[XSL-LIST Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message] Re: XSLT repetition constructs
> On 8 Mar 2019, at 08:44, Dave Pawson dave.pawson@xxxxxxxxx <xsl-list-service@xxxxxxxxxxxxxxxxxxxxxx> wrote: > > On Fri, 8 Mar 2019 at 08:40, Michael Kay mike@xxxxxxxxxxxx > <xsl-list-service@xxxxxxxxxxxxxxxxxxxxxx> wrote: >> >> Because if $a is an empty sequence (which is the situation on the first call), then $a[last()] is an empty sequence, and ($x + ()) returns an empty sequence, not $x. > > I'm sure I'm not the only one thinking that sounds illogical Mike? (DC?) > I'd appreciated the 'logic' behind that decision (maths?) > The design intent was that () should have similar behaviour to null in SQL; in SQL (simplifying greatly), any operation that has null as an operand produces null as a result. The idea is that if you do a query //customer[location = 'UK'] then a customer whose location is absent (null) does not get selected; you don't have to treat an absent location as some kind of exception condition. Similarly //customer[abs(amount-outstanding) > 10000] selects nothing if amount-outstanding is absent/null, because abs(()) is (). There's one significant difference between the way SQL does it and the way XPath does it; SQL has full three-valued logic in which not(null) is null, whereas in XPath, not(()) is true. It's all about making it convenient to handle datasets in which some of the values are not always present. There is of course a logic behind this, but it's not the classical two-valued logic that they teach in elementary school. Michael Kay Saxonica
|
PURCHASE STYLUS STUDIO ONLINE TODAY!Purchasing Stylus Studio from our online shop is Easy, Secure and Value Priced! Download The World's Best XML IDE!Accelerate XML development with our award-winning XML IDE - Download a free trial today! Subscribe in XML format
|