Merge remote-tracking branch 'upstream/master'

bbc.co.uk.txt

@@ -90,3 +90,11 @@ test_url: http://feeds.bbci.co.uk/news/rss.xml
 test_url: http://feeds.bbci.co.uk/sport/0/football/rss.xml?edition=int
 # video entry
 test_url: http://www.bbc.co.uk/news/world-asia-22056933
+
+# bbcthree long articles / gel (?) articles?
+title: //h1[contains(concat(' ',normalize-space(@class),' '),' LongArticle-headline ')]
+body: //div[contains(concat(' ',normalize-space(@class),' '),' LongArticle-body ')]
+author: //span[contains(concat(' ',normalize-space(@class),' '),' Info-authorName ')]
+date: //time[contains(concat(' ',normalize-space(@class),' '),' Info-timestamp ')]
+test_url: https://www.bbc.co.uk/bbcthree/article/991ea351-1e67-46dc-824d-a13033526ca6
+test_contains: In essence, Professor Rippon believes that our 'male' and 'female' behaviour is not innate but learned as we develop.

bbc.com.txt

@@ -1,5 +1,5 @@
 body: //article[contains(@class, "-ArticleWrapper")]
-body: (//div[contains(@class, 'article__body-content')])[1]
+body: //article[@class='article__body'] | (//div[contains(@class, 'article__body-content')])[1]
 body: (//div[@id="story-body"])[1]
 
 # for video entries

cleafy.com.txt

@@ -0,0 +1,6 @@
+body: //div[contains(@class, 'full blog')]
+title: //h1[@class='clabtitle']
+
+strip_id_or_class: author-block
+
+test_url: https://www.cleafy.com/cleafy-labs/on-device-fraud-on-the-rise-exposing-a-recent-copybara-fraud-campaign

digital-photography-school.com.txt

@@ -5,4 +5,6 @@ body: //section[@class='content']
 strip: //div[@class='post-meta']
 strip: //div[contains(@class, 'topics-list')]
 
+replace_string(wp-block-heading): 
+
 test_url: http://www.digital-photography-school.com/10-ways-to-develop-yourself-photographically

fmhy.net.txt

@@ -0,0 +1,11 @@
+body: //main[@class='main']
+
+strip: //h1[1]
+
+prune: no
+
+# for wallabag
+wrap_in(blockquote): //div[@class='tip custom-block']
+wrap_in(blockquote): //div[@class='warning custom-block']
+
+test_url: https://fmhy.net/unsafesites

labs.ripe.net.txt

@@ -0,0 +1,4 @@
+body: //article[@class="prose"]
+
+test_url: https://labs.ripe.net/author/kathleen_moriarty/the-llm-misinformation-problem-i-was-not-expecting/
+test_contains: The problems around vetting AI results won't be going away anytime soon.

lemonde.fr.txt

@@ -43,6 +43,7 @@ strip: //a[contains(@class, 'Header__offer')]
 
 # Remove gift modal
 strip_id_or_class: article__gift-modal
+strip_id_or_class: reuse__content
 
 # Remove the insane "conjugaison.lemonde.fr" links:
 find_string: <a target='_blank' onclick='return false;' class='lien_interne conjug'
@@ -77,3 +78,5 @@ test_url: http://www.lemonde.fr/big-browser/article/2017/10/27/assassinat-de-ken
 test_url: https://www.lemonde.fr/pixels/article/2018/07/14/douze-jeux-video-pour-s-amuser-a-plusieurs_5331269_4408996.html
 test_url: https://www.lemonde.fr/sante/video/2016/04/07/diabete-pourquoi-une-telle-progression-de-l-epidemie_4898147_1651302.html
 test_url: https://www.lemonde.fr/pixels/article/2023/12/20/manettes-de-playstation-sony-condamne-a-13-5-millions-d-euros-d-amende-pour-des-pratiques-anticoncurrentielles_6206941_4408996.html
+test_url: https://www.lemonde.fr/planete/article/2024/02/29/contre-le-frelon-asiatique-les-apiculteurs-appellent-a-la-mobilisation-generale_6219215_3244.html
+

medscape.com.txt

@@ -0,0 +1,10 @@
+body: //div[contains(concat(' ',normalize-space(@class),' '),' article__main-content ')]
+title: //h1[@class='article__title']
+author: //p[@class='meta__author-name']
+date: //p[@class='meta__date']
+
+prune: false
+tidy: true
+
+test_url: https://www.medscape.com/viewarticle/new-evidence-suggests-long-covid-could-be-brain-injury-2024a10002v0?form=fpf
+test_contains: Brain fog is one of the most common, persistent complaints in patients with long COVID.

msn.com.txt

@@ -1,7 +1,10 @@
-http_header(user-agent): curl/7.54.0
+http_header(user-agent): Mozilla/5.0 (compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm)
 
-body: //div[contains(concat(' ',normalize-space(@class),' '),' articlecontent ')]/div[contains(concat(' ',normalize-space(@class),' '),' richtext ')]
+body: //*[@class='article-body']
 
-strip: //div[contains(concat(' ',normalize-space(@class),' '),' articlecontent ')]/div[contains(concat(' ',normalize-space(@class),' '),' richtext ')]/div[not(contains(concat(' ',normalize-space(@class),' '),' e-image '))]
+strip_id_or_class: continue-reading-slot
 
-test_url: https://www.msn.com/en-us/news/technology/followers-are-a-valuable-currency-who-should-own-them/ar-AAXfrUU
+prune: no
+tidy: no
+
+test_url: https://www.msn.com/en-us/news/us/mass-casualty-incident-at-us-mexico-border/ar-BB1jice0

publications.aap.org.txt

@@ -0,0 +1,10 @@
+titl: //h1
+body: //div[contains(concat(' ', normalize-space(@class), ' '), ' article-body ')]
+author: //div[contains(concat(' ', normalize-space(@class), ' '), ' al-authors-list ')]
+date: //span[contains(concat(' ', normalize-space(@class), ' '), ' article-date ')]
+
+prune: false
+tidy: false
+
+test_url: https://publications.aap.org/pediatrics/article/doi/10.1542/peds.2023-062570/196606/Postacute-Sequelae-of-SARS-CoV-2-in-Children?autologincheck=redirected
+test_contains: The coronavirus disease 2019 (COVID-19) pandemic has caused unprecedented devastating medical, social, and economic impacts

ritimo.org.txt

@@ -0,0 +1,12 @@
+strip: //aside
+strip_id_or_class: liste-articles
+author: //span[@class='authors']
+title: //h1
+body: //div[@class='main']
+date: //p[@class='publication']/time
+
+tidy: false
+prune: false
+
+test_url: https://www.ritimo.org/Guide-de-survie-des-aventures-sur-internet-10313
+test_contains: Une surveillance abusive est à craindre, qu’elle soit ciblée ou massive

trendmicro.com.txt

@@ -0,0 +1,16 @@
+body: //p[@class='article-details__description'] | //article/main
+author: substring-after( //p[@class='article-details__author-by']/text() , 'By: ')
+
+strip_id_or_class: article-details__bar
+strip_id_or_class: article-details__display-tag
+strip_id_or_class: article-details__title
+strip_id_or_class: article-details__author-by
+strip_id_or_class: article-details__icons
+strip_id_or_class: research-layout-divider
+
+strip: //sidebar
+
+prune: no
+
+test_url: https://www.trendmicro.com/en_us/research/24/c/unveiling-earth-kapre-aka-redcurls-cyberespionage-tactics-with-t.html
+test_url: https://www.trendmicro.com/en_us/research/17/c/winnti-abuses-github.html