RFR [8054221] StringJoiner imlementation optimization

[Martin Buchholz]

Thanks, Ivan.


On Mon, Aug 4, 2014 at 9:16 AM, Ivan Gerasimov <ivan.gerasimov at oracle.com>
wrote:

> Hello!
>
> Here's the proposal by Martin Buchholz to improve implementation of
> StringJoiner class:
> http://cr.openjdk.java.net/~martin/webrevs/openjdk9/
> StringJoiner-optimization/
>
> I think that result of concatenation in the merge() method can be saved to
> be reused later.
> This would be useful, for example, if a StringJoiner is merged into other
> StringJoiners several times.
>
>
Looks good - saving the result of compaction is a good idea.

---

I'm truly unsure whether it's worth optimizing for size < 2.  But if we do
that, I think we should optimize size == 0 as well, thus:

if (addLen == 0 && size <= 1)
 return (size == 0) ? "" : elts[0];

---

 240         if (other == this) {


I doubt that optimizing specially for a self-merge is worthwhile.  When
you're thinking of doing that, you probably really want
Strings.repeat(String, int copies)
https://code.google.com/p/guava-libraries/source/browse/guava/src/com/google/common/base/Strings.java?r=5cc3b0b33392b02ff95a4ce5f291c439b5b1406e#154

I suggest adding repeat to String!

So, here's my slightly modified version of the Martin's webrev:
> http://cr.openjdk.java.net/~igerasim/8054221/0/webrev/
>
> I've also modified the test to regularly test cases with empty
> prefixes/suffixes.
>
> Sincerely yours,
> Ivan
>