[
Lists Home |
Date Index |
Thread Index
]
- From: David Brownell <david-b@pacbell.net>
- To: David Megginson <david@megginson.com>
- Date: Tue, 21 Dec 1999 10:28:35 -0800
David Megginson wrote:
>
> Miles Sabin <msabin@cromwellmedia.co.uk> writes:
>
> > > If I were doing it over, though, I would actually call
> > > java.lang.String.intern once for each of the strings in the
> > > intern table so that they were == to the regular intern'ed
> > > versions.
> >
> > Try it, but I think you'll be more likely to lose than gain.
Having done this, I'll disagree. The cost wasn't observable.
Except ... that processing some documents blew up a fixed-size
table in at least one incarnation of the JDK 1.1.x JVM. I saw
that in a pathological stress test case (which generated random
documents), never once "in the wild", but it does provide yet
another reason to use a better JVM (like JDK 1.2.x) on servers
and in other cases where you expect the JVM to live a long time.
> I'd appreciate more information here -- if I call
> java.lang.String.intern the first time I add a string to the intern
> table, then the cost is proportional to the number of entries in the
> table, not the number of accesses.
Right. The man cost being mapping from a portion of a buffer
to the String ... a "local" intern. Going one step beyond that
to a "global" intern (using the JVM) is cheap, since it's hardly
ever done.
- Dave
xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev@ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To unsubscribe, mailto:majordomo@ic.ac.uk the following message;
unsubscribe xml-dev
To subscribe to the digests, mailto:majordomo@ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa@ic.ac.uk)
|
