RE: Tamil glyphs

From: Marco.Cimarosti@icl.com
Date: Fri Sep 08 2000 - 06:08:48 EDT


Michael (michka) Kaplan wrote:
> From: "Rick McGowan" <rmcgowan@apple.com>
> [...]
> > I suppose if you just want to display the non-ligature type
> thing in a
> situation where the font wants to give you the ligature type
> thing, you
> might be able to use a ZWNJ or ZWNBSP between the chars.
> [...]
>
> Unfortunately, this will not work, as vowel reordering must
> happen, and will
> not if ZWNJ or ZWNBSP is included. [...]

ZWNBSP is not a good idea here; after all it is just a *space*.

But ZWJ and ZWNJ should be the way to go for this kind of things, especially
since they have been recently promoted to also represent ZWL and ZWNL
(zero-width [non] ligator).

It is very unfortunate that ZWNJ prevents Indic vowel reordering in this
case.

In ordinary cases, a ZW[N]J inside a consonant cluster does not prevent
matra reordering. E.g., in Devanagari:

        U+0915, U+094D, U+200C, U+0915, U+093F (ka, virama, ZWNJ, ka, i
matra)

is regularly reordered around the cluster:

        093F, 0915, 094D, 200D, 0915 (i matra, ka, virama, ZWNJ, ka)

and rendered with this sequence of glyphs:

        i_matra, ka_nominal, virama, ka_nominal

The ZWNJ is there just to mandate the non-default visualization of virama,
forbidding the regular renderings, that would be one of:

        i_matra, ka_half, ka_nominal
or
        i_matra, kka_ligature

depending on the font availability of a kka_ligature.

It would be nice if the possibility of reordering matras around a ZW[N]J
could be generalized, e.g. if:

        U+0915, U+200C, U+093F (ka, ZWNJ, i matra)

would regularly reorders as:

        093F, 0915, 200C (i matra, ka, ZWNJ)

producing the following sequence of glyphs

        i_matra, ka

The ZWNJ would simply be there to prevent a hypothetical single-glyph
sequence:

        ka_i_matra_ligature
        
But, by which rule is this not possible? Can you refer to a passage in the
book or in a UTR?

Is it because of the rule that a ZWNJ preceding a "non spacing mark"
represents the mark in isolation?

In this case, I wonder, should that rule really apply to a character like
U+093F? After all, it is actually classified as a "reordrant vowel sign",
not as a "non spacing mark".

_ Marco



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:21:13 EDT