Allow duplicate property names in classes

Allow duplicate property names in classes

ES6 no longer makes duplicate properties an error. However, we
continue to treat duplicate properties in strict mode object
literals as errors. With this change we allow duplicate properties
in class bodies. We continue to flag duplicate constructors as an
error as required by ES6.

BUG=v8:3570
LOG=Y
R=marja@chromium.org

Committed: https://code.google.com/p/v8/source/detail?r=24933

[arv (Not doing code reviews), 2014-10-27 15:42:07]

arv@chromium.org changed reviewers:
+ dslomov@chromium.org, marja@chromium.org

[arv (Not doing code reviews), 2014-10-27 15:42:07]

PTAL

Another option would be to make the checker handle this but since the checker is
going away I didn't want to put the logic in there.

[marja, 2014-10-28 08:30:22]

lgtm %

https://codereview.chromium.org/677953004/diff/1/src/preparser.h
File src/preparser.h (right):

https://codereview.chromium.org/677953004/diff/1/src/preparser.h#newcode1973
src/preparser.h:1973: if (*has_seen_constructor) {
It's not trivial to see that has_seen_constructor cannot be NULL here... so the
caller must always pass non-NULL if in_class and !is_static, but otherwise it's
ok to pass NULL? This is quite subtle. CHECK that this is followed? Document it
better? Or make this if (has_seen_constructor && *has_seen_constructor)? Always
pass a dummy non-NULL ptr? Idk what would be the best solution. Wdyt?

https://codereview.chromium.org/677953004/diff/1/test/cctest/test-parsing.cc
File test/cctest/test-parsing.cc (right):

https://codereview.chromium.org/677953004/diff/1/test/cctest/test-parsing.cc#...
test/cctest/test-parsing.cc:4042: // We currently do not allow any duplicate
properties in class bodies. This
This comment seems out of date now.

[arv (Not doing code reviews), 2014-10-28 09:25:55]

Thanks

https://codereview.chromium.org/677953004/diff/1/src/preparser.h
File src/preparser.h (right):

https://codereview.chromium.org/677953004/diff/1/src/preparser.h#newcode1973
src/preparser.h:1973: if (*has_seen_constructor) {
On 2014/10/28 08:30:22, marja wrote:
> It's not trivial to see that has_seen_constructor cannot be NULL here... so
the
> caller must always pass non-NULL if in_class and !is_static, but otherwise
it's
> ok to pass NULL? This is quite subtle. CHECK that this is followed? Document
it
> better? Or make this if (has_seen_constructor && *has_seen_constructor)?
Always
> pass a dummy non-NULL ptr? Idk what would be the best solution. Wdyt?

I agree that this is ugly. The least I can do is to add an assert.

[arv (Not doing code reviews), 2014-10-28 11:53:38]

Remove out dated comment and add/fix assert

[arv (Not doing code reviews), 2014-10-28 12:23:35]

Committed patchset #2 (id:20001) manually as 24933 (presubmit successful).