Subject: RE: Problems with apersand (&) when trying to dynamically define a character value as an attribute value.
From: "Michael Kay" <mike@xxxxxxxxxxxx>
Date: Thu, 26 Oct 2006 13:46:36 +0100
|
> you switched base half way through that expression:-)
> (decimal)32 not (hex)20 sneaky though, and probably more
> efficient than my key suggestion?
Actually it's difficult to predict whether substring($s, $n, 1) will take
constant time or time proportional to string length. Likely to vary from one
implementation to another. The difficulty is surrogate pairs.
Michael Kay
http://www.saxonica.com/
|