RFR : 8013528 : (XS) Provide SharedSecrets access to String(char[], boolean) constructor

Mike Duigou mike.duigou at oracle.com
Tue Apr 30 23:31:29 UTC 2013


Hello all;

Since this code will be introduced without any usages I decided it was critical to make a stand alone unit test. I've updated the webrev:

http://cr.openjdk.java.net/~mduigou/JDK-8013528/1/webrev/

The webrev mistakes my hg copy for a rename... Ignore that. Capturing the provenance of the test file probably isn't critical since the file is repurposed for a different test, but I prefer to have the origin tracked rather than use a non-vcs copy.

I also made the other changes suggested by reviewers.

Thanks

Mike

On Apr 29 2013, at 21:30 , Mike Duigou wrote:

> Hello all;
> 
> This change originated as part of JDK-8006627 (which was also previously split into JDK-8007398 as well). It adds an internal mechanism for performance sensitive usages to create a string from a provided character array without copying that array. This saves both in the allocation (and subsequent GC) as well as the copying of the characters. There are a few places in the JDK that return Strings which can benefit from this change.
> 
> http://cr.openjdk.java.net/~mduigou/JDK-8013528/0/webrev/
> 
> Fear not, JDK-8006627 and JDK-8007398 will be revisited... For now it would be to get this change in to allow other potential users to move forward with their changes.
> 
> Mike




More information about the core-libs-dev mailing list