<html><head></head><body>The expected shape is e2|p1l3|v1|a2 and I get this. You will have to add a p1xx(half) glyph to break it. -suresh <div class="gmail_quote">On 28 February 2015 9:04:24 am IST, Rajeesh K V <rajeeshknambiar@gmail.com> wrote:<blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;"> <pre class="k9mail">On Sat, Feb 28, 2015 at 3:34 AM, Behdad Esfahbod <behdad@behdad.org> wrote: <blockquote class="gmail_quote" style="margin: 0pt 0pt 1ex 0.8ex; border-left: 1px solid #729fcf; padding-left: 1ex;"> Hi Rajeesh, Please send the font file, so I can confirm this is behaving differently from Uniscribe.</blockquote> Hi Behdad, Link to the font <a href="http://smc.org.in/downloads/fonts/meera/Meera.ttf">http://smc.org.in/downloads/fonts/meera/Meera.ttf</a> <blockquote class="gmail_quote" style="margin: 0pt 0pt 1ex 0.8ex; border-left: 1px solid #729fcf; padding-left: 1ex;"> Thanks, behdad On 15-02-27 02:15 PM, Rajeesh K V wrote: <blockquote class="gmail_quote" style="margin: 0pt 0pt 1ex 0.8ex; border-left: 1px solid #ad7fa8; padding-left: 1ex;"> Though found an interesting case: hb-unicode-encode 0D2A 0D4D 0D32 0D4D 0D35 0D4B | hb-shape --font-file=Meera.ttf gives the output: [e2=0+871|p1=0+1440|xx=0+318|l3v1=0+1438|a2=0+848] The expected outcome is [p1|xx|e2|l3v1|a2]. Note the e2 (prebase part of two part matra 0D4B) reordered in front of p1 where it should have been placed in front of glyph l3v1. The last codepoint in input sequence 0D4B is a two part matra, split to 0D47 (e2) <glyph> 0D3E (a2). It's not a regression, an earlier build yields the same outcome as far as I can see. </blockquote></blockquote> </pre></blockquote></div></body></html>