Re: [Ejbca-develop] se.anatom.ejbca.util.Base64.decode strange bug

[Tomas G. <to...@pr...>]

Hi, this bug has now been fixed in upcoming ejbca 3.1 (cvs HEAD).

Cheers,
Tomas

Simon Lebettre wrote:

> Hello.
> 
> Using the last 3.1beta (cvs HEAD) on windows XP  :
> 
> If you run the following junit test, you will see that the method
> se.anatom.ejbca.util.Base64.decode is not able to decode this
> particular array (arrayIndexOufofBoundsException),
> 
> whereas org.bouncycastle.util.encoders.Base64.decode (on wich ejbca's
> base 64 is based) is able to do it.
> 
> The problem is not the size of this array (an array with the same size
> full of '56' is working ! )
> 
> the question is : is there still a good reason to fork from
> bouncycatle's base64 implementation ?
> 
> For your information, I detected this problem by exporting and
> re-importing my entire database to an xml file with dbunit
> (www.dbunit.org) :
> the dbunit import replaced every newline character in the table
> CADATA.DATA whith a space character, resulting in this 'special array'
> (by the way if I use the bouncycaste decoder my 'altered DB' works
> fine, but I'm investigating further to avoid any alteration at all)
>