Group Email Manager

login signup
Re: [Scheme-reports] mutable unicode strings Per Bothner (02 Jul 2014 04:45 UTC)
Re: [Scheme-reports] mutable unicode strings Michael Montague (02 Jul 2014 14:10 UTC)
Re: [Scheme-reports] mutable unicode strings Per Bothner (02 Jul 2014 15:10 UTC)
Re: [Scheme-reports] mutable unicode strings Alaric Snell-Pym (02 Jul 2014 15:21 UTC)
Re: [Scheme-reports] mutable unicode strings Bear (03 Jul 2014 02:21 UTC)
Re: [Scheme-reports] mutable unicode strings Per Bothner (03 Jul 2014 02:35 UTC)
Re: [Scheme-reports] mutable unicode strings Sam Tobin-Hochstadt (03 Jul 2014 13:01 UTC)
Re: [Scheme-reports] mutable unicode strings Per Bothner 03 Jul 2014 02:31 UTC 
On 07/02/2014 07:16 PM, Bear wrote:

> Are there any extant examples of 3-byte code units in strings?
> I would find that - interesting.

I too would be surprised.

What might make sense is packing 3 20-bit (or 21-bit) characters in a 64-bit long.
That would use 2.67 bytes per character and be quite friendly to
modern hardware.  Almost certainly faster that using 32 bits per character,
due to better locality.

Probably not a good choice for Kawa, though, since I think it is more important
to implement the java.lang.CharSequence interface efficiently.
--
	--Per Bothner
per@bothner.com   http://per.bothner.com/

_______________________________________________
Scheme-reports mailing list
Scheme-reports@scheme-reports.org
http://lists.scheme-reports.org/cgi-bin/mailman/listinfo/scheme-reports