[Fontconfig] Monospace detection

[Akira TAGOH]

Thanks for checking. I'm not planning to replace with it completely.
so that sould be better than nothing. but if there are any other, and
more effective way, that would be better than that.

On Sun, Jun 2, 2019 at 5:25 AM <jerrycasiano at gmail.com> wrote:
>
> On my system out of 7099 fonts only 975 contain Panose information.
>
> I don't think processing this info would have much of an impact anyway
> but just something to keep in mind.
>
>
> On Thu, 2019-05-30 at 18:25 +0900, Akira TAGOH wrote:
> > Hi,
> >
> > We have two values in spacing property to consider if a font is
> > monospaced or not. FC_MONO and FC_DUAL.  Right now fontconfig decides
> > it when:
> >
> > - single advance value available in a font -> FC_MONO
> > - two advance values available in a font and larger value is twice of
> > smaller value -> FC_DUAL
> >
> > This code worked for fonts which has smaller language coverage and
> > satisfy older Unicode specifications that non-Latin characters
> > assumes
> > double-widths or so.
> >
> > Nowadays, various glyphs are included in even monospaced fonts which
> > doesn't meet the above criteria (i.e. Noto Mono). it may be a time to
> > update and improve this.
> >
> > What do you think?
> >
> > I'm not sure what approach is better though, the idea that comes to
> > my
> > mind is to refer sFamilyClass and Panose values in OS2 table. if
> > there
> > are any better idea or opposed to this, please let me know.
> >
> > BTW given that we decide it like that, it isn't DUAL anymore. need a
> > good name for that too perhaps, but anyway.
> >
> > Regards,
>


-- 
Akira TAGOH