On Mon, May 02, 2005 at 07:00:05PM +0200, Georg Baum wrote: > Martin Vermeer wrote: > > > On Mon, May 02, 2005 at 03:25:39PM +0200, Juergen Spitzmueller wrote: > >> I kind of like the idea (untested), but I think you should trigger an > >> Alert which tells the user that some charstyle insets haven't been found > >> etc. > > > > Unfortunately that's not easy to do under this approach... you don't > > know it happened until after you saved and re-loaded. That's precisely > > the weakness of this approach. > > It is easy. Proof attached (replaces your patch). This is a case where the > dociterator stuff pays off. And I found a missing _() as well. > I think this should go in if it works for you, too (I have not tested it > thoroughly).
I think only the CutAndPaste part changed relative to my patch... so if that works for you... the beauty of this patch is that now we have the best of both worlds: a message at conversion time, and a reminder in the converted doc! I'll check it tomorrow if I find time. - Martin BTW another wild idea. In lyxtextclass.[hC] is defined CharStyle and CharStyles, a list of them. Adding a method addCharStyle would be immediately useful in readCharStyle, and then could be used (addCharStyle(tclass1.charstyle(name))) in SwitchBetweenClasses in place of where you put inset.setUndefined(). In that way one could add character styles "on the fly" to the CharStyles list... as a move towards user charstyles. Ah well, not now.
pgpoOyvvIaHhk.pgp
Description: PGP signature
