Nice! I missed that blog. So that this mean that plain vanilla Android still does not have proper BiDi support?
You didn't say it explicitely, but I understand from what you wrote that you choose JNI to access ICU for your BiDi solution. Was that JNI layer ever introduced to the open source Android to replace their buggy Java implementation? Regards, Dov On Sun, Sep 26, 2010 at 10:48, Shachar Shemesh <shac...@shemesh.biz> wrote: > Dov Grobgeld wrote: > > I'm curious. Did your Hebrew support modifications include full BiDi > support? Did you write the bidi code yourself, or did you use some known > library? > > I wrote about it extensively at the time: > http://blog.shemesh.biz/?p=726 > > As for the BiDi support, each Android installation already arrives with TWO > BiDi implementation. One inside the framework and another inside ICU for > C++, which is compiled for the platform in its completion, but only used, as > far as I can tell, to perform the BiDi character classification. It might be > used by non-Java code, or else why they did not use ICU for Java is beyond > me. Why they re-implemented that BiDi algorithm (badly) and not used JNI is > also beyond me. > > As such, I saw it as a smart move not to add a third implementation. In > fact, I always try to avoid re-implementing complicated algorithms for which > good implementations already exist. As such, in all of my years of dong BiDi > work on various platforms, the number of complete implementations of the UBA > I have coded is zero. This is a fact I am, actually, proud of. > > Shachar > > -- > Shachar Shemesh > Lingnu Open Source Consulting Ltd.http://www.lingnu.com > >
_______________________________________________ Linux-il mailing list Linux-il@cs.huji.ac.il http://mailman.cs.huji.ac.il/mailman/listinfo/linux-il
