digit)
results.append({'number_of_results': number_of_results})
except Exception as e: # pylint: disable=broad-except
logger.debug("did not 'number_of_results'")
logger.error(e, exc_info=True)
# parse results # parse results
for result in eval_xpath_list(dom, results_xpath): for result in eval_xpath_list(dom, results_xpath): # pylint: disable=too-many-nested-blocks
# google *sections* # google *sections*
if extract_text(eval_xpath(result, g_section_with_header)): if extract_text(eval_xpath(result, g_section_with_header)):
@ -342,13 +364,31 @@ def response(resp):
url = eval_xpath_getindex(result, href_xpath, 0, None) url = eval_xpath_getindex(result, href_xpath, 0, None)
if url is None: if url is None:
continue continue
content = extract_text(eval_xpath_getindex(result, content_xpath, 0, default=None), allow_none=True)
if content is None: content = []
img_list = []
for content_feature in eval_xpath(result, content_xpath):
val = content_feature.attrib['data-content-feature']
if val in ['1', '2']:
txt = extract_text(content_feature, allow_none=True)
if txt:
content.append(txt)
elif '0' in val:
img = content_feature.xpath('.//img/@src')
if img:
img = img[0]
if img.startswith('data:image'):
img_id = content_feature.xpath('.//img/@id')
if img_id:
img = data_image_map.get(img_id[0])
img_list.append(img)
if not content:
logger.debug('ignoring item from the result_xpath list: missing content of title "%s"', title) logger.debug('ignoring item from the result_xpath list: missing content of title "%s"', title)
continue continue
content = ' / '.join(content)
logger.debug('add link to results: %s', title) img_src = img_list[0] if img_list else None
results.append({'url': url, 'title': title, 'content': content}) results.append({'url': url, 'title': title, 'content': content, 'img_src': img_src})
except Exception as e: # pylint: disable=broad-except except Exception as e: # pylint: disable=broad-except
logger.error(e, exc_info=True) logger.error(e, exc_info=True)
@ -364,18 +404,6 @@ def response(resp):
# get supported languages from their site # get supported languages from their site
def _fetch_supported_languages(resp):
ret_val = {}
dom = html.fromstring(resp.text)
radio_buttons = eval_xpath_list(dom, '//*[@id="langSec"]//input[@name="lr"]')
for x in radio_buttons:
name = x.get("data-name")
code = x.get("value").split('_')[-1]
ret_val[code] = {"name": name}
return ret_val
skip_countries = [ skip_countries = [
@ -404,25 +432,21 @@ skip_countries = [
] ]
def fetch_traits(engine_traits: EngineTraits): def fetch_traits(engine_traits: EngineTraits, add_domains: bool = True):
"""Fetch languages from Google.""" """Fetch languages from Google."""
# pylint: disable=import-outside-toplevel # pylint: disable=import-outside-toplevel, too-many-branches
engine_traits.data_type = 'supported_languages' # deprecated engine_traits.custom['supported_domains'] = {}
import babel
import babel.languages
from searx import network
from searx.locales import language_tag, region_tag, get_offical_locales
resp = network.get('https://www.google.com/preferences') resp = network.get('https://www.google.com/preferences')
if not resp.ok: if not resp.ok:
print("ERROR: response from Google is not OK.") raise RuntimeError("Response from Google's preferences is not OK.")
dom = html.fromstring(resp.text) dom = html.fromstring(resp.text)
lang_map = {'no': 'nb'} # supported language codes
lang_map = {'no': 'nb'}
for x in eval_xpath_list(dom, '//*[@id="langSec"]//input[@name="lr"]'): for x in eval_xpath_list(dom, '//*[@id="langSec"]//input[@name="lr"]'):
eng_lang = x.get("value").split('_')[-1] eng_lang = x.get("value").split('_')[-1]
@ -443,6 +467,8 @@ def fetch_traits(engine_traits: EngineTraits):
# alias languages # alias languages
engine_traits.languages['zh'] = 'lang_zh-CN' engine_traits.languages['zh'] = 'lang_zh-CN'
# supported region codes
for x in eval_xpath_list(dom, '//*[@name="region"]/..//input[@name="region"]'): for x in eval_xpath_list(dom, '//*[@name="region"]/..//input[@name="region"]'):
eng_country = x.get("value") eng_country = x.get("value")
@ -459,4 +485,26 @@ def fetch_traits(engine_traits: EngineTraits):
continue continue
for sxng_locale in sxng_locales: for sxng_locale in sxng_locales:
engine_traits.regions[region_tag(sxng_locale)] = 'country' + eng_country engine_traits.regions[region_tag(sxng_locale)] = eng_country
# alias regions
engine_traits.regions['zh-CN'] = 'HK'
# supported domains
if add_domains:
resp = network.get('https://www.google.com/supported_domains')
if not resp.ok:
raise RuntimeError("Response from https://www.google.com/supported_domains is not OK.")
for domain in resp.text.split():
domain = domain.strip()
if not domain or domain in [
'.google.com',
]:
continue
region = domain.split('.')[-1].upper()
engine_traits.custom['supported_domains'][region] = 'www' + domain
if region == 'HK':
# There is no google.cn, we use .com.hk for zh-CN
engine_traits.custom['supported_domains']['CN'] = 'www' + domain

49
searx/engines/google_images.py
View file

@ -1,31 +1,38 @@
# SPDX-License-Identifier: AGPL-3.0-or-later # SPDX-License-Identifier: AGPL-3.0-or-later
# lint: pylint # lint: pylint
"""This is the implementation of the google images engine using the google """This is the implementation of the Google Images engine using the internal
internal API used the Google Go Android app. Google API used by the Google Go Android app.
This internal API offer results in This internal API offer results in
- JSON (_fmt:json) - JSON (``_fmt:json``)
- Protobuf (_fmt:pb) - Protobuf_ (``_fmt:pb``)
- Protobuf compressed? (_fmt:pc) - Protobuf_ compressed? (``_fmt:pc``)
- HTML (_fmt:html) - HTML (``_fmt:html``)
- Protobuf encoded in JSON (_fmt:jspb). - Protobuf_ encoded in JSON (``_fmt:jspb``).
.. _Protobuf: https://en.wikipedia.org/wiki/Protocol_Buffers
""" """
from typing import TYPE_CHECKING
from urllib.parse import urlencode from urllib.parse import urlencode
from json import loads from json import loads
from searx.engines.google import fetch_traits # pylint: disable=unused-import
from searx.engines.google import ( from searx.engines.google import (
get_lang_info, get_google_info,
time_range_dict, time_range_dict,
detect_google_sorry, detect_google_sorry,
) )
# pylint: disable=unused-import if TYPE_CHECKING:
from searx.engines.google import supported_languages_url, _fetch_supported_languages, fetch_traits import logging
from searx.enginelib.traits import EngineTraits
logger: logging.Logger
traits: EngineTraits
# pylint: enable=unused-import
# about # about
about = { about = {
@ -40,7 +47,6 @@ about = {
# engine dependent config # engine dependent config
categories = ['images', 'web'] categories = ['images', 'web']
paging = True paging = True
use_locale_domain = True
time_range_support = True time_range_support = True
safesearch = True safesearch = True
send_accept_language_header = True send_accept_language_header = True
@ -51,20 +57,18 @@ filter_mapping = {0: 'images', 1: 'active', 2: 'active'}
def request(query, params): def request(query, params):
"""Google-Image search request""" """Google-Image search request"""
lang_info = get_lang_info(params, supported_languages, language_aliases, False) google_info = get_google_info(params, traits)
query_url = ( query_url = (
'https://' 'https://'
+ lang_info['subdomain'] + google_info['subdomain']
+ '/search' + '/search'
+ "?" + "?"
+ urlencode( + urlencode(
{ {
'q': query, 'q': query,
'tbm': "isch", 'tbm': "isch",
**lang_info['params'], **google_info['params'],
'ie': "utf8",
'oe': "utf8",
'asearch': 'isch', 'asearch': 'isch',
'async': '_fmt:json,p:1,ijn:' + str(params['pageno']), 'async': '_fmt:json,p:1,ijn:' + str(params['pageno']),
} }
@ -77,9 +81,8 @@ def request(query, params):
query_url += '&' + urlencode({'safe': filter_mapping[params['safesearch']]}) query_url += '&' + urlencode({'safe': filter_mapping[params['safesearch']]})
params['url'] = query_url params['url'] = query_url
params['headers'].update(lang_info['headers']) params['cookies'] = google_info['cookies']
params['headers']['User-Agent'] = 'NSTN/3.60.474802233.release Dalvik/2.1.0 (Linux; U; Android 12; US) gzip' params['headers'].update(google_info['headers'])
params['headers']['Accept'] = '*/*'
return params return params
@ -111,7 +114,11 @@ def response(resp):
copyright_notice = item["result"].get('iptc', {}).get('copyright_notice') copyright_notice = item["result"].get('iptc', {}).get('copyright_notice')
if copyright_notice: if copyright_notice:
result_item['source'] += ' / ' + copyright_notice result_item['source'] += ' | ' + copyright_notice
freshness_date = item["result"].get("freshness_date")
if freshness_date:
result_item['source'] += ' | ' + freshness_date
file_size = item.get('gsa', {}).get('file_size') file_size = item.get('gsa', {}).get('file_size')
if file_size: if file_size:

253
searx/engines/google_news.py
View file

@ -1,24 +1,40 @@
# SPDX-License-Identifier: AGPL-3.0-or-later # SPDX-License-Identifier: AGPL-3.0-or-later
# lint: pylint # lint: pylint
"""This is the implementation of the google news engine. The google news API """This is the implementation of the Google News engine.
ignores some parameters from the common :ref:`google API`:
- num_ : the number of search results is ignored Google News has a different region handling compared to Google WEB.
- the ``ceid`` argument has to be set (:py:obj:`ceid_list`)
- the hl_ argument has to be set correctly (and different to Google WEB)
- the gl_ argument is mandatory
If one of this argument is not set correctly, the request is redirected to
CONSENT dialog::
https://consent.google.com/m?continue=
The google news API ignores some parameters from the common :ref:`google API`:
- num_ : the number of search results is ignored / there is no paging all
results for a query term are in the first response.
- save_ : is ignored / Google-News results are always *SafeSearch* - save_ : is ignored / Google-News results are always *SafeSearch*
.. _hl: https://developers.google.com/custom-search/docs/xml_results#hlsp
.. _gl: https://developers.google.com/custom-search/docs/xml_results#glsp
.. _num: https://developers.google.com/custom-search/docs/xml_results#numsp .. _num: https://developers.google.com/custom-search/docs/xml_results#numsp
.. _save: https://developers.google.com/custom-search/docs/xml_results#safesp .. _save: https://developers.google.com/custom-search/docs/xml_results#safesp
""" """
# pylint: disable=invalid-name from typing import TYPE_CHECKING
import binascii import binascii
import re import re
from urllib.parse import urlencode from urllib.parse import urlencode
from base64 import b64decode from base64 import b64decode
from lxml import html from lxml import html
import babel
from searx import locales
from searx.utils import ( from searx.utils import (
eval_xpath, eval_xpath,
eval_xpath_list, eval_xpath_list,
@ -26,19 +42,19 @@ from searx.utils import (
extract_text, extract_text,
) )
# pylint: disable=unused-import from searx.engines.google import fetch_traits as _fetch_traits # pylint: disable=unused-import
from searx.engines.google import ( from searx.engines.google import (
fetch_traits, get_google_info,
supported_languages_url,
_fetch_supported_languages,
)
# pylint: enable=unused-import
from searx.engines.google import (
get_lang_info,
detect_google_sorry, detect_google_sorry,
) )
from searx.enginelib.traits import EngineTraits
if TYPE_CHECKING:
import logging
logger: logging.Logger
traits: EngineTraits
# about # about
about = { about = {
@ -50,70 +66,77 @@ about = {
"results": 'HTML', "results": 'HTML',
} }
# compared to other google engines google-news has a different time range
# support. The time range is included in the search term.
time_range_dict = {
'day': 'when:1d',
'week': 'when:7d',
'month': 'when:1m',
'year': 'when:1y',
}
# engine dependent config # engine dependent config
categories = ['news'] categories = ['news']
paging = False paging = False
use_locale_domain = True time_range_support = False
time_range_support = True
# Google-News results are always *SafeSearch*. Option 'safesearch' is set to # Google-News results are always *SafeSearch*. Option 'safesearch' is set to
# False here, otherwise checker will report safesearch-errors:: # False here, otherwise checker will report safesearch-errors::
# #
# safesearch : results are identitical for safesearch=0 and safesearch=2 # safesearch : results are identitical for safesearch=0 and safesearch=2
safesearch = False safesearch = True
send_accept_language_header = True # send_accept_language_header = True
def request(query, params): def request(query, params):
"""Google-News search request""" """Google-News search request"""
lang_info = get_lang_info(params, supported_languages, language_aliases, False) sxng_locale = params.get('searxng_locale', 'en-US')
ceid = locales.get_engine_locale(sxng_locale, traits.custom['ceid'], default='US:en')
google_info = get_google_info(params, traits)
google_info['subdomain'] = 'news.google.com' # google news has only one domain
# google news has only one domain ceid_region, ceid_lang = ceid.split(':')
lang_info['subdomain'] = 'news.google.com' ceid_lang, ceid_suffix = (
ceid_lang.split('-')
+ [
None,
]
)[:2]
ceid = "%s:%s" % (lang_info['country'], lang_info['language']) google_info['params']['hl'] = ceid_lang
# google news redirects en to en-US if ceid_suffix and ceid_suffix not in ['Hans', 'Hant']:
if lang_info['params']['hl'] == 'en':
lang_info['params']['hl'] = 'en-US'
# Very special to google-news compared to other google engines, the time if ceid_region.lower() == ceid_lang:
# range is included in the search term. google_info['params']['hl'] = ceid_lang + '-' + ceid_region
if params['time_range']: else:
query += ' ' + time_range_dict[params['time_range']] google_info['params']['hl'] = ceid_lang + '-' + ceid_suffix
elif ceid_region.lower() != ceid_lang:
if ceid_region in ['AT', 'BE', 'CH', 'IL', 'SA', 'IN', 'BD', 'PT']:
google_info['params']['hl'] = ceid_lang
else:
google_info['params']['hl'] = ceid_lang + '-' + ceid_region
google_info['params']['lr'] = 'lang_' + ceid_lang.split('-')[0]
google_info['params']['gl'] = ceid_region
query_url = ( query_url = (
'https://' 'https://'
+ lang_info['subdomain'] + google_info['subdomain']
+ '/search' + "/search?"
+ "?" + urlencode(
+ urlencode({'q': query, **lang_info['params'], 'ie': "utf8", 'oe': "utf8", 'gl': lang_info['country']}) {
'q': query,
**google_info['params'],
}
)
# ceid includes a ':' character which must not be urlencoded
+ ('&ceid=%s' % ceid) + ('&ceid=%s' % ceid)
) # ceid includes a ':' character which must not be urlencoded )
params['url'] = query_url params['url'] = query_url
params['cookies'] = google_info['cookies']
params['cookies']['CONSENT'] = "YES+" params['headers'].update(google_info['headers'])
params['headers'].update(lang_info['headers'])
params['headers']['Accept'] = 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8'
return params return params
def response(resp): def response(resp):
"""Get response from google's search request""" """Get response from google's search request"""
results = [] results = []
detect_google_sorry(resp) detect_google_sorry(resp)
# convert the text to dom # convert the text to dom
@ -153,8 +176,8 @@ def response(resp):
# The pub_date is mostly a string like 'yesertday', not a real # The pub_date is mostly a string like 'yesertday', not a real
# timezone date or time. Therefore we can't use publishedDate. # timezone date or time. Therefore we can't use publishedDate.
pub_date = extract_text(eval_xpath(result, './article/div[1]/div[1]/time')) pub_date = extract_text(eval_xpath(result, './article//time'))
pub_origin = extract_text(eval_xpath(result, './article/div[1]/div[1]/a')) pub_origin = extract_text(eval_xpath(result, './article//a[@data-n-tid]'))
content = ' / '.join([x for x in [pub_origin, pub_date] if x]) content = ' / '.join([x for x in [pub_origin, pub_date] if x])
@ -175,3 +198,127 @@ def response(resp):
# return results # return results
return results return results
ceid_list = [
'AE:ar',
'AR:es-419',
'AT:de',
'AU:en',
'BD:bn',
'BE:fr',
'BE:nl',
'BG:bg',
'BR:pt-419',
'BW:en',
'CA:en',
'CA:fr',
'CH:de',
'CH:fr',
'CL:es-419',
'CN:zh-Hans',
'CO:es-419',
'CU:es-419',
'CZ:cs',
'DE:de',
'EG:ar',
'ES:es',
'ET:en',
'FR:fr',
'GB:en',
'GH:en',
'GR:el',
'HK:zh-Hant',
'HU:hu',
'ID:en',
'ID:id',
'IE:en',
'IL:en',
'IL:he',
'IN:bn',
'IN:en',
'IN:hi',
'IN:ml',
'IN:mr',
'IN:ta',
'IN:te',
'IT:it',
'JP:ja',
'KE:en',
'KR:ko',
'LB:ar',
'LT:lt',
'LV:en',
'LV:lv',
'MA:fr',
'MX:es-419',
'MY:en',
'NA:en',
'NG:en',
'NL:nl',
'NO:no',
'NZ:en',
'PE:es-419',
'PH:en',
'PK:en',
'PL:pl',
'PT:pt-150',
'RO:ro',
'RS:sr',
'RU:ru',
'SA:ar',
'SE:sv',
'SG:en',
'SI:sl',
'SK:sk',
'SN:fr',
'TH:th',
'TR:tr',
'TW:zh-Hant',
'TZ:en',
'UA:ru',
'UA:uk',
'UG:en',
'US:en',
'US:es-419',
'VE:es-419',
'VN:vi',
'ZA:en',
'ZW:en',
]
"""List of region/language combinations supported by Google News. Values of the
``ceid`` argument of the Google News REST API."""
_skip_values = [
'ET:en', # english (ethiopia)
'ID:en', # english (indonesia)
'LV:en', # english (latvia)
]
_ceid_locale_map = {'NO:no': 'nb-NO'}
def fetch_traits(engine_traits: EngineTraits):
_fetch_traits(engine_traits, add_domains=False)
engine_traits.custom['ceid'] = {}
for ceid in ceid_list:
if ceid in _skip_values:
continue
region, lang = ceid.split(':')
x = lang.split('-')
if len(x) > 1:
if x[1] not in ['Hant', 'Hans']:
lang = x[0]
sxng_locale = _ceid_locale_map.get(ceid, lang + '-' + region)
try:
locale = babel.Locale.parse(sxng_locale, sep='-')
except babel.UnknownLocaleError:
print("ERROR: %s -> %s is unknown by babel" % (ceid, sxng_locale))
continue
engine_traits.custom['ceid'][locales.region_tag(locale)] = ceid

117
searx/engines/google_scholar.py
View file

@ -1,19 +1,18 @@
# SPDX-License-Identifier: AGPL-3.0-or-later # SPDX-License-Identifier: AGPL-3.0-or-later
# lint: pylint # lint: pylint
"""Google (Scholar) """This is the implementation of the Google Scholar engine.
For detailed description of the *REST-full* API see: `Query Parameter Compared to other Google services the Scholar engine has a simple GET REST-API
Definitions`_. and there does not exists `async` API. Even though the API slightly vintage we
can make use of the :ref:`google API` to assemble the arguments of the GET
.. _Query Parameter Definitions: request.
https://developers.google.com/custom-search/docs/xml_results#WebSearch_Query_Parameter_Definitions
""" """
# pylint: disable=invalid-name from typing import TYPE_CHECKING
from typing import Optional
from urllib.parse import urlencode from urllib.parse import urlencode
from datetime import datetime from datetime import datetime
from typing import Optional
from lxml import html from lxml import html
from searx.utils import ( from searx.utils import (
@ -23,20 +22,21 @@ from searx.utils import (
extract_text, extract_text,
) )
from searx.exceptions import SearxEngineCaptchaException
from searx.engines.google import fetch_traits # pylint: disable=unused-import
from searx.engines.google import ( from searx.engines.google import (
get_lang_info, get_google_info,
time_range_dict, time_range_dict,
detect_google_sorry,
) )
from searx.enginelib.traits import EngineTraits
# pylint: disable=unused-import if TYPE_CHECKING:
from searx.engines.google import ( import logging
fetch_traits,
supported_languages_url,
_fetch_supported_languages,
)
# pylint: enable=unused-import logger: logging.Logger
traits: EngineTraits
# about # about
about = { about = {
@ -52,53 +52,62 @@ about = {
categories = ['science', 'scientific publications'] categories = ['science', 'scientific publications']
paging = True paging = True
language_support = True language_support = True
use_locale_domain = True
time_range_support = True time_range_support = True
safesearch = False safesearch = False
send_accept_language_header = True send_accept_language_header = True
def time_range_url(params): def time_range_args(params):
"""Returns a URL query component for a google-Scholar time range based on """Returns a dictionary with a time range arguments based on
``params['time_range']``. Google-Scholar does only support ranges in years. ``params['time_range']``.
To have any effect, all the Searx ranges (*day*, *week*, *month*, *year*)
are mapped to *year*. If no range is set, an empty string is returned. Google Scholar supports a detailed search by year. Searching by *last
Example:: month* or *last week* (as offered by SearXNG) is uncommon for scientific
publications and is not supported by Google Scholar.
To limit the result list when the users selects a range, all the SearXNG
ranges (*day*, *week*, *month*, *year*) are mapped to *year*. If no range
is set an empty dictionary of arguments is returned. Example; when
user selects a time range (current year minus one in 2022):
.. code:: python
{ 'as_ylo' : 2021 }
&as_ylo=2019
""" """
# as_ylo=2016&as_yhi=2019 ret_val = {}
ret_val = ''
if params['time_range'] in time_range_dict: if params['time_range'] in time_range_dict:
ret_val = urlencode({'as_ylo': datetime.now().year - 1}) ret_val['as_ylo'] = datetime.now().year - 1
return '&' + ret_val return ret_val
def detect_google_captcha(dom):
"""In case of CAPTCHA Google Scholar open its own *not a Robot* dialog and is
not redirected to ``sorry.google.com``.
"""
if eval_xpath(dom, "//form[@id='gs_captcha_f']"):
raise SearxEngineCaptchaException()
def request(query, params): def request(query, params):
"""Google-Scholar search request""" """Google-Scholar search request"""
offset = (params['pageno'] - 1) * 10 google_info = get_google_info(params, traits)
lang_info = get_lang_info(params, supported_languages, language_aliases, False)
# subdomain is: scholar.google.xy # subdomain is: scholar.google.xy
lang_info['subdomain'] = lang_info['subdomain'].replace("www.", "scholar.") google_info['subdomain'] = google_info['subdomain'].replace("www.", "scholar.")
query_url = ( args = {
'https://' 'q': query,
+ lang_info['subdomain'] **google_info['params'],
+ '/scholar' 'start': (params['pageno'] - 1) * 10,
+ "?" 'as_sdt': '2007', # include patents / to disable set '0,5'
+ urlencode({'q': query, **lang_info['params'], 'ie': "utf8", 'oe': "utf8", 'start': offset}) 'as_vis': '0', # include citations / to disable set '1'
) }
args.update(time_range_args(params))
query_url += time_range_url(params) params['url'] = 'https://' + google_info['subdomain'] + '/scholar?' + urlencode(args)
params['url'] = query_url params['cookies'] = google_info['cookies']
params['headers'].update(google_info['headers'])
params['cookies']['CONSENT'] = "YES+"
params['headers'].update(lang_info['headers'])
params['headers']['Accept'] = 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8'
# params['google_subdomain'] = subdomain
return params return params
@ -139,19 +148,15 @@ def parse_gs_a(text: Optional[str]):
def response(resp): # pylint: disable=too-many-locals def response(resp): # pylint: disable=too-many-locals
"""Get response from google's search request""" """Parse response from Google Scholar"""
results = [] results = []
detect_google_sorry(resp)
# which subdomain ?
# subdomain = resp.search_params.get('google_subdomain')
# convert the text to dom # convert the text to dom
dom = html.fromstring(resp.text) dom = html.fromstring(resp.text)
detect_google_captcha(dom)
# parse results # parse results
for result in eval_xpath_list(dom, '//div[@data-cid]'): for result in eval_xpath_list(dom, '//div[@data-rp]'):
title = extract_text(eval_xpath(result, './/h3[1]//a')) title = extract_text(eval_xpath(result, './/h3[1]//a'))
@ -159,7 +164,7 @@ def response(resp): # pylint: disable=too-many-locals
# this is a [ZITATION] block # this is a [ZITATION] block
continue continue
pub_type = extract_text(eval_xpath(result, './/span[@class="gs_ct1"]')) pub_type = extract_text(eval_xpath(result, './/span[@class="gs_ctg2"]'))
if pub_type: if pub_type:
pub_type = pub_type[1:-1].lower() pub_type = pub_type[1:-1].lower()

115
searx/engines/google_videos.py
View file

@ -1,6 +1,6 @@
# SPDX-License-Identifier: AGPL-3.0-or-later # SPDX-License-Identifier: AGPL-3.0-or-later
# lint: pylint # lint: pylint
"""This is the implementation of the google videos engine. """This is the implementation of the Google Videos engine.
.. admonition:: Content-Security-Policy (CSP) .. admonition:: Content-Security-Policy (CSP)
@ -14,9 +14,8 @@
""" """
# pylint: disable=invalid-name from typing import TYPE_CHECKING
import re
from urllib.parse import urlencode from urllib.parse import urlencode
from lxml import html from lxml import html
@ -27,20 +26,22 @@ from searx.utils import (
extract_text, extract_text,
) )
from searx.engines.google import fetch_traits # pylint: disable=unused-import
from searx.engines.google import ( from searx.engines.google import (
get_lang_info, get_google_info,
time_range_dict, time_range_dict,
filter_mapping, filter_mapping,
g_section_with_header,
title_xpath,
suggestion_xpath, suggestion_xpath,
detect_google_sorry, detect_google_sorry,
) )
from searx.enginelib.traits import EngineTraits
# pylint: disable=unused-import if TYPE_CHECKING:
from searx.engines.google import supported_languages_url, _fetch_supported_languages, fetch_traits import logging
# pylint: enable=unused-import logger: logging.Logger
traits: EngineTraits
# about # about
about = { about = {
@ -55,70 +56,32 @@ about = {
# engine dependent config # engine dependent config
categories = ['videos', 'web'] categories = ['videos', 'web']
paging = False paging = True
language_support = True language_support = True
use_locale_domain = True
time_range_support = True time_range_support = True
safesearch = True safesearch = True
send_accept_language_header = True
RE_CACHE = {}
def _re(regexpr):
"""returns compiled regular expression"""
RE_CACHE[regexpr] = RE_CACHE.get(regexpr, re.compile(regexpr))
return RE_CACHE[regexpr]
def scrap_out_thumbs_src(dom):
ret_val = {}
thumb_name = 'dimg_'
for script in eval_xpath_list(dom, '//script[contains(., "google.ldi={")]'):
_script = script.text
# "dimg_35":"https://i.ytimg.c....",
_dimurl = _re("s='([^']*)").findall(_script)
for k, v in _re('(' + thumb_name + '[0-9]*)":"(http[^"]*)').findall(_script):
v = v.replace(r'\u003d', '=')
v = v.replace(r'\u0026', '&')
ret_val[k] = v
logger.debug("found %s imgdata for: %s", thumb_name, ret_val.keys())
return ret_val
def scrap_out_thumbs(dom):
"""Scrap out thumbnail data from <script> tags."""
ret_val = {}
thumb_name = 'dimg_'
for script in eval_xpath_list(dom, '//script[contains(., "_setImagesSrc")]'):
_script = script.text
# var s='data:image/jpeg;base64, ...'
_imgdata = _re("s='([^']*)").findall(_script)
if not _imgdata:
continue
# var ii=['dimg_17']
for _vidthumb in _re(r"(%s\d+)" % thumb_name).findall(_script):
# At least the equal sign in the URL needs to be decoded
ret_val[_vidthumb] = _imgdata[0].replace(r"\x3d", "=")
logger.debug("found %s imgdata for: %s", thumb_name, ret_val.keys())
return ret_val
def request(query, params): def request(query, params):
"""Google-Video search request""" """Google-Video search request"""
lang_info = get_lang_info(params, supported_languages, language_aliases, False) google_info = get_google_info(params, traits)
query_url = ( query_url = (
'https://' 'https://'
+ lang_info['subdomain'] + google_info['subdomain']
+ '/search' + '/search'
+ "?" + "?"
+ urlencode({'q': query, 'tbm': "vid", **lang_info['params'], 'ie': "utf8", 'oe': "utf8"}) + urlencode(
{
'q': query,
'tbm': "vid",
'start': 10 * params['pageno'],
**google_info['params'],
'asearch': 'arc',
'async': 'use_ac:true,_fmt:html',
}
)
) )
if params['time_range'] in time_range_dict: if params['time_range'] in time_range_dict:
@ -127,9 +90,8 @@ def request(query, params):
query_url += '&' + urlencode({'safe': filter_mapping[params['safesearch']]}) query_url += '&' + urlencode({'safe': filter_mapping[params['safesearch']]})
params['url'] = query_url params['url'] = query_url
params['cookies']['CONSENT'] = "YES+" params['cookies'] = google_info['cookies']
params['headers'].update(lang_info['headers']) params['headers'].update(google_info['headers'])
params['headers']['Accept'] = 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8'
return params return params
@ -141,43 +103,30 @@ def response(resp):
# convert the text to dom # convert the text to dom
dom = html.fromstring(resp.text) dom = html.fromstring(resp.text)
vidthumb_imgdata = scrap_out_thumbs(dom)
thumbs_src = scrap_out_thumbs_src(dom)
logger.debug(str(thumbs_src))
# parse results # parse results
for result in eval_xpath_list(dom, '//div[contains(@class, "g ")]'): for result in eval_xpath_list(dom, '//div[contains(@class, "g ")]'):
# ignore google *sections* img_src = eval_xpath_getindex(result, './/img/@src', 0, None)
if extract_text(eval_xpath(result, g_section_with_header)): if img_src is None:
logger.debug("ignoring <g-section-with-header>")
continue continue
# ingnore articles without an image id / e.g. news articles title = extract_text(eval_xpath_getindex(result, './/a/h3[1]', 0))
img_id = eval_xpath_getindex(result, './/g-img/img/@id', 0, default=None) url = eval_xpath_getindex(result, './/a/h3[1]/../@href', 0)
if img_id is None:
logger.error("no img_id found in item %s (news article?)", len(results) + 1)
continue
img_src = vidthumb_imgdata.get(img_id, None)
if not img_src:
img_src = thumbs_src.get(img_id, "")
title = extract_text(eval_xpath_getindex(result, title_xpath, 0))
url = eval_xpath_getindex(result, './/div[@class="dXiKIc"]//a/@href', 0)
length = extract_text(eval_xpath(result, './/div[contains(@class, "P7xzyf")]/span/span'))
c_node = eval_xpath_getindex(result, './/div[@class="Uroaid"]', 0) c_node = eval_xpath_getindex(result, './/div[@class="Uroaid"]', 0)
content = extract_text(c_node) content = extract_text(c_node)
pub_info = extract_text(eval_xpath(result, './/div[@class="Zg1NU"]')) pub_info = extract_text(eval_xpath(result, './/div[@class="P7xzyf"]'))
length = extract_text(eval_xpath(result, './/div[@class="J1mWY"]'))
results.append( results.append(
{ {
'url': url, 'url': url,
'title': title, 'title': title,
'content': content, 'content': content,
'length': length,
'author': pub_info, 'author': pub_info,
'thumbnail': img_src, 'thumbnail': img_src,
'length': length,
'template': 'videos.html', 'template': 'videos.html',
} }
) )

5
searx/search/processors/online.py
View file

@ -187,11 +187,6 @@ class OnlineProcessor(EngineProcessor):
self.handle_exception(result_container, e, suspend=True) self.handle_exception(result_container, e, suspend=True)
self.logger.exception('CAPTCHA') self.logger.exception('CAPTCHA')
except SearxEngineTooManyRequestsException as e: except SearxEngineTooManyRequestsException as e:
if "google" in self.engine_name:
self.logger.warn(
"Set to 'true' the use_mobile_ui parameter in the 'engines:'"
" section of your settings.yml file if google is blocked for you."
)
self.handle_exception(result_container, e, suspend=True) self.handle_exception(result_container, e, suspend=True)
self.logger.exception('Too many requests') self.logger.exception('Too many requests')
except SearxEngineAccessDeniedException as e: except SearxEngineAccessDeniedException as e:

13
searx/settings.yml
View file

@ -731,22 +731,9 @@ engines:
- name: google - name: google
engine: google engine: google
shortcut: go shortcut: go
# see https://docs.searxng.org/src/searx.engines.google.html#module-searx.engines.google
use_mobile_ui: false
# additional_tests: # additional_tests:
# android: *test_android # android: *test_android
# - name: google italian
# engine: google
# shortcut: goit
# use_mobile_ui: false
# language: it
# - name: google mobile ui
# engine: google
# shortcut: gomui
# use_mobile_ui: true
- name: google images - name: google images
engine: google_images engine: google_images
shortcut: goi shortcut: goi

3
utils/templates/etc/searxng/settings.yml
View file

@ -52,9 +52,6 @@ enabled_plugins:
engines: engines:
- name: google
use_mobile_ui: true
# - name: fdroid # - name: fdroid
# disabled: false # disabled: false
# #