From: TAKAHASHI Masayoshi <maki@...>
Date: 2002-02-08T03:01:06+09:00
Subject: Re: XML::SAX2 critique

Hello,

First of all, I'm very glad to read a critique and suggestion by
heavy user of SAX (and also author of XML::SAX::PurePerl!),
especially SAX in Perl, one of lightweight-languages like Ruby.


msergeant@star.net.uk (Matt Sergeant) wrote:
> XML::SAX2 seems to be an almost verbatim copy of the Java SAX2 APIs.

Yes, almost, with some exceptions.
For example, characters(chars) doesn't use start and length.

> I strongly believe that they have made a very poor choice in their API
> by passing the node information to the callbacks as a list of
> parameters.

Hmm...

> Now picture a cleaner API:
> 
>   handler.start_element(el)
> 
> Where element is a lightweight class Element [1], which encapsulates
> the above information and the attributes. Simple to construct, but
> most importantly: extensible. This is how the Perl SAX APIs have been
> designed, right from the start, and it meant that Perl SAX1 was mostly
> compatible with Perl SAX2 [2].
> 
> I really urge people to consider the implications of something like
> this - it's a major design decision - if you carve in stone the Ruby
> SAX2 API now, and copy the Java design, you may regret it.

I consider this issue before. The reasons I choose this
interface are:

 1. compatibility of other language's SAX API
 2. speed

I know that Perl use not strings but object as parameter of
startElement, but C, Java and Python use strings, and
I don't know other language which use Perl sytle API.

In addition, creating objects need cost. Yes, this cost is
not so expensive, and Ruby usually pays many many cost :-)
But IMO, SAX should be sensitive about cost. If we can ignore
cost, we can use, say, tree API like REXML. It maybe more
Ruby-stic and easy to use.


But....

I think that compatibility is not so important, and
simplicity and extensibility is very important. Matt's
suggestion is thoughtful.

At least, I should write other API using Element object
and compare them.


By the way:
I don't catch up with Perl-XML ML, but (maybe last year)
I found discussion about ParserDetails.ini in Perl SAX2.
I think Ruby also need similar configuration, but I don't
know what conclution they come to.

Could you tell me how to use (or not use) ParserDetails.ini
in Perl?  Is there a document about it?


Thanks,

TAKAHASHI 'Maki' Masayoshi     E-mail: maki@rubycolor.org