Mailing-List: contact issues-help@commons.apache.org; run by ezmlm
Precedence: bulk
Reply-To: issues@commons.apache.org
Date: Tue, 20 Mar 2012 11:45:40 +0000 (UTC)
From: "Sebb (Commented) (JIRA)" <jira@apache.org>
To: issues@commons.apache.org
Message-ID: 
 <995053134.36117.1332243940311.JavaMail.tomcat@hel.zones.apache.org>
Subject: [jira] [Commented] (CODEC-96) Base64 encode() method is no longer
 thread-safe, breaking clients using it as a shared BinaryEncoder
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/CODEC-96?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13233362#comment-13233362 ] 

Sebb commented on CODEC-96:
---------------------------

I think it might be possible to move the mutable fields out of the lower level classes and back into the streaming classes which are the ones that actually need them.

Whether it is possible to do this in a compatible way, I don't yet know, but I'd like to give it a try over then next week or so.
                
> Base64 encode() method is no longer thread-safe, breaking clients using it as a shared BinaryEncoder
> ----------------------------------------------------------------------------------------------------
>
>                 Key: CODEC-96
>                 URL: https://issues.apache.org/jira/browse/CODEC-96
>             Project: Commons Codec
>          Issue Type: Bug
>    Affects Versions: 1.4
>            Reporter: Matt Ryall
>            Assignee: Julius Davies
>             Fix For: 1.x
>
>         Attachments: codec-96-3rd-attempt.patch
>
>
> Streaming support was added to Base64 in commons-codec 1.4 with CODEC-69. This introduced instance variables to Base64 which means the class can no longer be used as a shared BinaryEncoder instance.
> For example, BinaryEncoder has an interface which could be (and was) used like this with Base64:
> {code:java}
> class Example {
>     private BinaryEncoder encoder = new Base64();
>     byte[] someMethod(byte[] data) {
>         try {
>             return encoder.encode(data);
>         }
>         catch (EncoderException e) {
>             throw new RuntimeException(e);
>         }
>     } 
> }
> {code}
> Base64 is no longer thread-safe in commons-codec 1.4, so code like the above which is accessed by multiple threads can throw NullPointerException:
> {noformat}
> java.lang.NullPointerException
> 	at org.apache.commons.codec.binary.Base64.encode(Base64.java:469)
> 	at org.apache.commons.codec.binary.Base64.encode(Base64.java:937)
> 	at ... (application code)
> {noformat}
> Looking at the implementation of Base64, I think making it thread-safe for this kind of usage would be quite tricky. I haven't attempted to prepare a patch.
> I would be happy if it was indicated in the Javadoc that Base64 is not thread-safe and should not be shared. However, some other users of commons-codec might be more worried about this regression.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira