Re: numberformat07.xsl failure on gump

[Myriam_Midy]

I believe that that's what we do too.

Myriam


                                                                                       
  
                    Morten Jorgensen                                                   
  
                    <morten.jorgense       To:     [EMAIL PROTECTED]            
  
                    [EMAIL PROTECTED]>             cc:     (bcc: Myriam Midy/CAM/Lotus)        
  
                    Sent by:               Subject:     Re: numberformat07.xsl failure 
  
                    Morten.Jorgensen        on gump                                    
  
                    @sun.com                                                           
  
                                                                                       
  
                                                                                       
  
                    06/26/01 11:52                                                     
  
                    AM                                                                 
  
                    Please respond                                                     
  
                    to xalan-dev                                                       
  
                                                                                       
  
                                                                                       
  




Scott,

I don't know how the format-number() function is implemented in
Xalan, but I know that we used a standard Java 1.1 class to do
the work for us on XSLTC. We did this because the rules for
number formatting in the XSLT spec and the Java 1.1 API spec are
very close. But, this means that the actual output may differ if
the DecimalFormat class in your JDK does not conform to the Java
standards.

Morten


[EMAIL PROTECTED] wrote:
>
> (see http://nagoya.apache.org/~rubys/gump/xml-xalan2-smoketest.html)
>
> numberformat07.xsl succeeds fine on my machine.  I can't tell why it
didn't
> succeed on the gump machine.  Is this a precision problem?  Should we
> exclude this test from the smoketest?
>
> mdocs16 is a new test and is missing a referenced file.
>
> -scott