def parse_mkdocs(content, section, url): # pylint: disable=unused-argument
"""Get the embed content for the section."""
ret = []
headers = []
if not content or not content.get('content'):
return (None, None, section)
body = content['content']
for element in PQ(body)('h2'):
headers.append(recurse_while_none(element))
if not section and headers:
# If no section is sent, return the content of the first one
section = list(headers[0].keys())[0].lower()
if section:
body_obj = PQ(body)
escaped_section = escape_selector(section)
section_list = body_obj(
':header:contains("{title}")'.format(title=str(escaped_section)))
for num in range(len(section_list)):
header2 = section_list.eq(num)
# h2_title = h2.text().strip()
# section_id = h2.attr('id')
h2_content = ""
next_p = header2.next()
while next_p:
if next_p[0].tag == 'h2':
break
h2_html = next_p.outerHtml()
if h2_html:
h2_content += "\n%s\n" % h2_html
next_p = next_p.next()
if h2_content:
ret.append(h2_content)
# ret.append({
# 'id': section_id,
# 'title': h2_title,
# 'content': h2_content,
# })
return (ret, headers, section)
def parse_sphinx(content, section, url):
"""Get the embed content for the section."""
body = content.get('body')
toc = content.get('toc')
if not content or not body or not toc:
return (None, None, section)
headers = [recurse_while_none(element) for element in PQ(toc)('a')]
if not section and headers:
# If no section is sent, return the content of the first one
# TODO: This will always be the full page content,
# lets do something smarter here
section = list(headers[0].keys())[0].lower()
if not section:
return [], headers, None
body_obj = PQ(body)
escaped_section = escape_selector(section)
elements_id = [
escaped_section,
slugify(escaped_section),
make_id(escaped_section),
f'module-{escaped_section}',
]
query_result = []
for element_id in elements_id:
if not element_id:
continue
query_result = body_obj(f'#{element_id}')
if query_result:
break
if not query_result:
selector = f':header:contains("{escaped_section}")'
query_result = body_obj(selector).parent()
# Handle ``dt`` special cases
if len(query_result) == 1 and query_result[0].tag == 'dt':
parent = query_result.parent()
if 'glossary' in parent.attr('class'):
# Sphinx HTML structure for term glossary puts the ``id`` in the
# ``dt`` element with the title of the term. In this case, we
# need to return the next sibling which contains the definition
# of the term itself.
# Structure:
# <dl class="glossary docutils">
# <dt id="term-definition">definition</dt>
# <dd>Text definition for the term</dd>
# ...
# </dl>
query_result = query_result.next()
elif 'citation' in parent.attr('class'):
# Sphinx HTML structure for sphinxcontrib-bibtex puts the ``id`` in the
# ``dt`` element with the title of the cite. In this case, we
# need to return the next sibling which contains the cite itself.
# Structure:
# <dl class="citation">
# <dt id="cite-id"><span><a>Title of the cite</a></span></dt>
# <dd>Content of the cite</dd>
# ...
# </dl>
query_result = query_result.next()
else:
# Sphinx HTML structure for definition list puts the ``id``
# the ``dt`` element, instead of the ``dl``. This makes
# the backend to return just the title of the definition. If we
# detect this case, we return the parent (the whole ``dl``)
# Structure:
# <dl class="confval">
# <dt id="confval-config">
# <code class="descname">config</code>
# <a class="headerlink" href="#confval-config">¶</a></dt>
# <dd><p>Text with a description</p></dd>
# </dl>
query_result = parent
def dump(obj):
"""Handle API-based doc HTML."""
if obj[0].tag in ['span', 'h2']:
return obj.parent().outerHtml()
return obj.outerHtml()
ret = [dump(clean_links(PQ(obj), url)) for obj in query_result]
return ret, headers, section