General comments on parsers

Don Park donpark at quake.net
Thu Dec 11 16:23:35 GMT 1997


The situation is complicated by the fact that W3C is working on and has not
yet released its own version of Java XML Object Model.  Since it will be
difficult to have all existing Java XML parsers to conform to a single
object model, I think the best approach is for someone to write a new Java
parser framework which provides a reasonable object model and acts as the
Universal XML Parser (UXP?:-).

UXP should use some kind of simple registry scheme and a UI to allow users
to plug in new UXP compatible parsers.  Writing UXP adapters for each of
existing Java XML parsers should not be too hard.  Once UXP is in place, new
parsers will start to conform.  When W3C XML API is out, all we need to do
is write two adapters:

1) UXP to W3C adapter so programs using W3C XML API can use UXP parsers
(i.e. JavaScript).
2) W3C to UXP adapter so programs using UXP can use any XML parsers
providing W3C XML API.

BTW, I have taken a look at Xapi-J and W3C OM API and, frankly, I am not
satisfied with either of them.  Enumeration by index is problematic and
callbacks are either not supported or primitive.  Not that I can offer any
better in the near future <g>.  Call me a stuck up critic, if you will.

Don

>Yes, please. This list (especially John Tigue) worked hard to come up with
>Xapi-J - everyone seemed to think it was a good way forward, but no parsers
>implement it.  Instead we have an increasing (and rather difficult) variety
>of approaches (and especially terminology). For example, it's clear that
>AElfred and Lark use 'Entity' in different ways [I'm slightly confused by
>Lark's use of Entity].
>
>Parsers are NOT equivalent, and there are many reasons why an application
>may wish to use more than one.
> - different interfaces, giving different views of the document
> - different optimisations of speed, memory, etc.
> - different treatment of entities
> - different features
>
>It's very tedious to have to implement different interfaces for each
>(AElfred has about 30 methods - and they are all valuable). So:
> - Chris
> - David
> - James
> - John
> - Norbert
> - Tim
>any comments on a common interface :-)?
>
> P.
>
>Peter Murray-Rust, Director Virtual School of Molecular Sciences, domestic
>


xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/
To (un)subscribe, mailto:majordomo at ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)




More information about the Xml-dev mailing list