[
Lists Home |
Date Index |
Thread Index
]
Jeni Tennison wrote:
> It's interesting how close you are to Walter Perry's position here
> --that nothing matters aside from the output of your process, and that
> the input can be anything at all; it's up to the processor to work out
> what.
I believe that I have been consistently a stickler for well-formedness in
XML: my first exchange with Peter Murray-Rust on this list emphasized the
importance I attach to well-formedness, though in that case by contrast
with those who believed that something more than simple well-formedness
was required for XML to be truly useful. That said, well-formedness
checking is in the domain of the parser, and a parser of itself falls far
short of the XML processor described in the Recommendation. Patrick
Durusau can speak for himself, and may well disagree with my
understanding, but in my opinion the JITTs processor operates downstream
of XML parsing and of necessity has other inputs, and input assumptions,
than the XML which a parser by itself processes. I build processors
downstream from XML parsing in just this way and, yes, you are correct
that for the inputs for which they are designed--and which they are
specifically expert in handling--such processors must enjoy the autonomous
authority to proceed on their own rules. It is only that whatever portion
of their input is XML is subject to XML parsing, including the expected
well-formedness checks, upon every use.
Respectfully,
Walter Perry
|
