From 7d8c20c80d86ac47eac86f11533fee038cd48190 Mon Sep 17 00:00:00 2001 From: Bnyro Date: Thu, 3 Aug 2023 19:07:22 +0200 Subject: [PATCH] [feat] new engine: wikispecies --- searx/engines/mediawiki.py | 7 ++++--- searx/settings.yml | 12 ++++++++++++ 2 files changed, 16 insertions(+), 3 deletions(-) diff --git a/searx/engines/mediawiki.py b/searx/engines/mediawiki.py index 9002e9ba7..27ce36e87 100644 --- a/searx/engines/mediawiki.py +++ b/searx/engines/mediawiki.py @@ -3,10 +3,11 @@ General mediawiki-engine (Web) """ -from json import loads from string import Formatter from urllib.parse import urlencode, quote +from searx.utils import html_to_text + # about about = { "website": None, @@ -69,7 +70,7 @@ def request(query, params): def response(resp): results = [] - search_results = loads(resp.text) + search_results = resp.json() # return empty array if there are no results if not search_results.get('query', {}).get('search'): @@ -86,7 +87,7 @@ def response(resp): ) # append result - results.append({'url': url, 'title': result['title'], 'content': ''}) + results.append({'url': url, 'title': result['title'], 'content': html_to_text(result.get('snippet', ''))}) # return results return results diff --git a/searx/settings.yml b/searx/settings.yml index 209b32b4d..e498a9ea2 100644 --- a/searx/settings.yml +++ b/searx/settings.yml @@ -1579,6 +1579,18 @@ engines: website: https://www.wikisource.org/ wikidata_id: Q263 + - name: wikipecies + engine: mediawiki + shortcut: wsp + categories: [general, science] + base_url: "https://species.wikimedia.org/" + number_of_results: 5 + search_type: text + disabled: true + about: + website: https://species.wikimedia.org/ + wikidata_id: Q13679 + - name: wiktionary engine: mediawiki shortcut: wt