RE: Question regarding the typehirarchy

Such questions are best asked on the www-ql alias.

 

XQuery is mainly a strongly typed language which means that your
expression below leads to a type error. 

 

Exceptions are the treatment of untyped data, the notion of atomization,
and the implicit promotion of numeric data (and xs:anyURI to be added).

 

These rules are described in the corresponding sections in the language
document or the formal semantics (look for atomization, numeric type
promotion, xdt:untypedAtomic).

 

HTH
Michael

 

________________________________

From: public-qt-comments-request@w3.org
[mailto:public-qt-comments-request@w3.org] On Behalf Of Kirmse, Daniel
Sent: Tuesday, March 23, 2004 12:13 AM
To: 'public-qt-comments@w3.org'
Subject: Question regarding the typehirarchy

 

Hi,

don't know whether this is the right place for the question. If not,
please take my apologies.

I wonder about the XQuery type hierarchy. There are a lot of documents
around but I did not really find some kind of implicit conversion rule
for types. Suppose the expression:

"20" > 1000

A string literal is compared with an integer literal. In "XQuery 1.0 and
XPath 2.0 Formal Semantics" section [B.2 Mapping of Overloaded Internal
Functions]. A list of internal functions is provided. According to it
there are comparison functions that take either numerics or strings as
parameters. But what is to take in the provided example. For the
practical use the numeric variant seems to be the best. But the question
is: Is there an explicit rule that describes when to convert which type
to what other type and when not? Is there a document that contains the
ruling? Which section?

Cheers & Thanks

Daniel Kirmse