Get python-markdown extension to output html tags

问题:

I initially thought that how to pass class attribute and value to markdown syntax and How do I set an HTML class attribute in Markdown? asked the question that I want to ask here, but I want to know how to write an python-markdown extension that outputs an html span element with specific class attribute in Django.

I've written the following extension (file saved as "mdx_debatesyntax.py"):

import markdown
from markdown.inlinepatterns import Pattern

PREMISE_RE = r'(\[p )(.*?)\]'

class AttrTagPattern(Pattern):
    """
    Return element of type `tag` with a text attribute of group(3) 
    of a Pattern and with the html attributes defined with the constructor.

    """
    def __init__ (self, pattern, tag, attrs):
        Pattern.__init__(self, pattern)
        self.tag = tag
        self.attrs = attrs

    def handleMatch(self, m):
        el = markdown.util.etree.Element(self.tag)
        el.text = m.group(3)
        for (key,val) in self.attrs.items():
            el.set(key,val)
        return el

class DebateSyntaxExtension(markdown.Extension):
    def extendMarkdown(self, md, md_globals):
        premise_tag = AttrTagPattern(PREMISE_RE, 'span',{'class':'premise'})
        md.inlinePatterns.add('premise', premise_tag, '_begin')

def makeExtension(configs=None):
    return DebateSyntaxExtension(configs=configs)

And when I test this directly (not through Django) it works:

import markdown
md = markdown.Markdown(extensions=['debatesyntax'])
md.convert('This is a [p test]!')

Returns

u'<p>This is a <span class="premise">test</span>!</p>'

However, when I run this through Django, the attribute class gets stripped, giving me:

<p>This is a <span>test</span>!</p>

How do I prevent the class attribute (or other attributes) from being stripped?

Resolution

The Django app that I was modifying (OSQA), was manually "sanitizing" the HTML. The AttrTagPattern class code does, after all, work.

回答1:

try

{{ value|markdown:"debatesyntax" }}