Issue 11031045: Reject uses of lexical for-loop variable on the RHS.

LGTM (with nits). https://codereview.chromium.org/11031045/diff/1/src/parser.cc File src/parser.cc (right): https://codereview.chromium.org/11031045/diff/1/src/parser.cc#newcode2794 src/parser.cc:2794: top_scope_->NewUnresolved(factory(), name, interface); For consistency we ...

8 years, 2 months ago (2012-10-05 08:42:36 UTC) #2

rossberg

8 years, 2 months ago (2012-10-05 09:15:02 UTC) #3

https://codereview.chromium.org/11031045/diff/1/src/parser.cc
File src/parser.cc (right):

https://codereview.chromium.org/11031045/diff/1/src/parser.cc#newcode2794
src/parser.cc:2794: top_scope_->NewUnresolved(factory(), name, interface);
On 2012/10/05 08:42:36, Michael Starzinger wrote:
> For consistency we could also move this loop variable and interface down.

Done.

https://codereview.chromium.org/11031045/diff/1/src/parser.cc#newcode2841
src/parser.cc:2841: Interface* interface = Interface::NewValue();
On 2012/10/05 08:42:36, Michael Starzinger wrote:
> Can we also move the interface down right before the loop variable is created?

Inlined instead.

https://codereview.chromium.org/11031045/diff/1/test/mjsunit/regress/regress-...
File test/mjsunit/regress/regress-2322.js (right):

https://codereview.chromium.org/11031045/diff/1/test/mjsunit/regress/regress-...
test/mjsunit/regress/regress-2322.js:30: assertThrows("for (let x in x);",
SyntaxError);
On 2012/10/05 08:42:36, Michael Starzinger wrote:
> You probably want to add 'use strict' to this test case, to make sure that the
> original problem repros.

Indeed.

Expand Messages | Collapse Messages