Error fixed in esql 1.22 reintroduced in 1.24?

[Marco Pauck <pa...@wmd.de>]

While browsing the CVS I noticed that the newest version 1.24
of the esql logic sheet reversed a patch that was made in 1.22
to fix a problem resulting in duplicate variable declarations
when using multiple logicsheets with class-level logic.

Here is the change introduced with 1.24 (see line 132/130):
http://xml.apache.org/websrc/cvsweb.cgi/xml-cocoon/src/org/apache/cocoon/processor/xsp/library/sql/esql.xsl.diff?r1=1.23&r2=1.24

And here's the patch that fixed the problem in 1.22:
http://xml.apache.org/websrc/cvsweb.cgi/xml-cocoon/src/org/apache/cocoon/processor/xsp/library/sql/esql.xsl.diff?r1=1.21&r2=1.22

Is this an oversight or is the problem now fixed in another way?

	Marco

Re: Error fixed in esql 1.22 reintroduced in 1.24?

[Donald Ball <ba...@webslingerZ.com>]

On Tue, 7 Nov 2000, Marco Pauck wrote:

> While browsing the CVS I noticed that the newest version 1.24
> of the esql logic sheet reversed a patch that was made in 1.22
> to fix a problem resulting in duplicate variable declarations
> when using multiple logicsheets with class-level logic.
> 
> Here is the change introduced with 1.24 (see line 132/130):
> http://xml.apache.org/websrc/cvsweb.cgi/xml-cocoon/src/org/apache/cocoon/processor/xsp/library/sql/esql.xsl.diff?r1=1.23&r2=1.24
> 
> And here's the patch that fixed the problem in 1.22:
> http://xml.apache.org/websrc/cvsweb.cgi/xml-cocoon/src/org/apache/cocoon/processor/xsp/library/sql/esql.xsl.diff?r1=1.21&r2=1.22
> 
> Is this an oversight or is the problem now fixed in another way?

complete oversight, thanks. must've reverted the repository when patching
something. it's patched again now. by the way, jeremy - in the ftp
logicsheet, you fix this problem like so:

<xsl:template match="xsp:page/*[not(starts-with(name(.), 'xsp:'))]">

i think that's incorrect, actually. i think you really want to be doing
this:

<xsl:template 
match="xsp:page/*[not(namespace-uri(.)='http://www.apache.org/1999/XSP/Core')]">

e.g. matching the namespace URI, not the prefix chosen by the page
author. make sense? or does XSLT actually work such that your expression
will match elements in the xsp namespace, since the namespace is
associated with the prefix xsp in the stylesheet, even though the page
author might have associated it with another prefix?

this is why namespaces are loosely tied to prefixes, right? ;)

- donald