[sword-devel] Likely miss in Arabic vowel point filter.

[Paul A. Martel]

While looking at the source code (for no really good reason) I happened to
run into the following test in the Arabic Vowel Points filter

    if (((*from == 0xFC) && (*(from + 1) >= 0xE5 && *(from + 1) <= 0x63))
|| ...

Unless I'm reading this wrong, this clause is impossible to satisfy.
So it effectively _disables_ the intended filtering of the 0xFC code page
characters.
They're probably intended to be 0x5E (not 0xE5) through 0x63, which
according to http://www.utf8-chartable.de/unicode-utf8-table.pl?start=64512all
have descriptions of the form "ARABIC LIGATURE SHADDA WITH ..."
whatever that means.

I am a little gun-shy about logging bugs that I don't know how to manifest.

But I thought I'd bring this out here rather than forget it, in case
someone else may find it useful.

Would sending this to sword-bugs at crosswire.org have been a better idea?

--paul
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.crosswire.org/pipermail/sword-devel/attachments/20111216/3edb71c8/attachment.html>