mirror of
https://github.com/searxng/searxng.git
synced 2024-11-26 12:51:02 +00:00
916739d6b4
The use of img_src AND thumbnail in the default results makes no sense (only a thumbnail is needed). In the current state this is rather confusing, because img_src is displayed like a thumbnail (small) and thumbnail is displayed like an image (large). Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
115 lines
3.5 KiB
Python
115 lines
3.5 KiB
Python
# SPDX-License-Identifier: AGPL-3.0-or-later
|
|
"""Google Play Apps & Google Play Movies
|
|
"""
|
|
|
|
from urllib.parse import urlencode
|
|
from lxml import html
|
|
from searx.utils import (
|
|
eval_xpath,
|
|
extract_url,
|
|
extract_text,
|
|
eval_xpath_list,
|
|
eval_xpath_getindex,
|
|
)
|
|
|
|
about = {
|
|
"website": "https://play.google.com/",
|
|
"wikidata_id": "Q79576",
|
|
"use_official_api": False,
|
|
"require_api_key": False,
|
|
"results": "HTML",
|
|
}
|
|
|
|
send_accept_language_header = True
|
|
|
|
play_categ = None # apps|movies
|
|
base_url = 'https://play.google.com'
|
|
search_url = base_url + "/store/search?{query}&c={play_categ}"
|
|
|
|
|
|
def request(query, params):
|
|
|
|
if play_categ not in ('movies', 'apps'):
|
|
raise ValueError(f"unknown google play category: {play_categ}")
|
|
|
|
params["url"] = search_url.format(
|
|
query=urlencode({"q": query}),
|
|
play_categ=play_categ,
|
|
)
|
|
params['cookies']['CONSENT'] = "YES+"
|
|
|
|
return params
|
|
|
|
|
|
def response(resp):
|
|
|
|
if play_categ == 'movies':
|
|
return response_movies(resp)
|
|
if play_categ == 'apps':
|
|
return response_apps(resp)
|
|
|
|
raise ValueError(f"Unsupported play category: {play_categ}")
|
|
|
|
|
|
def response_movies(resp):
|
|
|
|
results = []
|
|
dom = html.fromstring(resp.text)
|
|
|
|
for section in eval_xpath(dom, '//c-wiz/section/header/..'):
|
|
sec_name = extract_text(eval_xpath(section, './header'))
|
|
for item in eval_xpath(section, './/a'):
|
|
url = base_url + item.get('href')
|
|
div_1, div_2 = eval_xpath(item, './div')[:2]
|
|
title = extract_text(eval_xpath(div_2, './div[@title]'))
|
|
metadata = extract_text(eval_xpath(div_2, './div[@class]'))
|
|
img = eval_xpath(div_1, './/img')[0]
|
|
thumbnail = img.get('src')
|
|
results.append(
|
|
{
|
|
"url": url,
|
|
"title": title,
|
|
"content": sec_name,
|
|
"thumbnail": thumbnail,
|
|
'metadata': metadata,
|
|
'template': 'videos.html',
|
|
}
|
|
)
|
|
return results
|
|
|
|
|
|
def response_apps(resp):
|
|
|
|
results = []
|
|
dom = html.fromstring(resp.text)
|
|
|
|
if eval_xpath(dom, '//div[@class="v6DsQb"]'):
|
|
return []
|
|
|
|
spot = eval_xpath_getindex(dom, '//div[@class="ipRz4"]', 0, None)
|
|
if spot is not None:
|
|
url = extract_url(eval_xpath(spot, './a[@class="Qfxief"]/@href'), search_url)
|
|
title = extract_text(eval_xpath(spot, './/div[@class="vWM94c"]'))
|
|
content = extract_text(eval_xpath(spot, './/div[@class="LbQbAe"]'))
|
|
img = extract_text(eval_xpath(spot, './/img[@class="T75of bzqKMd"]/@src'))
|
|
|
|
results.append({"url": url, "title": title, "content": content, "img_src": img})
|
|
|
|
more = eval_xpath_list(dom, '//c-wiz[@jsrenderer="RBsfwb"]//div[@role="listitem"]', min_len=1)
|
|
for result in more:
|
|
url = extract_url(eval_xpath(result, ".//a/@href"), search_url)
|
|
title = extract_text(eval_xpath(result, './/span[@class="DdYX5"]'))
|
|
content = extract_text(eval_xpath(result, './/span[@class="wMUdtb"]'))
|
|
img = extract_text(
|
|
eval_xpath(
|
|
result,
|
|
'.//img[@class="T75of stzEZd" or @class="T75of etjhNc Q8CSx "]/@src',
|
|
)
|
|
)
|
|
|
|
results.append({"url": url, "title": title, "content": content, "img_src": img})
|
|
|
|
for suggestion in eval_xpath_list(dom, '//c-wiz[@jsrenderer="qyd4Kb"]//div[@class="ULeU3b neq64b"]'):
|
|
results.append({"suggestion": extract_text(eval_xpath(suggestion, './/div[@class="Epkrse "]'))})
|
|
|
|
return results
|