[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: Semantics (was Re: Inheritance in XML [^*])

• From: "Michael Kay" <M.H.Kay@e...>
• To: <xml-dev@i...>
• Date: Thu, 23 Apr 1998 19:28:26 +0100

>If XML had no semantics, then XSL, XLL and the DOM would
have to
>explicitly describe the mapping from syntactic features to
the abstract
>nodes that they work on. But they do not, because XML has
semantic
>concepts like "element, "element type", "notation" and
"attribute" that
>are *described by* the syntax.


If those are semantic concepts, then what are the syntactic
concepts?

The syntax of a language defines which sequences of symbols
constitute legitimate sentences in that language. The
semantics of the language define how to interpret the
sentences of the language as statements about objects in
some external world.

So if you take my GedML as an example, the syntax of GedML
is defined partly by the XML specification and partly by the
GedML DTD. The semantics of GedML are the rules that say an
<INDI> element asserts the existence of a person and a <FAM>
element asserts the existence of a married or unmarried
couple, etc. Of course some of the semantic constraints (a
person can only have one father) can be expressed as
syntactic constraints, while others (a person must be born
after their father) cannot, because the syntactic
metalanguage used for XML is insufficiently powerful.

Tim is surely right: there are no semantics implied by the
XML spec. XSL, XLL, etc don't change this. They define
syntactic (or symbolic) manipulations of XML constructs
which are probably most useful if you used XML the way its
designers expected you to, but they do not make any
assumptions about the semantics of your XML content. This
would become obvious if the XSL spec were rewritten (as it
could be) in a purely formal way to define a transformation
from one sequence of symbols into another.

Mike Kay, ICL


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

• Prev by Date: XLink comment and queries
• Next by Date: help me understand DOM XML Level 1
• Previous by thread: XLink comment and queries
• Next by thread: Re: Semantics (was Re: Inheritance in XML [^*])
• Index(es):
  • Date
  • Thread