Re: TMQL, next round

Hi Robert,

I have also an e-mail with a list of issues in the queue, but it will
need one more day.

[...]
>    TMQL Issue: predefined data types
[...]
>    Structured Discussion

>       ? should all types from XSD be understood?

I don't see a reason to support them all. Additionally, it is
difficult to have "native" support for the binary ones. Further, some
values, i.e. 1, may be xs:byte, xs:short, xs:int or xs:long or
xs:integer or its supertype xs:decimal. How to distinguish a xs:gYear
from an integer etc.

We should collect some of the primary datatypes and all other
datatypes can be represented as (quoted) lexical representation and
the datatype URI (or QName).

IMO the following datatypes should make it into TMQL:
- One of the floating point numbers
- one integer datatype
- xs:string
- xs:anyURI
- xs:boolean
- xs:date
- xs:dateTime ???
- xs:time ???

>         + very rich collection, strong big-vendor support
>         + constants can be written 'naturally' (3 instead of "3"^xsd:integer)
>       => also all XPath 2.0 functions and operators have to be supported

I don't understand why support of XPath 2.0 functions/ops is a logical
consequence.

Best regards,
Lars
-- 
http://www.semagia.com