Re: Should we extend xsd:decimal?

On 30/10/2022 14:57, Dan Brickley wrote:
>
>
> On Sun, 30 Oct 2022 at 10:39, Pierre-Antoine Champin 
> <pierre-antoine@w3.org> wrote:
>
>     Dear Semantic Web community,
>
>     an issue has been raised some time ago about the DCAT-3 specification
>     [1], related to the use of `xsd:decimal` as the range a property, and
>     how a seemingly correct JSON-LD document may produce ill-formed
>     `xsd:decimal` literals [2].
>
>     Exploring the possible solutions to this issue, I came to the
>     conclusion
>     that
>
>     - it is a shame that `xsd:decimal` does not support the E notation
>     [3]
>     just like `xsd:double` does, and
>     - actually, many implementations (of RDF, SPARQL, OWL...) do
>     support the
>     E notation for `xsd:decimal`.
>
>     I created a github repo at
>
>     http://github.com/pchampin/xsd_decimal

>
>     to document this issue and the current state of implementations, and
>     discuss whether updating the normative definition of `xsd:decimal`
>     would
>     be a good idea or not.
>
>     Please chime in if you are interested
>
>
> Since we are talking about XML Schema datatyping, it is important to 
> figure out what the XML community think about the idea.
You are absolutely correct, and I should not have overlooked that...
> Could you stick a blog post somewhere on w3.org <http://w3.org> with 
> an overview of what’s being suggested (beyond RDF details), then we 
> can bounce it around XML hangouts like XML-DEV.
I could do that. However, the pointers you gave below seem to indicate 
that the "no scientific notation" is well anchored in the XML community, 
so we might have our answer already...
>
> I assume you are thinking of updating the actual xsd:decimal spec 
> rather than just its RDF representation,
yes
> otherwise it would be confusing if the new variation leaked back into 
> XML settings where it isn’t considered valid.
>
> From a quick look around eg
> http://zvon.org/xxl/XMLSchemaTutorial/Output/ser_types_st7.html my 
> sense is that there is too much out there for this change to be feasible.
>
> “The element "A" represents a number (decimal), which must not be in 
> the scientific format (e.g. 1.5E+10 is forbidden - such numbers are 
> not processed by XSLT, for example).”
>
> |
> https://www.oreilly.com/library/view/xml-schema/0596002521/re62.html 
> teaches that |"1234.456E+2"| (scientific notation (|"E+2"|) is forbidden)
> |
>
> http://books.xmlschemata.org/relaxng/ch19-77057.html

>
> “The following values are invalid:1 234.456(spaces are 
> forbidden),1234.456E+2(scientific notation—E+2—is forbidden),+ 
> 1234.456(spaces are forbidden), or+1,234.456(delimiters between 
> thousands are forbidden).”
>
> This is without considering sites like Stack Overflow, or specific XML 
> languages like XAuery here:
> https://stackoverflow.com/questions/36337725/scientific-notation-to-whole-numbers

>
> Cheers,
>
> Dan
>
> p.s. also nearby,
> https://ontology2.com/the-book/decimal-the-missing-datatype.html


thanks for all these links.

   pa

>
>
>
>
>
>        pa
>
>     [1] https://github.com/w3c/dxwg/issues/1536

>     [2]
>     https://json-ld.org/playground/#startTab=tab-nquads&json-ld=%7B%22%40context%22%3A%7B%22ex%22%3A%22http%3A%2F%2Fexample.com%2Fns%2F%22%2C%22xsd%22%3A%22http%3A%2F%2Fwww.w3.org%2F2001%2FXMLSchema%23%22%2C%22ex%3Afoo%22%3A%7B%22%40type%22%3A%22xsd%3Adecimal%22%7D%7D%2C%22ex%3Afoo%22%3A%5B12.3%5D%7D

>     <https://json-ld.org/playground/#startTab=tab-nquads&json-ld=%7B%22%40context%22%3A%7B%22ex%22%3A%22http%3A%2F%2Fexample.com%2Fns%2F%22%2C%22xsd%22%3A%22http%3A%2F%2Fwww.w3.org%2F2001%2FXMLSchema%23%22%2C%22ex%3Afoo%22%3A%7B%22%40type%22%3A%22xsd%3Adecimal%22%7D%7D%2C%22ex%3Afoo%22%3A%5B12.3%5D%7D>
>     [3] https://en.wikipedia.org/wiki/Scientific_notation#E_notation

>