This is the mail archive of the
java-patches@gcc.gnu.org
mailing list for the Java project.
Re: RFC: caching for I/O converters
- To: Per Bothner <per at bothner dot com>
- Subject: Re: RFC: caching for I/O converters
- From: Tom Tromey <tromey at redhat dot com>
- Date: 30 Jul 2001 14:47:03 -0600
- Cc: minyard at acm dot org, Java Patch List <java-patches at gcc dot gnu dot org>
- References: <87snhn7lir.fsf@creche.redhat.com> <m2k82zid4a.fsf@kelso.bothner.com> <m3itibg4ur.fsf@wf-rch.cirr.com> <m2itib0y74.fsf@kelso.bothner.com>
- Reply-To: tromey at redhat dot com
>>>>> "Per" == Per Bothner <per@bothner.com> writes:
[ This is a continuation of an old thread.
The context is having a stateless BytesToUnicode which can be used
concurrently ]
Per> I'm thinking that if my idea makes sense, it might make even more
Per> sense to do it at the C++ level, perhaps with an interface
Per> similar to (compatible with) iconv. That has the big advantage
Per> that it can work on buffers that are not Java arrays. For
Per> example you could implement a FileReader or FileWriter without
Per> actually allocating a Java byte[]; instead it could copy in/out
Per> of lower-level page-aligned buffers. But I think such an idea
Per> will have to wait for some other time.
In 1.4 there is a new I/O system which has its own converting scheme.
I imagine the next step for us would be to become compatible with
that. I haven't looked at its to see if its performance
characteristics are good.
Tom