[Indic] Do Reph repositioning in final reordering like the spec says
authorBehdad Esfahbod <behdad@behdad.org>
Thu, 10 May 2012 11:45:52 +0000 (13:45 +0200)
committerBehdad Esfahbod <behdad@behdad.org>
Thu, 10 May 2012 11:45:52 +0000 (13:45 +0200)
commitdbb105883c6e9b83e78dc8b10766cd56b98cd7e1
tree3a019bbc1954534eb3b54c922763f4803332f7e7
parent4705a7026900e51f6430f03a73c87f2df035df92
[Indic] Do Reph repositioning in final reordering like the spec says

This introduced a failure, which we tracked down to a test case like this:

  U+092E,U+094B,U+094D,U+0930

The final character is a Ra that should be put in a syllable of it's
own.  And we do.  But it will interact with the Halant before it.  So
now we finally are convinced that we have to limit features to syllable
boundaries.  That's coming after lunch!
src/hb-ot-shape-complex-indic-private.hh
src/hb-ot-shape-complex-indic.cc