Re: Joining hamzah

From: Arno Schmitt (arno@zedat.fu-berlin.de)
Date: Tue Aug 26 2008 - 23:35:07 CDT

Next message: Rick McGowan: "New Public Review Issue #127, Proposed Update UAX #44 Unicode Character Database"

Previous message: John Hudson: "Re: Joining hamzah"
In reply to: Andreas Prilop: "Joining hamzah"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
Mail actions: [ respond to this message ] [ mail a new topic ]

Andreas Prilop wrote:
AP> Is it okay to write the sequence
AP> U+0640 U+0654
AP> for a "joining hamzah"? Example:
AP> ??????

Bad idea -- I agree with John.
This idea occurs naturally because hamza is wrongly defined.

Hamza is a letter that is non-joining when
a) initial
b) final
c) isolated i.e. final after a right-joining letter
    -- 'non-joining,' not 'breaking' as some call it:
       because there IS a break, its shape is non joining.
Between lam and alef (which form a ligature) it is transparent.
After a dual joining letter
  AND before a dual- or right joining letter it sits on a connecting
  stroke i.e. forces the first letter to have a longer connecting
  stroke to its left.

So the font must know that something like U+0640 is needed
graphically -- without appearing in the code.

As to the *basic* shaping behaviour -- i.e. which of the four shapes
a letter takes -- hamza is ALWAYS transparent.

Arno

Next message: Rick McGowan: "New Public Review Issue #127, Proposed Update UAX #44 Unicode Character Database"
Previous message: John Hudson: "Re: Joining hamzah"
In reply to: Andreas Prilop: "Joining hamzah"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
Mail actions: [ respond to this message ] [ mail a new topic ]

This archive was generated by hypermail 2.1.5 : Tue Aug 26 2008 - 23:38:25 CDT