Question

I am using regex in Python to pull comments, formatted with Markdown, from a CoffeeScript file. CS uses ### as the start and end markers for multiline comments. This conflicts with the markdown formatting because # is used to denote a header class. Is it be possible make Markdown interpret the % or ^ or some other character as the header class token?

Answer 1

It is possible to override many elements of Python Markdown by taking advantage of the markdown.extensions.Extension class. After a little digging and experimenting I got this to work by writing a custom header parser which extended the generic BlockParser.

Essentially it looks something like this

class CustomHeaderParser( BlockParser )
    def run( params )
        # implementation

class CustomExtension( Extension )
    def extendMarkdown( self, md, md_globals )
        md.parser.blockprocessor.add( 'custom_header_parser', CustomHeaderParser(md.parser), '<hashheader' )
        md.registerExtension( self )


markdown_parser = markdown.Markdown( extensions=[ CaretHeaderExtension() ] )
markdown_parser.convert( markdown_text )

About half way through researching this I also found another, less desirable, work around. There are two types of header styles that markdown understands Setext and atx. Setext uses ==== and —– on the line below the text to provide two header levels h1 and h2.