[
Lists Home |
Date Index |
Thread Index
]
> In XML there are three (not two) syntactic devices that
> people have at various times asserted have some connection to
> the absence of a tuple in the relational model.
>
> XML has:
> 1. Omission of an optional element.
> 2. Presence of an element having the attribute xsi:nil
> 3. Presence of an element having empty content.
>
> SQL does make a distinction between omission and NULL in some
> contexts (e.g. an UPDATE statement) but not in others (e.g. a rowset).
I don't think you mean "the absence of a tuple". The relational model
has the vast edifice of "null" to support absent data in a cell of a
table, but it has never had any way of representing a missing or unknown
row.
An element with empty content is surely the parallel of a cell
containing a zero-length string in SQL. Users may use this with similar
semantic intent to omission of the element, but it's hard to see it as
an analog of SQL's NULL.
For my part, I have always thought that xsi:nil is an abomination. If I
want to represent uncertainty in my XML data, I want to choose myself
how to model it, for example
<age estimate="30" plus-or-minus="10" confidence="low"
source="hearsay"/>
The whole point of XML is that it doesn't define the semantics of the
tags: whoever invented xsi:nil seems not to have realized this.
Michael Kay
|
