[racket] Unicode Question

Yes, that's a bug. I've pushed a repair.

The bug show up when the composed form is more than 16 characters
shorter than the decomposed form, due to a mistake in an
avoid-wasted-space step of the implementation.

At Tue, 22 Jan 2013 17:33:21 -0800, Nick Sivo wrote:
> Hi,
> 
> In the attached file, use of string-normalize-nfc truncates the
> string.  Is this expected?  I tried on Mac and Windows, using 5.2.1,
> 5.3.1, and the latest nightly.  According to
> http://unicode.org/reports/tr15/ form C might alter the bytes, but
> shouldn't change the displayed form, let alone truncate.
> 
> Here's a gist if that's preferable: https://gist.github.com/7de39c77f98fc45cd03d
> 
> Also, I have no idea what the text says.  If it's
> inappropriate/offensive, I apologize.
> 
> Thanks,
> Nick