On 9/11/2013 5:24 PM, Khaled Hosny wrote:
On Wed, Sep 11, 2013 at 02:37:35PM +0200, Hans Hagen wrote:
anyway, it would be nice to see where the three methods fail:

Of the three, method one seems to give correct results on all the given
tests (I’m yet to test with my own documents).

I’m skeptical that bidi can be implemented in a one pass algorithm,
there have been a one pass algorithm called “Pretty Good Bidi
Algorithm”, but it has its limitation (I never tested it myself).

Sure, although it can be close to okay with some backward and forward scanning but I'm nbot really in the mood for that now. Anyhow, for the occasional mix of arabic and latin this method works ok. For more extreme cases method 'one' will do and method 'two' ... well it depends on developments in unicode as this method will be the more configurable one. (And I can probably make a faster implementation of method two when performance matters.)


                                          Hans Hagen | PRAGMA ADE
              Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
    tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com
                                             | www.pragma-pod.nl
If your question is of interest to others as well, please add an entry to the 

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context
webpage  : http://www.pragma-ade.nl / http://tex.aanhet.net
archive  : http://foundry.supelec.fr/projects/contextrev/
wiki     : http://contextgarden.net

Reply via email to