[Fribidi-discuss] CapRTL bug

[Behdad Esfahbod]

On Sun, 5 Oct 2003, Eran Tromer wrote:

> Hi,
>
> The CapRTL test charset produces different results than ISO8859-8 for
> the case "A-5". Is this a bug?
>
> $ echo 'A-5' | fribidi --test --charset=3DCapRTL -
> A-5    =3D>    5-A

Perhaps not.  As doing 'fribidi --charsetdesc CapRTL' shows that
'-' is an ON, but in almost all *real* charsets, '-' is ES I
guess.  For the reason, I have borrowed CapRTL from the Unicode
reference implementation, to test conformance, so their fault if
any.  In future perhaps we're gonna drop that CapRTL and invent a
better one (to add to the mess ;) ).

behdad

> $ echo -e '\340-5' | ./fribidi --test  --charset=3DISO8859-8 -
> =E0-5    =3D>    -5=E0
>
>
> Regards,
>    Eran