W3C home > Mailing lists > Public > www-xml-schema-comments@w3.org > April to June 2002

RE: Is +0 a nonPositiveInteger?

From: <noah_mendelsohn@us.ibm.com>
Date: Fri, 26 Apr 2002 19:54:33 -0400
To: "Ashok Malhotra" <ashokma@microsoft.com>
Cc: "James Clark" <jjc@jclark.com>, "XML Schema Comments" <www-xml-schema-comments@w3.org>
Message-ID: <OFBA5C2D76.304BA551-ON85256BA8.00000F8A@lotus.com>
I'm afraid I'm not quick enough with regex's to fully grok the intent of 
the specific example, but I presume it allows a subset of the lexical 
forms allowed for the base dateTime type.  The value space of the derived 
type is the set of values corresponding to those lexical forms, IMO. 

In general, my understanding was that in the absence of additional 
restricting facets on the derived type, the lexical space of any derived 
type is the intersection of the lexical space of the base and the set of 
lexical forms allowed by the regex.  If there are other constraining 
facets (e.g. minExclusive) they further constrain the restricted value 
space, and thus indirectly the lexical space.

>> I'm sorry you are troubled but consider this case.

I'm more than troubled about the example I gave, I'm asking:  what is the 
lexical space?  We've gone to recommendation.  People have built and 
deployed ro our recommendation, and you're implying that there is an 
opportunity to consider the lexical space of the derived type is not even 
defined?  I have always understood our recommendation to work as described 
above (though as you know, I have long suggested that we should devote 
some of our formalization efforts to our Datatypes specification -- such 
formalization would surely clarify any ambiguities.)  Many thanks...maybe 
I'm missing something.

------------------------------------------------------------------
Noah Mendelsohn                              Voice: 1-617-693-4036
IBM Corporation                                Fax: 1-617-693-8676
One Rogers Street
Cambridge, MA 02142
------------------------------------------------------------------







"Ashok Malhotra" <ashokma@microsoft.com>
04/26/2002 06:10 PM

 
        To:     <noah_mendelsohn@us.ibm.com>
        cc:     "James Clark" <jjc@jclark.com>, "XML Schema Comments" 
<www-xml-schema-comments@w3.org>
        Subject:        RE: Is +0 a nonPositiveInteger?


I'm sorry you are troubled but consider this case.

<xsd:simpleType name="myTime">
  <xsd:restriction base="xsd:dateTime">
    <xsd:pattern 
 
value="[0-9]{2}:[0-9]{2}:[0-9]2.[0-9]*(Z|[+-]?[0-9]{2}:[0-9]{2})?."/>
  </xsd:restriction>
</xsd:simpleType>

What is the value space and lexical spece of the derived type and how
are they related?

All the best, Ashok 
===========================================================


-----Original Message-----
From: noah_mendelsohn@us.ibm.com [mailto:noah_mendelsohn@us.ibm.com] 
Sent: Friday, April 26, 2002 2:35 PM
To: Ashok Malhotra
Cc: James Clark; XML Schema Comments
Subject: RE: Is +0 a nonPositiveInteger?

Ashok Malhotra writes:

>> I have come to the conclusion that when you 
>> derive a simple type you should not inherit 
>> the lexical representation of the base type. 

I find this very problematic.  The primer shows:

<xsd:simpleType name="myInteger">
  <xsd:restriction base="xsd:integer">
    <xsd:minInclusive value="10000"/>
    <xsd:maxInclusive value="99999"/>
  </xsd:restriction>
</xsd:simpleType>

Are you implying that there is no specified lexical space for this type?

How can that work?

------------------------------------------------------------------
Noah Mendelsohn                              Voice: 1-617-693-4036
IBM Corporation                                Fax: 1-617-693-8676
One Rogers Street
Cambridge, MA 02142
------------------------------------------------------------------







"Ashok Malhotra" <ashokma@microsoft.com>
Sent by: www-xml-schema-comments-request@w3.org
04/24/2002 09:34 AM

 
        To:     "James Clark" <jjc@jclark.com>, "XML Schema Comments" 
<www-xml-schema-comments@w3.org>
        cc:     (bcc: Noah Mendelsohn/Cambridge/IBM)
        Subject:        RE: Is +0 a nonPositiveInteger?


James:
I have come to the conclusion that when you derive a simple type you
should not inherit the lexical representation of the base type.  If you
do,
you open your self to a host of problems which we can discuss another
time.

Thus, simple type derivation merely gives you a new, more restricted
value 
space.  You can then go ahead an specify a lexical space for this
restricted
value space and specify a mapping from the lexical to the value space.

If you look at it this way, nonPositiveInteger has a value space
consisting
of 0 and the negative integers.  Its lexical space consists of 0 and
strings of digits preceded by a minus sign.

All the best, Ashok 
===========================================================


-----Original Message-----
From: James Clark [mailto:jjc@jclark.com] 
Sent: Wednesday, April 24, 2002 3:23 AM
To: XML Schema Comments
Subject: Is +0 a nonPositiveInteger?

Is +0 allowed as a nonPositiveInteger?  At the moment there's a
contradiction. 3.3.14.1 says "nonPositiveInteger has a lexical
representation consisting of a negative sign ("-") followed by a
finite-length sequence of decimal digits (#x30-#x39). If the sequence of
digits consists of all zeros then the sign is optional."  This doesn't
allow
+0.  On the other hand 0 is in the value space of nonPositiveInteger and
+0
is a legal representation of ) in the lexical space of integer.

Either

(a) the prose in 3.3.14.1 needs fixing, or

(b) the schema for schema needs to add a pattern facet to the definition
of
nonPositiveInteger that excludes +0

If you do (b), then you will probably want to fix nonNegativeInteger to
disallow "-0". However, at the moment there's no contradiction since the
prose for nonNegativeInteger allows "an optional sign" not just an
optional
positive sign.

James
Received on Friday, 26 April 2002 20:06:40 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Sunday, 6 December 2009 18:13:00 GMT