Re: RFR 8027640 : String.indexOf(String, int) for the empty string case not specified

2014-02-25 Thread Paul Sandoz 

On Feb 25, 2014, at 12:52 AM, Brent Christian brent.christ...@oracle.com 
wrote:

 Please review my changes for:
 https://bugs.openjdk.java.net/browse/JDK-8027640
 
 The webrev is here:
 http://cr.openjdk.java.net/~bchristi/8027640/webrev.00/
 

In AbstractStringBuilder there is a small typo, k:

1348  * pThe returned index is the largest value {@code k} for which:

Otherwise looks ok.

Paul.

 
 This is a small cleanup to reconcile the String[last]indexOf() spec with 
 longstanding behavior.  There are no changes to executable code.
 
 In String, the Math.min(fromIndex, this.length()) clause is brought over 
 from String[Buffer|Builder] for indexOf(String,int) and 
 lastIndexOf(String,int).  This resolves the issue for which the bug was filed.
 
 To further sync up the [last]indexOf() docs, I ported the 6940381 wording 
 improvements from String to String[Buffer|Builder].  I also made a few 
 doclint updates to the code in question.
 
 Thanks,
 -Brent

Re: RFR 8027640 : String.indexOf(String, int) for the empty string case not specified

2014-02-25 Thread Brent Christian 

On 2/25/14 4:41 AM, Paul Sandoz wrote:

On Feb 25, 2014, at 12:52 AM, Brent Christian brent.christ...@oracle.com 
wrote:


Please review my changes for:
https://bugs.openjdk.java.net/browse/JDK-8027640

The webrev is here:
http://cr.openjdk.java.net/~bchristi/8027640/webrev.00/



In AbstractStringBuilder there is a small typo, k:

1348  * pThe returned index is the largest value {@code k} for which:



Great catch, Paul.  Thanks!

-Brent

RFR 8027640 : String.indexOf(String,int) for the empty string case not specified

2014-02-24 Thread Brent Christian 

Please review my changes for:
https://bugs.openjdk.java.net/browse/JDK-8027640

The webrev is here:
http://cr.openjdk.java.net/~bchristi/8027640/webrev.00/


This is a small cleanup to reconcile the String[last]indexOf() spec with 
longstanding behavior.  There are no changes to executable code.


In String, the Math.min(fromIndex, this.length()) clause is brought 
over from String[Buffer|Builder] for indexOf(String,int) and 
lastIndexOf(String,int).  This resolves the issue for which the bug was 
filed.


To further sync up the [last]indexOf() docs, I ported the 6940381 
wording improvements from String to String[Buffer|Builder].  I also made 
a few doclint updates to the code in question.


Thanks,
-Brent