[r6rs-discuss] Unicode issues

Shiro Kawai shiro at lava.net
Tue Aug 28 16:04:48 EDT 2007


From: John Cowan <cowan at ccil.org>
Subject: Re: [r6rs-discuss] Unicode issues
Date: Tue, 28 Aug 2007 15:22:33 -0400

> William D Clinger scripsit:
> 
> > Implementations of the R6RS may use variable-width
> > representations while achieving O(1) amortized time for
> > string-length, string-ref, and string-set!.
> 
> Indeed.  What's more, there are four classes of programs:
> 
> a) those which need O(1) read/write access to characters in strings;
> 
> b) those which need O(1) read (but not write) access;
> 
> c) those which need amortized O(1) access;
> 
> d) those which do not need O(1) access at all.
> 
> Alas, there is no knowing a priori which Scheme implementations will be
> used to execute which applications, so it's hard to make the appropriate
> tradeoffs in advance.

True.
I'd rather cound on higher-level string/text srfis which allows
implementations to provide optimized high-level string operations,
bypassing index-based access.

--shiro





More information about the r6rs-discuss mailing list