DFDL support for Infinity and Nan need clarification. There are two aspects
  1. Representation of infinity and Nan in the infoset
  2. Representation of Infinity and Nan in the datastream


Representation of infinity and Nan in the infoset

DFDL should follow XML which allows +inf, -inf and Nan for Float and Double

From XML Schema Part 2: Datatypes Second Edition

[Definition:]  float is patterned after the IEEE single-precision 32-bit floating point type [IEEE 754-1985]. The basic ˇvalue spaceˇ of float consists of the values m × 2^e, where m is an integer whose absolute value is less than 2^24, and e is an integer between -149 and 104, inclusive. In addition to the basic ˇvalue spaceˇ described above, the ˇvalue spaceˇ of float also contains the following three special values: positive and negative infinity and not-a-number (NaN). The ˇorder-relationˇ on float is: x < y iff y - x is positive for x and y in the value space. Positive infinity is greater than all other non-NaN values. NaN equals itself but is ˇincomparableˇ with (neither greater than nor less than) any other value in the ˇvalue spaceˇ.
The special values positive and negative infinity and not-a-number have lexical representations INF, -INF and NaN, respectively. Lexical representations for zero may take a positive or negative sign.  

Similarly description for Double

Minor changes to infoset [dataValue] description

1.        [dataValue] The value in the value space (as defined by XML Schema Part 2: Datatypes [XSDLV1] ) of the [datatype] member  or special value nil. In a complex element information item this member has no value. For information items of datatype xs:string, the value will be the  ISO10646 character codes of the string and 'implicit' (also known as logical), left-to-right bidirectional ordering and orientation.


Is it sufficient just to refer to XML schema rather than spell out the value spaces in this specification?


Representation of Infinity and Nan in the data stream

ICU allows a single character for Nan and infinity. The infinity character can be prefixed by the positive or negative prefix (ie the prefix part of he pattern)

This is too restrictive for DFDL as we need to be able to support at least  'INF' '-INF' and 'NaN'


Only minor changes to current description.

numberInfinityRep String

The value used to represent infinity.

Infinity is represented as string with the positive or negative prefixes and suffixes from the numberPattern applied

This property is applicable when dfdl: textNumberRepresentation is 'standard' and the simple type  type is float or double.

Annotation: dfdl:textNumberFormat

numberNaNRep String

The value used to represent NaN.

NaN is represented as string and the positive or negative prefixes and suffixes from the numberPattern are not used

This property is applicable when dfdl: textNumberRepresentation is 'standard' and the simple type type is float or double.

Annotation: dfdl:textNumberFormat


 



Regards
 
Alan Powell
 
Development - MQSeries, Message Broker, ESB
IBM Software Group, Application and Integration Middleware Software
-------------------------------------------------------------------------------------------------------------------------------------------
IBM
MP211, Hursley Park
Hursley, SO21 2JN
United Kingdom
Phone: +44-1962-815073
e-mail: alan_powell@uk.ibm.com



Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU