[XSL-LIST Mailing List Archive Home]
[By Thread]
[By Date]
Christopher R. Maden wrote:
so the issue is what is more effective and robust. First expression to be evaluated needs 2 string2numbe
conversions (may be optimized to only 1) and 1 number comparison, while the second one needs 1 string2number conversion, 1 number2string and 1 string comparison, so it seems to me that hack expression is a little more effective.
The only hope is XPath 2, which must support lexical representation of NaN.
XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list
Re: [xsl] variable going out of scope in a for-each loop?
|
Subject: Re: [xsl] variable going out of scope in a for-each loop? From: Oleg Tkachenko <olegt@xxxxxxxxxxxxx> Date: Mon, 10 Dec 2001 12:04:40 +0200 |
Christopher R. Maden wrote:
No, he didn't. This is a clever little bit of hackery. NaN is not equal to itself. So if the node value is a number, number(.)=number(.) is a tautology, but if the node value *isn't* a number, then NaN=NaN is false.Well, number() = number() is an analog of more readable expression string(number()) = 'NaN',
so the issue is what is more effective and robust. First expression to be evaluated needs 2 string2numbe
conversions (may be optimized to only 1) and 1 number comparison, while the second one needs 1 string2number conversion, 1 number2string and 1 string comparison, so it seems to me that hack expression is a little more effective.
The only hope is XPath 2, which must support lexical representation of NaN.
-- Oleg Tkachenko Multiconn International, Israel
XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list
| Current Thread |
|---|
|
| <- Previous | Index | Next -> |
|---|---|---|
| Re: [xsl] variable going out of sco, Mike Brown | Thread | RE: [xsl] variable going out of sco, Michael Kay |
| [xsl] Distinct problem, "Kreyßel, Conny" | Date | Re: [xsl] whitespaces, Oleg Tkachenko |
| Month |
Keywords