TODO.txt

- py3: py2.4 test (broken?)
- add "markdown-in-html" extra to wiki
- add "smarty-pants" extra to wiki
- add "html-classes" extra to wiki
- more on the "code-color" extra wiki page
- better "Extras" wiki page that shows a quick example of each with link for
  more details. Mention "extensions" for gjuice.
- add description of demote-headers extras to wiki
- find more unicode edge cases (look for any usage of md5() and make that
  unicode)
- bug: can't have '<\w+' in a code span or code block with safe_mode if there
  is a '>' somewhere later in the document. E.g. code.as.com-beta/CHANGES.md.
  It captures all of that. Right answer is to not count code spans or code
  blocks.
  - add an issue for this
  - test cases
  - idea: better sanitation re-write? lot of work
  - idea: Change all <,>,& emission from markdown processing to something
    like {LT}, {GT}, {AMP}, {OPENTAG:$tag[:$class]} (first checking for
    conflicts and escaping those out of the way). Then do sanitization at the
    end:
        escape: escape all <,>,& with entities
        remove: not supported
        whitelist: (new) build a reasonable default whitelist of patterns to
            keep. Takes "extras" argument (and hook for subclassing) to
            for custom whitelist. Google Code (was it?) had some list
            of reasonable whitelist stuff.
    Then unescape these special chars. The use of OPENTAG above would make
    "html-classes" extra trivial.
- test safe_mode on HTML in footnotes
- http://www.freewisdom.org/projects/python-markdown/Available_Extensions
- Extras.wiki desc of code-color option. Not sure I love the ":::name"
  markup for the lexer name.
- update MDTest 1.1? (see
  http://six.pairlist.net/pipermail/markdown-discuss/2007-September/000815.html)
  update MDTest tests from http://git.michelf.com/mdtest/
- I see ref to Markdown.pl 1.0.2
  (http://six.pairlist.net/pipermail/markdown-discuss/2007-August/000756.html)
  Update to that? Yes. Copy, at least, in showdown package.
- take a look at other examples/test-cases from
  http://adlcommunity.net/help.php?file=advanced_markdown.html
- get on http://en.wikipedia.org/wiki/Markdown
- ask about remaining two MarkdownTest test failures
- fix the r135 xml option, add xml extra for it (see email)
- perhaps some extras from Maruku and PHP Markdown extra
  (http://maruku.rubyforge.org/maruku.html#extra)
    - tables (tho I don't really like the syntax, prefer google codes, see
      below)
    - markdown inside literal HTML (if 'markdown="1|true"' attr)
    - automatic toc generation (wanted that anyway, no a fan of maruku syntax
      for this)
    - weird markup in headers and links (does markdown2.py handle this?)
    - meta-data syntax? One example of this is ids for headers. How about
      automatically assigning header ids from the name (a la rest)?
    - at-the-top email-style headers?
    - maruku's footnote links are 'fn:1' and 'fnref:1' for a footnote id of
      'blah'. If this is the PHP Markdown Extras way, then should follow
      that.
- googlecode wiki markup ideas?
  (http://code.google.com/p/support/wiki/WikiSyntax)
    - ~~strikeout~~
- <http://daringfireball.net/2004/12/markdown_licensing> at bottom has a wish
  list:
    - simple "cite" for blockquote. How about:
        [Zaphod Breeblebrox]
        > blah blah
        > blah
- look at http://markedapp.com/   screencast?