From 581d6f665760decc2cfa77e9024ba18dd4d36bfb Mon Sep 17 00:00:00 2001 From: Jim Miller Date: Mon, 5 Feb 2024 10:36:41 -0600 Subject: [PATCH] adapter_literotica: Allow for empty div.aa_ht tags, remove extra None from text --- fanficfare/adapters/adapter_literotica.py | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/fanficfare/adapters/adapter_literotica.py b/fanficfare/adapters/adapter_literotica.py index 48891e33..a6d93fd0 100644 --- a/fanficfare/adapters/adapter_literotica.py +++ b/fanficfare/adapters/adapter_literotica.py @@ -341,13 +341,13 @@ class LiteroticaSiteAdapter(BaseSiteAdapter): [comment.extract() for comment in page_soup.findAll(string=lambda text:isinstance(text, Comment))] fullhtml = "" for aa_ht_div in page_soup.find_all('div', 'aa_ht'): - html = unicode(aa_ht_div.div) - # logger.debug(html) - # Strip some starting and ending tags, - html = re.sub(r'^', r'', html) - html = re.sub(r'$', r'', html) - html = re.sub(r'

$', r'', html) - fullhtml = fullhtml + html + if aa_ht_div.div: + html = unicode(aa_ht_div.div) + # Strip some starting and ending tags, + html = re.sub(r'^', r'', html) + html = re.sub(r'$', r'', html) + html = re.sub(r'

$', r'', html) + fullhtml = fullhtml + html # logger.debug('getPageText - fullhtml: %s' % fullhtml) return fullhtml