Dov Feldstern wrote:
Abdelrazak Younes wrote:
PS: Dov, there could be some problems with this last commit and bidi
support. I've tried hard to only change the interfaces and to maintain
the logic but bugs happen. Please let me know if there's anything wrong.
Yeah, I'm sitting here cringing about this commit ;-/ --- I'm afraid
it's going to break a whole bunch of things. For example, there are
places at painting where having the language information is very
important...
I don't think I changed the logic there, as I said, the place where
Language information was needed (mainly in rowpainter) should still work
the same.
ATM it doesn't compile, though, so it's hard to test it ;) .
What's the problem?
I agree very much with the fact that we should have a generalized range,
and that language and font should be separate specializations of it. But
(a) in practice, I think that that separation may be hard to maintain;
and (b) I would be happier if we had a good design for this before going
ahead and doing it. Especially if you say that there are places that you
know are problematic (like the InsetQuotes) --- I think that making the
code cleaner at the expense of regressions is not right...
Well, that's just screen painting here and as I wrote, it's easy enough
to fix though I am not settled on the right solution. In Unicode time, I
don't really understand why we have a new for InsetQuotes actually.
But we already know that you and I have very different approaches to
these issues... ;)
Oh I try very hard to not introduce any regression but the above problem
was not implemented correctly anyway...
Abdel.
