zaclys/searxng
1
0
Fork 1
mirror of https://github.com/searxng/searxng synced 2024-01-01 19:24:07 +01:00
Code Issues Projects Releases Packages Wiki Activity

[prep] indent Wikipedia functions

Browse source 
This commit indents the methods from the wikipedia engine
so that the following refactoring commit has a clean diff.
This commit is contained in:
Martin Fischer 2022-01-28 12:50:42 +01:00
parent 5dd28ff04b
commit 1e97bfab70
1 changed files with 52 additions and 52 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

104
searx/engines/wikipedia.py
View file

@ -25,74 +25,74 @@ supported_languages_url = 'https://meta.wikimedia.org/wiki/List_of_Wikipedias'
language_variants = {"zh": ("zh-cn", "zh-hk", "zh-mo", "zh-my", "zh-sg", "zh-tw")} language_variants = {"zh": ("zh-cn", "zh-hk", "zh-mo", "zh-my", "zh-sg", "zh-tw")}
# set language in base_url # set language in base_url
def url_lang(lang): def url_lang(lang):
lang_pre = lang.split('-')[0] lang_pre = lang.split('-')[0]
if lang_pre == 'all' or lang_pre not in supported_languages and lang_pre not in language_aliases: if lang_pre == 'all' or lang_pre not in supported_languages and lang_pre not in language_aliases:
return 'en' return 'en'
return match_language(lang, supported_languages, language_aliases).split('-')[0] return match_language(lang, supported_languages, language_aliases).split('-')[0]
# do search-request # do search-request
def request(query, params): def request(query, params):
if query.islower(): if query.islower():
query = query.title() query = query.title()
language = url_lang(params['language']) language = url_lang(params['language'])
params['url'] = search_url.format(title=quote(query), language=language) params['url'] = search_url.format(title=quote(query), language=language)
if params['language'].lower() in language_variants.get(language, []): if params['language'].lower() in language_variants.get(language, []):
params['headers']['Accept-Language'] = params['language'].lower() params['headers']['Accept-Language'] = params['language'].lower()
params['headers']['User-Agent'] = searx_useragent() params['headers']['User-Agent'] = searx_useragent()
params['raise_for_httperror'] = False params['raise_for_httperror'] = False
params['soft_max_redirects'] = 2 params['soft_max_redirects'] = 2
return params return params
# get response from search-request # get response from search-request
def response(resp): def response(resp):
if resp.status_code == 404: if resp.status_code == 404:
return [] return []
if resp.status_code == 400: if resp.status_code == 400:
try: try:
api_result = loads(resp.text) api_result = loads(resp.text)
except: except:
pass pass
else: else:
if ( if (
api_result['type'] == 'https://mediawiki.org/wiki/HyperSwitch/errors/bad_request' api_result['type'] == 'https://mediawiki.org/wiki/HyperSwitch/errors/bad_request'
and api_result['detail'] == 'title-invalid-characters' and api_result['detail'] == 'title-invalid-characters'
): ):
return [] return []
raise_for_httperror(resp) raise_for_httperror(resp)
results = [] results = []
api_result = loads(resp.text) api_result = loads(resp.text)
# skip disambiguation pages # skip disambiguation pages
if api_result.get('type') != 'standard': if api_result.get('type') != 'standard':
return [] return []
title = api_result['title'] title = api_result['title']
wikipedia_link = api_result['content_urls']['desktop']['page'] wikipedia_link = api_result['content_urls']['desktop']['page']
results.append({'url': wikipedia_link, 'title': title}) results.append({'url': wikipedia_link, 'title': title})
results.append( results.append(
{ {
'infobox': title, 'infobox': title,
'id': wikipedia_link, 'id': wikipedia_link,
'content': api_result.get('extract', ''), 'content': api_result.get('extract', ''),
'img_src': api_result.get('thumbnail', {}).get('source'), 'img_src': api_result.get('thumbnail', {}).get('source'),
'urls': [{'title': 'Wikipedia', 'url': wikipedia_link}], 'urls': [{'title': 'Wikipedia', 'url': wikipedia_link}],
} }
) )
return results return results
# get supported languages from their site # get supported languages from their site