Re: how can I write an arabic square root- I think I've understood a little.

From: Peter_Constable@sil.org
Date: Thu Apr 04 2002 - 17:16:25 EST


On 04/01/2002 07:24:40 PM Markus Scherer wrote:

>> I believe that the current mirrored and mirrored glyph properties are
>> useful only when no help can be obtained from the font; otherwise, the
>> resolved directionality should be provided to the font, which should
>> then select the appropriate shape for each and every character,
>> regardless of the mirrored and mirrored glyph properties.
>
>
>Correct, and this is how the BidiMirroring.txt file is intended: Only if
the
>layout engine+font are not smart enough, as a fallback mechanism.

This raises a question in my mind: how is an app to know whether the layout
engine+font are smart enough? What if the app makes the wrong assumption
about the rendering components, does a character transformation, and then
the rendering components do a glyph transformation on the transformed
characters? Consider an OpenType + Uniscribe implementation: Either
Uniscribe can do a character transformation before doing cmap lookups and
calling the OpenType engine, or Unscribe can set features (as Eric
suggested), assuming the font supports those features (a reasonable
assumption: OT fonts have to be created to a spec with regard to features
that need to be used to support a given script). But (I gather) Uniscribe
currently isn't doing either. So, an app might currently be doing
something, but it doesn't have any way (that I know of) to ask a future
version of Uniscribe whether it does handle mirroring through one mechanism
or the other. In other words, it seems to me that it must be agreed that an
app should assume it is handled by Uniscribe/OT, or should assume that it
is not.

- Peter

---------------------------------------------------------------------------
Peter Constable

Non-Roman Script Initiative, SIL International
7500 W. Camp Wisdom Rd., Dallas, TX 75236, USA
Tel: +1 972 708 7485
E-mail: <peter_constable@sil.org>



This archive was generated by hypermail 2.1.2 : Thu Apr 04 2002 - 18:24:30 EST