[Doc-SIG] More on docutils

[Tony J Ibbs (Tibs)]

Well, there are also some serious (stupid) problems with the
amalgamating of literal paragraphs (i.e., the subparagraphs of a
paragraph ending in '::') in the "old form" docutils code.

As it turned out, I spent some time over the weekend working on the code
rewrite - it's just too frustrating to try to work on the old code when
the new code is already cooking in my head. Not to mention embarassing.

*Unfortunately*, the floppy I brought in to work this morning only had
Friday evening's (well, Saturday morning's) version of DocNodes.py,
which is significantly out-of-date with respect to the end of Sunday.
I'll try to bring in a current version tomorrow and upload it (my modem
connection at home is, well, flaky).

As a taster, DocNodes.py already has the following:

* paragraph splitting, including at list items
* support for '::' literal paragraphs (which *appears to work*)
* support for '>>>' Python paragraphs
* much simpler object model, much simpler internal tree structure,
generally less faffing around
* much better docstrings
* docstrings include doctest "code", and the module is being written to
use doctest for self-testing (and yes, this is both fun (hmm, running
code that parses the text that includes the result of the parse - nice
stuff) and makes it easier to test)

It should rapidly (erm, remember this is Glacial Software Development
Unlimited, when interpreting that word) gain:

* support for list items

shortly to be followed by

* paragraph markup
* output to DOM tree

and so on, bringing it back to where the "old" code was, but (heh heh)
working this time...

Tibs

--
Tony J Ibbs (Tibs)      http://www.tibsnjoan.co.uk/
Feet first on 3 wheels...
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)