Gmane
Favicon
From: David Goodger <goodger <at> python.org>
Subject: Re: Re: The Trouble with Tables
Newsgroups: gmane.text.docutils.user
Date: 2005-04-05 22:44:37 GMT (4 years, 12 weeks, 6 days, 23 hours and 16 minutes ago)
[Felix Wiemann]
 > What about such a syntax::
 >
 >     .. table:: XOR
 >        :stub-columns: 1
 >
 >        +---+---+---+
 >        |   | T | F |
 >        +===+===+===+
 >        | T | F | T |
 >        +---+---+---+
 >        | F | T | F |
 >        +---+---+---+
 >
 > Or even::
 >
 >     .. table:: XOR
 >        :header-rows: 1
 >        :stub-columns: 1
 >
 >        +---+---+---+
 >        |   | T | F |
 >        +---+---+---+
 >        | T | F | T |
 >        +---+---+---+
 >        | F | T | F |
 >        +---+---+---+

Sure, +1.

 > I think the different table directives ("table", "list-table",
 > "csv-table") basically call for unification into a single "table"
 > directive.

+1 on unifying the *implementations* (and documentation).  See the
implementation of admonitions for an example.

-1 on unifying the directives themselves.  I'd much rather write

     .. list-table::

        ...

than

     .. table::
        :format: list

        ...

Think polymorphism!

 > Furthermore, it might be useful to have a "rotate" option (not sure
 > about the best name yet; maybe "column-wise") which causes the table
 > to be rotated, i.e. CSV tables and list tables would appear
 > column-wise instead of row-wise.

+1 on "column-wise".  That's an already-recorded idea though; see
http://docutils.sf.net/docs/dev/rst/alternatives.html#list-driven-tables
for this one and others.  Feel free to update it.

--
David Goodger <http://python.net/~goodger>