From: "James Britt (rubydev)" <james@...>
Date: 2001-11-23T03:48:45+09:00
Subject: [ruby-talk:26165] Re: Conformance Test of XML Parsers in Ruby(20011122)

> --- David Alan Black <dblack@candle.superlink.net
> > > > I don't think 99% conformant parser is useless
> > > > because it's just not 100%,
> >
> > It's pretty close to useless... at least it won't be
> > taken seriously by people who don't want to have to
> > worry about the 1%, possibly.
>
> I totally disagree. There are plenty of cases where the
> same app is responsible for both writing and reading all
> the XML. In many cases the XML is so simple that it would
> hard for it to break any parser. In fact, the few times
> that I have used XML so far have been in exactly those
> conditions.

Referring to something as an XML parser|DOM|whatever brings certain expectations.  If only 99% of the expectations are met, then the
details of that remaining 1% are critical.

I've had many cases where the internal manipulation of markup did not require a full-featured XML parser.  Typically there was no
need for entity resolution, or recognizing DTDs, or even whether the "document" had a single root element.  A custom parser that did
exactly what was needed was much faster than any off-the-shelf parser.

What the larger application accepted was XML, and what it emitted was XML, but there isn't a good name for the markup/data while in
that middle processing stage.  Perhaps it's a problem of terminology; there's no such thing as "almost XML".  But "almost XML"
processors can be very useful.  What do you call them? What do you call the stuff they handle?


James