On Tue, Mar 05, 2002 at 11:56:10AM +0100, Juergen Vigna wrote: > On 04-Mar-2002 Martin Vermeer wrote: > > > Good luck (I mean it.) In the meantime, find attached the simple version > > of the patch, which fixes the bug and does nothing more. > > I don't think there is luck involved. Anyway have a look at the fix I > just comitted if you're interested. > > Jug
Yes, this works for me. It does approximately what my first patch also did
and looks similarly ugly :-)
But I like the trick of making *par the previous of the first paragraph!
And you prevent "double deepenings" from appearing, using the useful
routine getMaxDepthAfter. I hadn't thought that far yet.
One thing I don't quite like, or find elegant, in your patch is the way
you handle the case where a depth-delta'd paragraph in the buffer
acquires a negative depth. Apparently you just switch off depth-deltaing
at that point!
Obviously you are going to lose something if you try to "graft" a
nested-lists cut with a shallower end point than the starting point, on a
too shallow target point, but...
Wouldn't it be better this way:
// set the right depth so that we are not too deep or shallow.
if ((static_cast<int>(tmpbuf->params().depth()) + depth_delta) < 0)
tmpbuf->params().depth(tmpbuf->params().depth());
else
tmpbuf->params().depth(tmpbuf->params().depth() + depth_delta);
or even (switching my optimizer on):
// set the right depth so that we are not too deep or shallow.
if ((static_cast<int>(tmpbuf->params().depth()) + depth_delta) >= 0)
tmpbuf->params().depth(tmpbuf->params().depth() + depth_delta);
This way depth_delta is remembered for if the outline goes deeper again
later on. Only the too shallow paragraphs are flattened.
Martin
msg33953/pgp00000.pgp
Description: PGP signature
