Re: Hyphenation Design (Was Re: Possible Pango 1.4 ideas)

[Owen Taylor <otaylor redhat com>]

Matthias Clasen <maclas gmx de> writes:

> On Thu, 2003-01-23 at 01:41, Owen Taylor wrote:
> > 
> > Matthias Clasen <maclas gmx de> writes:
> > 
> > > One question that just occurred to me is if layout-generated characters
> > > in the PangoLayout (or, if you want, PangoSuperDuperLayout) like hypens
> > > or ZWNBSPs inserted for kerning won't interfere with uses of PangoLayout
> > > for visual navigation like gtk_text_layout_move_iter_visually() does ?
> > 
> > The set of possible cursor positions is defined by the results of
> > pango_break() which would be called on the pre-hyphenation text, so 
> > the hyphen glyph should't effect cursor positioning at all. (*)
> > 
> 
> I guess that also applies to justification-inserted glyphs like Arabic
> Kashidas, but what about your proposal to represent manual kerning by 
> ZWNBSPs ? Even if the user explicitly specified the kerning, he would
> probably still be surprised to stumble over the extra invisible
> character when navigating with the cursor. 

Keynav around special characters is still something that we
don't handle very well. There was some extensive discussion of
this on this list earlier, but nothing very conclusive for
a result. My feeling though, is generally, that if a a character
doesn't have a visual representation, it shouldn't have a 
corresponding cursor position.

> > (*) Except for the problems with positioning at the ends of lines
> >     without trailing whitespace that we have now in GtkTextView.
> >     There should a bug open about it somewhere. 
> 
> That would be http://bugzilla.gnome.org/show_bug.cgi?id=75270 and maybe
> http://bugzilla.gnome.org/show_bug.cgi?id=81412

81412 is the one I was referring to.

75270 is just an off-by-one error or something.

Regards,
                                        Owen