Martin, rather than catch the number format exception here, how about we
add the existing validate-pom target into the verify-project target called
before anything gets run, and if there are errors in the POM according to
the schema, the whole build stops.

This would get around this issue and other 'digester' warnings I've seen
this week.
--
dIon Gillard, Multitask Consulting
Work:      http://www.multitask.com.au
Developers: http://adslgateway.multitask.com.au/developers


                                                                                       
                            
                    Martin van den                                                     
                            
                    Bemt                 To:     Turbine Maven Developers List         
                            
                    <[EMAIL PROTECTED]        <[EMAIL PROTECTED]>       
                            
                    et>                  cc:                                           
                            
                                         Subject:     [PATCH] resubmit of 2 patches..  
                            
                    06/05/02 08:33                                                     
                            
                    AM                                                                 
                            
                    Please respond                                                     
                            
                    to "Turbine                                                        
                            
                    Maven                                                              
                            
                    Developers                                                         
                            
                    List"                                                              
                            
                                                                                       
                            
                                                                                       
                            




Hi,

I redid the patch that handles an invalid pomversion better (catches the
numberformatexception thrown and shows a "nice" message saying what was
wrong).

Also the patch which allows to no fail on the testscases, but just
continue building the system. The -1 that was there, was changed to -0.

They are both included in one diff.

Mvgr,
Martin van den Bemt



(See attached file: various.diff)--
To unsubscribe, e-mail:   <
mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <
mailto:[EMAIL PROTECTED]>

Attachment: various.diff
Description: Binary data

--
To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>

Reply via email to