[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qexo-general] qexo speed
From: |
Per Bothner |
Subject: |
Re: [Qexo-general] qexo speed |
Date: |
Sat, 17 May 2003 09:46:05 -0700 |
User-agent: |
Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.4b) Gecko/20030507 |
fl wrote:
A simple 500K xml file is processed at the same speed by both tools,
compiled qexo being a bit faster.
That's reassuring, though we can probably do better.
Do you have any numbers or predictions about qexo versus compiled qexo
versus xslt versus stx or others ?
Even "non-compiled" Qexo ends up being compiled. However,
pre-compilation says a little bit of compilation time,
plus a few more optimizations may be done for pre-compilation.
The Qexo parser is quite lean, though I have a few improvements
in mind. The DOM that Qexo uses is also very compact and fast.
(It is not compatible with the standard W3C DOM.)
Note: according to some, joost is somehow as fast as SAXON...
I haven't used Saxon - I mainly use xsltproc (written in C) when
I need an xslt processor.
Note that Qexo does include an implementation of a small subset
of xslt; it would be interesting to improve enough to do some
useful tasks - and run some benchmarks.
Generally, I seldom take the time to run benchmarks, or do
performance-tuning, so it's appreciate to hear about your
experience. Maybe later this Summer I'll do a little more
performance work, but there is so much of XQuery left to
implement that feature-completion usually comes first.
--
--Per Bothner
address@hidden http://per.bothner.com/