Re: [xslt] Problem with variables

[Igor Zlatkovic <igor stud fh-frankfurt de>]

Here is what java 1.4.0 on Windows gives for the same number and the same
format strings:

  C:\Home\Igor\tmp>java NrTest
  %25
  %25.5
  %25.480
  %25.48

  C:\Home\Igor\tmp>

Well, given that the XSLT spec says that the java behaviour is
authoritative, I would say that libxslt is conformant, MSXSL and Saxon are
not.

Oh, and thanks to you and Charlie for the info about '%'!

Ciao
Igor

> According to the XSLT spec, the format-number function behaves as in Java-
> http://java.sun.com/products/jdk/1.1/docs/api/java.text.DecimalFormat.html
>
> The per-cent (%) character multiplies the number by 100 then formats it,
> The per-mille (?) character multiplies the number by 1000 then formats it.
>
> Though it is odd you saying that msxml gives those results.
> I'm testing libxslt using a testsuite comparing results from our own old
> XSLT processor, Saxon and XT (These tests were built up a couple of years
> ago, so Saxon and XT might have changed since). All 3 processors returned
> the same results you found with MSXML.
>
> In fact, I've just tested Saxon 6.5.2, and that returns
>
>   %0
>   %.3
>   %.255
>   %.2548
>
> Is there someone with Java handy who could verify which results are
correct?
> I strongly suspect libxslt is and everyone else is wrong...
>
> I'll get me a latest copy from CVS to see if it fixes my variable problem.
>
> Thanks,
> Richard
>
> ----- Original Message -----
> From: "Igor Zlatkovic" <igor@stud.fh-frankfurt.de>
> To: <xslt@gnome.org>
> Sent: Tuesday, May 21, 2002 2:05 PM
> Subject: Re: [xslt] Problem with variables
>
>
> > Hello,
> >
> > strange, when I run it, I get
> >
> >   C:\Home\Igor\tmp>xsltproc b.xsl a.xml
> >
> >     %25
> >     %25.5
> >     %25.480
> >     %25.48
> >   C:\Home\Igor\tmp>
> >
> > But, I am using the current CVS version, not the 2.4.21 release. Perhaps
> the
> > problem you are having is gone in the meantime.
> >
> > I would ask a question: I am puzzled about the meaning of the '%'
> character
> > in the format string. My book says that per-cent and per-mille
characters
> > are allowed, but does not say what they are for, neither it provides an
> > example. To make things more puzzling, MSXSL gives a different result:
> >
> >   C:\Home\Igor\tmp>msxsl -u 4.0 a.xml b.xsl
> >
> >     %
> >     %.3
> >     %.255
> >     %.2548
> >   C:\Home\Igor\tmp>
> >
> > Now, can someone tell me what the per-cent and per-mille characters in
the
> > format string mean? Which processor from the two above produced the
> correct
> > result?
> >
> > Ciao
> > Igor
> >
> >
> > > Can someone suggest whats going wrong here? I've had a look through
the
> > > libxslt code, but can't find anything obvious.
> > >
> > > (Using Igor's Win32 binaries libxml2.2-4-21, libxslt1.0.17 on Win2000)
> > >
> > > a.xml:
> > > <?xml version="1.0"?>
> > > <AAA author="nicmila@idoox.com">
> > >      <BBB>bbb</BBB>
> > >      <CCC>ccc</CCC>
> > > </AAA>
> > >
> > > b.xsl:
> > > <?xml version="1.0"?>
> > > <xsl:stylesheet xmlns:xsl='http://www.w3.org/1999/XSL/Transform'
> > > version="1.0">
> > > <xsl:output method="text"/>
> > > <xsl:variable name="G">0.2548</xsl:variable>
> > > <xsl:template match="/">
> > >   <xsl:text>
> > >   </xsl:text>
> > >   <xsl:value-of select="format-number($G,'%')"/>
> > >   <xsl:text>
> > >   </xsl:text>
> > >   <xsl:value-of select="format-number($G,'%#.#')"/>
> > >   <xsl:text>
> > >   </xsl:text>
> > >   <xsl:value-of select="format-number($G,'%#.000')"/>
> > >   <xsl:text>
> > >   </xsl:text>
> > >   <xsl:value-of select="format-number($G,'%#.#####')"/>
> > > </xsl:template>
> > > </xsl:stylesheet>
> > >
> > > C:\libxml\libxslt-1.0.17.win32\util>xsltproc b.xsl a.xml
> > > xsltproc c.xsl a.xml
> > > runtime error: file c.xsl line 4 element variable
> > > Global variable G already defined
> > > %25%25.5%25.480%25.48
> > >
> > > There appears to be two problems here:
> > > 1) The runtime error stating that variable "G" is already defined.
> > > 2) I'd expect the formatted numbers to each appear on new lines (but
> this
> > > could quite easily go down to a gap in my understanding of
<xsl:text/>)
> > >
> > > Any ideas?