> I have been thinking about the issues involved though, and have a list of
> non-trivial tasks in my head that are blocking the filter from being
> feature-complete:
>
> 1 - image support (dependant on hooking up an importer for the WPG format--
> imagemagick has one, but doesn't seem to import bitmaps correctly. [ this
> might be easy to fix ] dia apparently also has a wpg export filter, which
> might form a good basis for an input filter, but I haven't really played with
> it)

I'm still studying this WPG format, although I can't devote much time (for
the same reason as you, lack of time). Based on my last examination of
ImageMagick and Dia, none has full and complete support for WPG (it's
rather a quick-and-dirty hack).  Therefore, I suggest that we might want
to start a library which does for WPG what libwmf does for WMF. Any
volunteer ?

> 2 - table support [ dependant on abiword supporting this feature, but
> shouldn't be terribly difficult once it does ]

Basic WP table parsing is easy. However, I still don't come to fully
understand the nitty gritty details (e.g border, inline frame placement,
...)

> 3 - footnote/endnote support  [ dependant on abiword supporting this feature,
> but shouldn't be terribly difficult once it does ]

Shouldn't be too difficult as well.

> 4 - font change support [ seemingly dependant on getting a list of fonts from
> abiword, and a font substitution table. i imagine this will be coming post
> 1.0, but this hasn't been much discussed on this last. ]

CMIIW but last time I think we agreed that font substitution is better
done on the Abiword/KWord side, not the filter itself (any comments,
Dom?).

> 5 - list support [ i pretty much managed to reverse engineer the abiword list
> format, but haven't found time to implement the damn thing. as i'll probably
> be rather busy for the next month at least, if anyone is feeling keen in the
> interim, they may go ahead and implement this without fear of duplicating
> effort. i have attached my findings from a while ago to the end of this
> message ]

Great work ! I never succedded in understanding this...

> Is the WP 5.x format a strict subset of the WP 6+ format? Or were there been
> more substantial changes? It looks from your code as if there are only minor
> differences, so much so that one might be able to subclass off a WP 5.x
> filter from a WP 6+ one (this would be the abi way of doing things).

It's not a subset, there are some obvious differences in the way WP 5 and
6 implement things. I suggest that you look at WP 5.x file format
documentation (available from wotsit.org) which is unfortunately is in WP
5 file. I hope I do have time to convert it to text/HTML version of this
(using KWord of course :-)

OTOH a few weeks ago I managed to create experimental export filter for WP
(not too difficult, it's just tricky).  I have to dig this stuff again (no
promise though). I will keep you inform about this.

> Good to see you on the list, in any event-- it seems to me as if more
> extensive collaboration between KWord and AbiWord can only be a good thing,
> especially on issues such as this.

I agree 100%. And I hope I am just welcomed, even if I don't code for
Abiword (hey, at least I grabbed the CVS!).

Regards,

Ariya

Reply via email to