Different Bidirectional Character Types

[Richard Wordingham]

On Wed, 13 Jul 2022 08:03:07 -0700
Asmus Freytag via Unicode <unicode at corp.unicode.org> wrote:

> On 7/13/2022 7:51 AM, r12a via Unicode wrote:
> Asmus Freytag wrote on 13/07/2022 15:43:
> > On 7/13/2022 2:58 AM, r12a wrote:
> >> The approach differs not only by script and which digits are used,
> >> but also by language.  Arabic and Persian both use the Arabic
> >> script, but do things differently when it comes to ordering
> >> components of a range or expression.
> >>> 
> >> Isn't that difference handled by having two different sets of
> >> digits? As opposed to relying on a language tag.
> >> 
> > See the cases in figs. 3 and 4 at
> > https://r12a.github.io/scripts/arabic/arb.html#expressions.  Same
> > digits, different expectations about directionality.
> > 
> > I wasn't talking about language tags or behaviour arising from
> > character properties (indeed the language tag doesn't make a
> > difference) – i was talking about the user expectations differing
> > from language to language about the order in which digits appear in
> > the text.
> > 
> If I understand correctly, this would be a case that's not handled by
> the UBA, then. Would that be worth calling out, you think?

And to answer the original question, it would be good to start with the
user expectations, and then explain how the UBA reduces (does it?) the
jiggery pokery required of the typist to get the desired outcome.  In
particular, we seem to be exploiting a difference in glyph styles and
promoting it it a character difference to get a left-to-right ordering.

Richard.

Richard.