[Python-markdown-discuss] Ticket 74

[Gerry L. <gjl...@gm...>]

Started looking into this one and I think I uncovered a somewhat related
bug.  If a child element's tail is assigned, then neither the text nor
tail show up in the output.

For example:

    import markdown
    from markdown.inlinepatterns import Pattern

    class MyPattern(Pattern):

        def handleMatch(self, m):

            el = markdown.etree.Element('p')
            el.text = markdown.AtomicString('an *atomic*')
            c1 = markdown.etree.SubElement(el, 'span')
            c1.text = markdown.AtomicString('*string*')
            c1.tail = markdown.AtomicString('`not code`')
            return el

    class MyExtension(markdown.Extension):

        def extendMarkdown(self, md, md_globals):
            patt = MyPattern(r'::(.*?)::')
            md.inlinePatterns.insert(0, 'foobar', patt)



    ext = MyExtension(None)
    md = markdown.Markdown(extensions=[ext], output_format="xhtml1")
    html = md.convert('here is ::mypattern::')

    print html

The resulting output is:
    <p>here is <p>an *atomic* </p>
    </p>

As opposed to:
    <p>here is <p>an *atomic*<span>*string*</span>`not code`</p>
    </p>

This is probably out in the weeds as far as usage goes.  Also, I'm
surprised that this doesn't cause normal usage issues- surely the tree
parser receives similar input in normal use situations(ie when parsing a
file of Markdown text).  An element with a child with text and tail
can't be that unusual.

Anyway, I'll be investigating it a bit more.

Regards-
Gerry LaMontagne