>>>  Rectangular selections pose a real problem.
>> Bah! It never worked well anyway...
> Can I kicck it out? Can I, can I? Pleeaaase?  :-P

Well rectangular selection only makes sense when dealing with a
fixed width font where all characters occupy one column, and we
have sparkaround's input in STR-2158 that that's not always true.

> I'll rip the rectangle selection out first which will greatly
> reduce the number of lines of code. I'll refactor the re-occuring
> code which should get us 50% there. Then the only real issue left
> would be efficient rendering of a line of text which must be
> rewritten.

There are 3 outstanding problems raised by STR-2158.

1. Incorrect counting of bytes instead of characters for UTF-8 glyphs.
   That's a one line fix.

2. Incorrect assumption that all [non-control] characters occupy just
   one column. sparkaround's CJK characters occupy 2 characters.
   This is fixable by incorporating Markus Kuhn's wcwidth.c code
   and changing some Fl_Text_* routines to use a char* or ucs parameter
   instead of just one char at a time. I was working out the best
   approach and consolidating this in light of...

3. wrap_mode(1, 0) should result in wrapping text at the widget width
   based on the pixel width of characters, but in fact this only works
   for ascii, and not higher UTF-8 characters. I was investigating...

D.
_______________________________________________
fltk-dev mailing list
[email protected]
http://lists.easysw.com/mailman/listinfo/fltk-dev

Reply via email to