fix #839 Iterable Java Object

[tuchida]

Closes #839
Implemented Symbol.iterator in java.lang.Iterable and java.util.Map.
java.util.Iterator is equivalent to the Iterator object in ECMAScript because it has state, so we did not implement Symbol.iterator.
java.util.Map can iterate over keys and values just like ECMAScript's Map. Nashorn works the same way.

[gbrail]

It seems like this is useful. Is it possible to update the commit message or this PR with a short sample of what is possible now that wasn't possible before so that we can put something in the release notes?

Also, how do we ensure that some existing code won't break now that iteration is easier on many types of Java objects?

[tuchida]

It seems like this is useful. Is it possible to update the commit message or this PR with a short sample of what is possible now that wasn't possible before so that we can put something in the release notes?

Can I write the release notes here?
https://github.com/mozilla/rhino/blob/master/RELEASE-NOTES.md

Also, how do we ensure that some existing code won't break now that iteration is easier on many types of Java objects?

Since Symbol.iterator is a Symbol, it does not override Java methods.
Up until now, it was not iterable, which caused a TypeError.

var l = new java.util.ArrayList([1, 2, 3]);
for (var v of l) print(v);
// js: uncaught JavaScript runtime exception: TypeError: [1.0, 2, 3] is not iterable

I don't think there are any side effects from this becoming iterable.

[gbrail]

Thanks -- you have given me enough to put in the release notes when it's time. I might ask for help with the PR at that time if it's not obvious. It's cool that we're starting to use lambdas to cut down on code clutter too!

[tonygermano]

@p-bakker Can this get added to the 1.7.14 milestone and properly linked to 839?

[p-bakker]

I've linked the PR to the issue. As the issue is already linked to the milestone, i haven't linked the PR, because otherwise both the issue and the PR show up in the milestone