From c00c0c54343e16540439341158affaf78b130c19 Mon Sep 17 00:00:00 2001 From: allendema_searxng_pi Date: Thu, 3 Aug 2023 19:15:24 +0200 Subject: [PATCH] [mod] remove discontinued petalsearch engines --- searx/engines/petal_images.py | 94 ----------------------------------- searx/settings.yml | 27 ---------- 2 files changed, 121 deletions(-) delete mode 100644 searx/engines/petal_images.py diff --git a/searx/engines/petal_images.py b/searx/engines/petal_images.py deleted file mode 100644 index 88853c1bd..000000000 --- a/searx/engines/petal_images.py +++ /dev/null @@ -1,94 +0,0 @@ -# SPDX-License-Identifier: AGPL-3.0-or-later -"""Petalsearch Images - -""" - -from json import loads -from urllib.parse import urlencode -from datetime import datetime - -from lxml import html - -from searx.utils import extract_text - -about = { - "website": 'https://petalsearch.com/', - "wikidata_id": 'Q104399280', - "official_api_documentation": False, - "use_official_api": False, - "require_api_key": False, - "results": 'JSON', -} - -categories = ['images'] -paging = True -time_range_support = False - -safesearch = True -safesearch_table = {0: 'off', 1: 'moderate', 2: 'on'} - -base_url = 'https://petalsearch.com/' -search_string = 'search?{query}&channel=image&ps=50&pn={page}®ion={lang}&ss_mode={safesearch}&ss_type=normal' - - -def request(query, params): - - search_path = search_string.format( - query=urlencode({'query': query}), - page=params['pageno'], - lang=params['language'].lower(), - safesearch=safesearch_table[params['safesearch']], - ) - - params['url'] = base_url + search_path - - return params - - -def response(resp): - results = [] - - tree = html.fromstring(resp.text) - root = tree.findall('.//script[3]') - - # Convert list to JSON - json_content = extract_text(root) - - # Manipulate with JSON - data = loads(json_content) - - for result in data['newImages']: - url = result['url'] - title = result['title'] - thumbnail_src = result['image'] - - pic_dict = result.get('extrainfo') - - date_from_api = pic_dict.get('publish_time') - width = pic_dict.get('width') - height = pic_dict.get('height') - img_src = pic_dict.get('real_url') - - # Continue if img_src is missing - if img_src is None or '': - continue - - # Get and convert published date - if date_from_api is not None: - publishedDate = datetime.fromtimestamp(int(date_from_api)) - - # Append results - results.append( - { - 'template': 'images.html', - 'url': url, - 'title': title, - 'img_src': img_src, - 'thumbnail_src': thumbnail_src, - 'width': width, - 'height': height, - 'publishedDate': publishedDate, - } - ) - - return results diff --git a/searx/settings.yml b/searx/settings.yml index 04ec3e466..2ac29351d 100644 --- a/searx/settings.yml +++ b/searx/settings.yml @@ -1845,33 +1845,6 @@ engines: categories: news brave_category: news - - name: petalsearch - shortcut: pts - engine: xpath - paging: true - search_url: https://petalsearch.com/search?query={query}&pn={pageno} - url_xpath: //div[@class='card-source'] - title_xpath: //div[@class='title-name'] - content_xpath: //div[@class='webpage-text'] - first_page_num: 1 - disabled: true - headers: - User-Agent: Mozilla/5.0 (Linux; Android 7.0;) \ - AppleWebKit/537.36 (KHTML, like Gecko) \ - Mobile Safari/537.36 (compatible; PetalBot;+https://webmaster.petalsearch.com/site/petalbot) - about: - website: https://petalsearch.com/ - wikidata_id: Q104399280 - use_official_api: false - require_api_key: false - results: HTML - - - name: petalsearch images - engine: petal_images - shortcut: ptsi - disabled: true - timeout: 3.0 - - name: lib.rs shortcut: lrs engine: xpath