Re: [gtkmm] Reusing gtk+ reference documentation



Murray Cumming wrote:

On Sat, 2001-12-01 at 23:58, Murray Cumming wrote:

My current plan was to copy all the gtk+ inline documentation to gtkmm2
when they seemed to have finished writing it themselves.

However, maybe this would work:

- A script could scan all the gtk+ code, outputting the text in to a
file that's a bit like gtk.defs. For instance, gtk.refdoc gtk-doc must
already do similar scanning to build the html.

- gtkmmproc would look at this gtk.refdoc file as well as gtk.defs when
it builds the .h/.cc files. It would put the gtk+ comments next to our
gtkmm methods, for Doxygen to parse. This would make gtkmmproc even
slower, of course.

- The gtk+ documentation would not always be appropriate, so we could
have a second file containing a few custom overrides. gtkmmproc would
use the custom text if it exists for the method.

This would allow us to reuse the code without a lot of tedious
copy-and-paste, and it would allow us to update our documentation when
gtk+ updates theirs.


Would any perl coders like to take this task on? I think that the best way might be to add a --output-docs-defs option to the gtkdoc-mkdb script in the gtk-doc
module. We could then add a suitable make target where this is used in
gtk/docs/reference.



You might want to look at the docgen.py script in the codegen/ subdirectory of pygtk. It uses a defs file to get information about how the C API maps to methods of objects, then pulls the documentation out of the source code and outputs fake python documentation :)

The script is a little out of date now because of the big class renaming I did in pygtk, but the script might be useful to you.

You will still need some form of language dependent overrides, as some of the inline docs are C specific (one or two include C example code).

James.

--
Email: james daa com au
WWW:   http://www.daa.com.au/~james/






[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]