Data constraints

Thread View: comp.text.sgml

2 messages

[T] Back to Threads [V] Tree View [G] Group Articles

2 total messages Started by null Fri, 06 Nov 1992 15:54

#378

Author: null
Date: Fri, 06 Nov 1992 15:54

26 lines
1192 bytes

Newcomer to SGML seeks assitance,

We are trying to define an element which may occur in text as one of a set 
of allowed character strings. That is a kind of PCDATA item constrained to
be a member of a certain set of values.

I would like to be able to write something like :-

<!ELEMENT element - O ("Option1" | "Option2" | "Option3") >

Or use parameter reference entities to replace the text:-

<!ELEMENT element - O ( %option1; | %option2; | %option3;) >

<!ENTITY % option1 "Option1" >
<!ENTITY % option2 "Option2" >
<!ENTITY % option3 "Option3" >

Unfortunately these naive approaches don't work with the SGMLS parser
that we are using and I havn't been able to find the answer in the 
ISO standard.  Any ideas please?

Regards

Steve Hook
s.hook@mcs.surrey.ac.uk

Re: Data constraints

#379

Author: drmacro
Date: Fri, 06 Nov 1992 17:35

59 lines
2788 bytes

In <1992Nov6.155434.24122@EE.Surrey.Ac.UK> Steve Hook writes:

>Newcomer to SGML seeks assitance,

>We are trying to define an element which may occur in text as one of a set
>of allowed character strings. That is a kind of PCDATA item constrained to
>be a member of a certain set of values.
>s.hook@mcs.surrey.ac.uk

SGML provides no way to constrain the values of PCDATA, however,
HyTime (ISO 10744) does, with the HyLex function.  (Of course,
your parser or application has to support HyLex, but we'll assume
it does for argument).  HyLex is an extension of the SGML
content model notation that allows literals and the use of
"lexical type names" defined elsewhere (either in the HyTime
standard or in the definition of your application, for example,
in an SGML-encoded style definition document).

For the designs I've been working on, I've defined a LEXMODEL=
attribute that defines the lexical rules for one or more attributes
and/or the element content, something like this:

<!-- Note:  in the lexmodel, '<'and '>' are really square brackets -->
<!ELEMENT  zipcode  - - (#PCDATA) -- Contains a zip code -->
<!ATTLIST  zipcode
             LexModel CDATA #FIXED
                   "#CONTENT (<0-9>, <0-9>, <0-9>, <0-9>, <0-9>,
                              ("-", <0-9>, <0-9>, <0-9>, <0-9>)?)"
--
  US ZIP code is a 5-digit number optionally followed by a 4-digit
  number separated by a dash.
--
>

The HyLex definition of LEXMODEL is:

<HyLex fn=lexmodelatt>
(("#CONTENT" | attname), lexmodel)+
</HyLex>
<HyLex fn=lexmodel>
LexicalTypeName | ("(", HyLexExpr, ")")
</HyLex>

In other words, one or more attribute name/lexmodel pairs,
with the reservered name "#CONTENT" letting you define
the rules for the element content.

I'm not sure the HyTime standard actually defines the LexModel=
attribute, at least as I've used it here, but it doesn't need to for
me to be able to define its use in my application.  I'm sure any
parser vendor could add generic HyLex support to their product without
any great difficulty.  They may already have the base for it in some
proprietary way.

Eliot Kimber                      Internet:  drmacro@ralvm13.vnet.ibm.com
Dept E14/B500                     IBMMAIL:   USIB2DK9@IBMMAIL
Network Programs Information Development     Phone: 1-919-543-7091
IBM Corporation
Research Triangle Park, NC 27709

Thread Navigation

This is a paginated view of messages in the thread with full content displayed inline.

Messages are displayed in chronological order, with the original post highlighted in green.

Use pagination controls to navigate through all messages in large threads.

Back to All Threads