I think your comparison should not be all encompassing but based upon the
problem you are trying to solve.  Object Oriented programming is naturally
pluggable so switching from one type of parser to another should be
relatively simple depending on the problem you are trying to solve.

Examples:
- If you have a project/program that is using XML for intra-processes
communication, then you probably don't need a validating parser because
there are no outside processes creating or consuming the XML. In that case
you might want to compare all the non-validating parsers for performance,
memory usage, stability, and conformance to the SAX1 and SAx2 API (I
recommend Aelfred from Saxon).

- If you have a project/program that needs to validate XML (extra-process
communication) then you might compare parsers that support DTDs and Schema.
You should still compare their performance (relative to validating parsers
not non validating), memory usage, and stability.  In addition, you may want
to consider (in the case of schemas) how closely they support the W3C's
specs.

- If you have a project/program that needs a DOM parser then the basic
comparisons should still apply but you might want to look at any additional
features that Implementation of the DOM provides.  For example, Xerces'
DocumentImpl class provides quite a few useful and cool features that the
DOM interface does not provide.

In a nutshell I'm saying, don't take a monolithic approach to solving any
problem or domain of problems.  It's better to find or write classes that
solve a particular problem or sets of problems well than a class that solves
a problem or sets of problems but does it mediocre.

Dane Foster
Equity Technology Group, Inc
http://www.equitytg.com.
954.360.9800
----- Original Message -----
From: "Andy Clark" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Tuesday, April 10, 2001 9:48 PM
Subject: Re: parsers


Ismail Nalwala wrote:
> I am looking for a reasonable way to compare XML parsers based on
performance,
> stability and functionality(e.g. XML schema validation).  The candidates
that I
> am interested in are JDOM, xerces and crimson.    Any help or information
in

JDOM is a data structure, not a parser. Also, you have to decide
what your required feature set is before selecting a parser. If
parser XYZ is the fastest parser in the universe but doesn't
have the features you need, who cares how fast it is.

--
Andy Clark * IBM, TRL - Japan * [EMAIL PROTECTED]

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]




---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to