Map xs:int to Int, xs:integer to BigInt Inbox X

[Leif Warner]

From what I've read it seems xs:integer is supposed to be an arbitrary-precision integer type.  It's a subtype of decimal.

Currently it gets mapped to Int, while xs:int doesn't look to be handled.

Here's a patch to sort that out.  Not sure what naming you'd like to use, e.g. for the /asInt or /asInteger (/asBigInt?) part.

I'm used to that convention from Haskell, where "Int" is the int32 type, and "Integer" is the arbitrary precision one, but on the Java platform that might be confusing, as historically "Int" is the primitive int32, and "Integer" is just an object wrapper for that.

FYI, I came across this when turning a Scardf augmentation rule into a Jena rule.  The Util.makeIntNode [1] method puts an xs:int in there, and /asInt would complain and refuse to work on those values because it was expecting something marked xs:integer.

xs:integer is what I get when I put a bare numeric literal into a turtle file and run rapper (Redland utils) on it, though.  If there's a bunch on needless BigInts in there, they could be checked (e.g. BigInt(foo).abs < MAX_INT) and converted to Int, at the expense of some static typing, however.

[1] http://jena.sourceforge.net/javadoc/com/hp/hpl/jena/reasoner/rulesys/Util.html#makeIntNode(int)

-Leif

[Hrvoje Simic]

Thanks for this.

I was thinking of naming the TypeNodeConverter[BigInt] converter "asBigInt".

Also, I'm thinking that TypeNodeConveters should be allowed to convert
literals of multiple types, since integers in RDF are usually
xsd:integer but Ints in Scala. So asInt could make Ints from both
xsd:int and xsd:integer typed literals - considerations of integer
size would be left to the developer.

Hrvoje

[Hrvoje Simic]

I just checked in the revision 218:
http://code.google.com/p/scardf/source/detail?r=218

It's mixed in with a bigger design change concerning predicate trees.
Sorry for that.

Now the TypeNodeConverter class takes the function UriRef => Boolean
which determines which XML Schema types can be converted. The idea is
that, although most RDF has integers as xsd:integer, you would want to
work with Ints in Scala. Now you can use either asInt or asBigInt,
depending on your needs.

If you have the time, please check if this implementation matches your
expectations.

Other converters could be extended to allow more XML types. E.g.
asBigDecimal could take xsd:integers.

Hrvoje

[Leif Warner]

Yes, this new code seems to work for all the cases - thanks!
-Leif

On Feb 7, 1:43 am, Hrvoje Simic <hrvoje.si...@gmail.com> wrote:
> I just checked in the revision 218:http://code.google.com/p/scardf/source/detail?r=218
>
> It's mixed in with a bigger design change concerning predicate trees.
> Sorry for that.
>
> Now the TypeNodeConverter class takes the function UriRef => Boolean
> which determines which XML Schema types can be converted. The idea is
> that, although most RDF has integers as xsd:integer, you would want to
> work with Ints in Scala. Now you can use either asInt or asBigInt,
> depending on your needs.
>
> If you have the time, please check if this implementation matches your
> expectations.
>
> Other converters could be extended to allow more XML types. E.g.
> asBigDecimal could take xsd:integers.
>
> Hrvoje
>
>
>
>
>
>
>
> On Thu, Feb 3, 2011 at 19:38, Hrvoje Simic <hrvoje.si...@gmail.com> wrote:
> > Thanks for this.
>
> > I was thinking of naming the TypeNodeConverter[BigInt] converter "asBigInt".
>
> > Also, I'm thinking that TypeNodeConveters should be allowed to convert
> > literals of multiple types, since integers in RDF are usually
> > xsd:integer but Ints in Scala. So asInt could make Ints from both
> > xsd:int and xsd:integer typed literals - considerations of integer
> > size would be left to the developer.
>
> > Hrvoje
>

> > On Thu, Feb 3, 2011 at 19:21, Leif Warner <abimel...@gmail.com> wrote:
> >> From what I've read it seems xs:integer is supposed to be an
> >> arbitrary-precision integer type.  It's a subtype of decimal.
> >> Currently it gets mapped to Int, while xs:int doesn't look to be handled.
> >> Here's a patch to sort that out.  Not sure what naming you'd like to use,
> >> e.g. for the /asInt or /asInteger (/asBigInt?) part.
> >> I'm used to that convention from Haskell, where "Int" is the int32 type, and
> >> "Integer" is the arbitrary precision one, but on the Java platform that
> >> might be confusing, as historically "Int" is the primitive int32, and
> >> "Integer" is just an object wrapper for that.
> >> FYI, I came across this when turning a Scardf augmentation rule into a Jena
> >> rule.  The Util.makeIntNode [1] method puts an xs:int in there, and /asInt
> >> would complain and refuse to work on those values because it was expecting
> >> something marked xs:integer.
> >> xs:integer is what I get when I put a bare numeric literal into a turtle
> >> file and run rapper (Redland utils) on it, though.  If there's a bunch on
> >> needless BigInts in there, they could be checked (e.g. BigInt(foo).abs <
> >> MAX_INT) and converted to Int, at the expense of some static typing,
> >> however.

> >> [1] http://jena.sourceforge.net/javadoc/com/hp/hpl/jena/reasoner/rulesys/...)
> >> -Leif