From 6ed414c1ca7d243d2ae8102938f7c539a973af9f Mon Sep 17 00:00:00 2001 From: Kovid Goyal Date: Fri, 15 Sep 2023 16:13:24 +0530 Subject: [PATCH] Update WSJ --- recipes/wsj.recipe | 8 +++++--- recipes/wsj_free.recipe | 8 +++++--- 2 files changed, 10 insertions(+), 6 deletions(-) diff --git a/recipes/wsj.recipe b/recipes/wsj.recipe index 22479e213b..f9de630b5c 100644 --- a/recipes/wsj.recipe +++ b/recipes/wsj.recipe @@ -66,13 +66,15 @@ class WSJ(BasicNewsRecipe): ''' keep_only_tags = [ - classes('wsj-article-headline-wrap articleLead bylineWrap bigTop-hero'), + classes('wsj-article-headline-wrap articleLead bylineWrap bigTop-hero article-container'), dict(name='section', attrs={'subscriptions-section':'content'}) ] remove_tags = [ classes('wsj-ad newsletter-inset media-object-video media-object-podcast podcast--iframe dynamic-inset-overflow-button'), - dict(name='amp-iframe') # interactive graphics + dict(role=["toolbar", "complementary"]), + dict(attrs={"aria-label": ["Sponsored Offers", "What to Read Next"]}), + dict(name='amp-iframe'), # interactive graphics ] def preprocess_html(self, soup): @@ -91,7 +93,7 @@ class WSJ(BasicNewsRecipe): def get_cover_url(self): from datetime import date cover = 'https://img.kiosko.net/' + date.today().strftime('%Y/%m/%d') + '/us/wsj.750.jpg' - br = BasicNewsRecipe.get_browser(self, verify_ssl_certificates=False) + br = BasicNewsRecipe.get_browser(self, verify_ssl_certificates=False) try: br.open(cover) except: diff --git a/recipes/wsj_free.recipe b/recipes/wsj_free.recipe index cea7a7a281..0653cb8534 100644 --- a/recipes/wsj_free.recipe +++ b/recipes/wsj_free.recipe @@ -66,13 +66,15 @@ class WSJ(BasicNewsRecipe): ''' keep_only_tags = [ - classes('wsj-article-headline-wrap articleLead bylineWrap bigTop-hero'), + classes('wsj-article-headline-wrap articleLead bylineWrap bigTop-hero article-container'), dict(name='section', attrs={'subscriptions-section':'content'}) ] remove_tags = [ classes('wsj-ad newsletter-inset media-object-video media-object-podcast podcast--iframe dynamic-inset-overflow-button'), - dict(name='amp-iframe') # interactive graphics + dict(role=["toolbar", "complementary"]), + dict(attrs={"aria-label": ["Sponsored Offers", "What to Read Next"]}), + dict(name='amp-iframe'), # interactive graphics ] def preprocess_html(self, soup): @@ -91,7 +93,7 @@ class WSJ(BasicNewsRecipe): def get_cover_url(self): from datetime import date cover = 'https://img.kiosko.net/' + date.today().strftime('%Y/%m/%d') + '/us/wsj.750.jpg' - br = BasicNewsRecipe.get_browser(self, verify_ssl_certificates=False) + br = BasicNewsRecipe.get_browser(self, verify_ssl_certificates=False) try: br.open(cover) except: