cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ivelin Ivanov" <ive...@iname.com>
Subject Re: [Schematron-love-in] Re: [Announcement] Fast Schematron Validation Here !
Date Thu, 28 Mar 2002 05:55:31 GMT

I'm cross-sending this email to Schematron and Cocoon dev-lists, because
we're discussing problems of common interest.


----- Original Message -----
From: "Rick Jelliffe" <ricko@allette.com.au>
To: <schematron-love-in@lists.sourceforge.net>
Sent: Wednesday, March 27, 2002 11:38 PM
Subject: Re: [Schematron-love-in] Re: [Announcement] Fast Schematron
Validation Here !


From: "Ivelin Ivanov" <ivelin@iname.com>

> A question was brought up on the Cocoon dev list.
>
> Can the phases tag be kept separate from the schema.

>Sure.

>All specific processing semantics of Schematron are
implementation-dependent:
>what happens when an assertion fails, which phases are active, which
elements
>are being tested, which order information items are traversed, etc.

>If you want to have externally-specified phases or to dynamically select
which
>patterns to run, that is fine.

>Once you get inside a pattern, it is a little different: you cannot
arbitrarily run
>rules as they are because they are lexically related: so if you have
  <pattern>
     <rule context="c1">
        ...
     </rule>
     <rule context="c2">
        ...
     </rule>
   </pattern>
>and you wanted to run the second rule only against a particular information
item,
>the actual context is
    not(c1) and c2
>which requires more testing than people may expect.

>The lack of semantics is why I try to encourage people to make general
>statements in <assert> statements "An X should have a Y" rather than
>"Error: you are hopeless, why don't you quit".  The <diagnostic> element
>is provided for that.

>> If the underlying model doesn't change and the full set of patterns is
the same,
>> then when adding support for new devices, wizards, etc. to build a
document
>> instance,  the rules for partial validation should be separate from the
description of
>> the model.

>Sorry, I don't understand this sentence: what do you mean by "devices"
here?

I mean browser, client. Different browsers (PC, PDA, cell-phone, etc.) may
support different human interfaces and therefore the document may be split
into different pieces which are gathered and put together at the end.
The validation of the pieces at each stage is device/client dependent.
Is the question more clear?

> > Does this question make sence? What do you suggest?

>I suggest people experiment and do whatever they can to get their jobs done
>and make life simpler and richer :-)    The point of Schematron is not to
>make a monolithic, ultimate validation system, but to provide a toolkit
>and a different vocabulary to help people solve some big practical problems
>with  minimal fuss, using technology that places human language at the
centre
>(rather than at the periphery, in "documentation' elements.)  If you come
>up with some nice new way to use the statements in a Schematron schema,
>you will only get respect.

>And please pass on to the coccoon people that if they have ideas for
>abstractions or hooks that might enhance Schematron, please feel free
>to prototype them and let us know.

We sure will.


Cheers,

Ivelin


>

Cheers
Rick Jelliffe



_______________________________________________
Schematron-love-in mailing list
Schematron-love-in@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/schematron-love-in


---------------------------------------------------------------------
To unsubscribe, e-mail: cocoon-dev-unsubscribe@xml.apache.org
For additional commands, email: cocoon-dev-help@xml.apache.org


Mime
View raw message