[XeTeX] Anomalies in using feature-enabled OT fonts

[Khaled Hosny]

On Sun, Nov 21, 2010 at 12:35:35AM +0900, Takayuki YATO (ZR) wrote:
> Thank you for reeponse.
> 
> > > 2. \fontcharwd / \fontcharht / \fontchardp
> > > \font\tgpNorm="TeX Gyre Pagella"       % default
> > > \font\tgpSmcp="TeX Gyre Pagella:+smcp" % small-caps
> > > % There two lines should not report the same measurement....
> > > \tgpNorm\measure{a}\measure{f}\measure{g}
> > > \tgpSmcp\measure{a}\measure{f}\measure{g}
> > > \bye
> > But `a will always resolve too 97, so in all cases you are asking for
> > the metrics of the glyph at 97 slot of the font; the meaning of `a is
> > font independent (will, I don't know the exact terms, some TeX guru
> > would explain better).
> 
> I understand your explanation and admit it's logical. But from the
> grammar of XeTeX I have the impression that XeTeX treats a font with
> some feature enabled as different from the original font; that is,
> there are two distinct fonts \tgpNorm and \tgpSmcp, and the latter has
> small-caps glyphs for lowercase characters. In this view, the glyph at
> slot 97 is different between these "two" fonts. And it is quite natural
> that one expect \fontcharwd, \XeTeXcharglyph, etc. applied to \tgpSmcp
> return the information on the small-caps glyphs.

No, the glyph at slot 97 is still normal lower case a glyph (unless you
are using an old smallcaps font, but that is another story), when you
select a +smcp or any other feature, the engine will then use the
smallcap glyph from its slot (which is not 97 normally) whenever it
encounters a lower case a character. The `a notation is just another way
to pass numbers to TeX (the same is "XXXX notation) and it does not
change its value when you change fonts; `a is always resolved to 97.

> > IMO the only reliable way to check for a specific un-encoded glyph is
> > either by glyph id or glyph name, but both are very unportable.
> 
> If I have ever to go in your view, then the variant of \XeTeXcharglyph
> which returns the id of the glyoh that results from feature application
> seems to me the more desirable. And it would be better if there were
> in addition \XeTeXglyphwd, \XeTeXglyphht and \XeTeXglyphdp. (Are the
> latter two equivalent to \XeTeXglyphbounds 2 and \XeTeXglyphbounds 4?)

I'm not saying it is right or wrong, I'm just trying to understand what
is going on.

Regards,
 Khaled

-- 
 Khaled Hosny
 Arabic localiser and member of Arabeyes.org team
 Free font developer