From: Dom Lachowicz (domlachowicz@yahoo.com)
Date: Mon Jun 23 2003 - 15:29:07 EDT
> All of this is a little much for a single POW, but I
> think Martin's
> emphasis is on round-trip XHTML table import/export.
Yes, indeed. Personally, I'm tempted to make your
libxml2-based HTML plugin mandatory and be done with
it.
> Question: Are custom clip-board types supported?
> XHTML is basically a
> document structure and the copy&paste of a single
> table is encumbered
> by the document superstructure, i.e.:
I done't understand what you mean by "custom". We can
support any type of data on the clipboard in theory.
In practice, the order in which you register things on
the clipboard has a profound effect on the types of
things you get back. This is why we don't support
dynamic clipboard registry for any importer/exporter.
Eg. if this clipboard ordering happens (which is
likely since HTML is strcmp ordered befor RTF):
text/html
application/rtf
and we cut+paste to ourselves, we'll get HTML
round-tripped instead of RTF, with greatly diminished
quality of result. There needs to be a *lot* more
thought on how to do this intelligently. Dataloss and
formatting loss is unacceptable.
> and I'm often tempted just to prune the peripheral
> elements, i.e.:
No please don't do this. We'll want to export the
document in its entirety to the clipboard, including
any <html> etc.. elements that you'd find
"unnecessary". The entity on the clipboard, regardless
of its format (rtf, html, ...), should be a valid,
standalone document.
> After all, what does gnumeric do when faced when a
> whole document? Or
> does it assume that the document contains a single
> table?
This needs to be codified moreso.
Dom
__________________________________
Do you Yahoo!?
SBC Yahoo! DSL - Now only $29.95 per month!
http://sbc.yahoo.com
This archive was generated by hypermail 2.1.4 : Mon Jun 23 2003 - 15:40:09 EDT