Improve remove headings content

This commit is contained in:
Kiryl
2022-06-22 18:18:58 +03:00
parent bc4055bdaf
commit de1246d890

View File

@@ -190,14 +190,32 @@ def _remove_headings_content(content_tag, title_of_chapter: str):
title_of_chapter = title_of_chapter.lower()
for tag in content_tag.contents:
text = tag if isinstance(tag, NavigableString) else tag.text
if text:
if re.sub(r'([\s\xa0])', '', text):
text = re.sub(r"[\s\xa0]", " ", text).lower()
text = text.strip() # delete extra spaces
if title_of_chapter == text or \
(title_of_chapter in text and re.findall(r"^h[1-3]$", tag.name)):
_add_span_to_save_ids_for_links(tag, content_tag)
tag.extract()
break
elif not isinstance(tag, NavigableString):
_remove_headings_content(tag, title_of_chapter)
break
def _tags_to_correspond_livecarta_tag(chapter_tag):
"""Function to replace all tags to correspond livecarta tags"""
for reg_key, to_replace_value in LiveCartaConfig.REPLACE_REGEX_WITH_LIVECARTA_CORRESPOND_TAGS.items():
for key in reg_key:
tags = chapter_tag.find_all(re.compile(key))
for tag in tags:
tag.name = to_replace_value
def _unwrap_tags(chapter_tag):
"""Function unwrap tags and move id to span"""
for tag in LiveCartaConfig. TAGS_TO_UNWRAP:
for s in chapter_tag.find_all(tag):
_add_span_to_save_ids_for_links(s, chapter_tag)
s.unwrap()
# todo remove