Mark Leisher wrote:
>Recent experimentation with bidi reordering algorithms brings up some
>questions. I am told (by native speakers) that the expression "-10%" is
>written "%-10" in Persian and "%10-" in Arabic. How is it written in
>Thaana? Syriac? Are implementers using explicit bidi codes to reorder
>expressions like this?
This has more to do with locale-dependent presentation of data (numbers,
money, date, time, etc.) than with the Unicode bidi algorithm.
I write applications for commerce and, in fact, we often wish that locale
settings also had an entry to specify the format of percentages. Especially
*negative* percentages that, in our world, mean "discounts".
But I don't see why the text you mention should need explicit bidi codes: if
backstore contains sequences like "10-%" or "-10%", after bidi reordering
the two strings should display as "%-10" and "%10-" respectively, with no
need of explicit bidi codes (or even "implicit" codes like LRM or RLM). But
my understanding of the bidi algo is quite poor: am I missing something?
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:56 EDT