Subject: Re: Identity transformation for XSLT 2.0
From: Colin Paul Adams <colin@xxxxxxxxxxxxxxxxxx>
Date: Mon, 17 Dec 2007 07:22:09 +0000
|
Also, you make use of E000/1/2. As you point out, the chance of these
being used is small, but not non-zero.
I would suggest using Private-use characters from plane 17 instead
(and somewhere in the middle of the plane, not at the start of the
block). This would reduce the chance of a collision further, I think.
Of course, it would still not be zero. In particular, trying to run
the transformation against itself is guaranteed to find a collision :-)
--
Colin Adams
Preston Lancashire
| Current Thread |
|
Colin Paul Adams - 17 Dec 2007 07:22:43 -0000 <=
Michael Kay - 17 Dec 2007 09:55:30 -0000
|
|