support language selection, detect google sorry, etc

This commit is contained in:
ta 2022-08-27 17:16:00 +07:00
parent 692855b442
commit 1dc6c86ffe
2 changed files with 162 additions and 0 deletions

View file

@ -1234,6 +1234,146 @@
"name": "\u4e2d\u6587 (\u7e41\u9ad4)"
}
},
"google shopping": {
"af": {
"name": "Afrikaans"
},
"ar": {
"name": "\u0627\u0644\u0639\u0631\u0628\u064a\u0629"
},
"be": {
"name": "\u0431\u0435\u043b\u0430\u0440\u0443\u0441\u043a\u0430\u044f"
},
"bg": {
"name": "\u0431\u044a\u043b\u0433\u0430\u0440\u0441\u043a\u0438"
},
"ca": {
"name": "catal\u00e0"
},
"cs": {
"name": "\u010de\u0161tina"
},
"da": {
"name": "dansk"
},
"de": {
"name": "Deutsch"
},
"el": {
"name": "\u0395\u03bb\u03bb\u03b7\u03bd\u03b9\u03ba\u03ac"
},
"en": {
"name": "English"
},
"eo": {
"name": "esperanto"
},
"es": {
"name": "espa\u00f1ol"
},
"et": {
"name": "eesti"
},
"fa": {
"name": "\u0641\u0627\u0631\u0633\u06cc"
},
"fi": {
"name": "suomi"
},
"fr": {
"name": "fran\u00e7ais"
},
"hi": {
"name": "\u0939\u093f\u0928\u094d\u0926\u0940"
},
"hr": {
"name": "hrvatski"
},
"hu": {
"name": "magyar"
},
"hy": {
"name": "\u0570\u0561\u0575\u0565\u0580\u0565\u0576"
},
"id": {
"name": "Indonesia"
},
"is": {
"name": "\u00edslenska"
},
"it": {
"name": "italiano"
},
"iw": {
"name": "\u05e2\u05d1\u05e8\u05d9\u05ea"
},
"ja": {
"name": "\u65e5\u672c\u8a9e"
},
"ko": {
"name": "\ud55c\uad6d\uc5b4"
},
"lt": {
"name": "lietuvi\u0173"
},
"lv": {
"name": "latvie\u0161u"
},
"nl": {
"name": "Nederlands"
},
"no": {
"name": "norsk"
},
"pl": {
"name": "polski"
},
"pt": {
"name": "portugu\u00eas"
},
"ro": {
"name": "rom\u00e2n\u0103"
},
"ru": {
"name": "\u0440\u0443\u0441\u0441\u043a\u0438\u0439"
},
"sk": {
"name": "sloven\u010dina"
},
"sl": {
"name": "sloven\u0161\u010dina"
},
"sr": {
"name": "\u0441\u0440\u043f\u0441\u043a\u0438"
},
"sv": {
"name": "svenska"
},
"sw": {
"name": "Kiswahili"
},
"th": {
"name": "\u0e44\u0e17\u0e22"
},
"tl": {
"name": "Filipino"
},
"tr": {
"name": "T\u00fcrk\u00e7e"
},
"uk": {
"name": "\u0443\u043a\u0440\u0430\u0457\u043d\u0441\u044c\u043a\u0430"
},
"vi": {
"name": "Ti\u1ebfng Vi\u1ec7t"
},
"zh-CN": {
"name": "\u4e2d\u6587 (\u7b80\u4f53)"
},
"zh-TW": {
"name": "\u4e2d\u6587 (\u7e41\u9ad4)"
}
},
"google videos": {
"af": {
"name": "Afrikaans"

View file

@ -6,6 +6,19 @@ from urllib.parse import urlencode
from lxml import html
from searx.utils import extract_text
from searx.engines.google import (
get_lang_info,
detect_google_sorry,
)
# pylint: disable=unused-import
from searx.engines.google import (
supported_languages_url,
_fetch_supported_languages,
)
# pylint: enable=unused-import
about = {
"website": "https://shopping.google.com",
"wikidata_id": "Q1433417",
@ -31,14 +44,23 @@ condition_xpath = './/span[@class="JkJxid HFeBod"]'
def request(query, params):
pageno = (params["pageno"] - 1) * 60
lang_info = get_lang_info(params, supported_languages, language_aliases, False)
params["url"] = search_url.format(query=urlencode({"q": query}), pageno=pageno)
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['cookies']['CONSENT'] = "YES+"
return params
def response(resp):
results = []
detect_google_sorry(resp)
dom = html.fromstring(resp.text)
res = dom.xpath(results_xpath)