On 2014/07/03 at 09:07:11, rossberg wrote:
On 2014/07/03 07:57:14, marja wrote:
> This CL makes V8 crash if the arrow function allowing flag is passed,
since
the
> actual implementation is missing (this will return NULL expressions,
and if
the
> upper layer tries to use them, it'll crash). Is this okay, or will we be
burned
> by this because of some fuzzing tests?
The fuzzer does not turn on the harmony flag, only the es-staging flag
(harmony containing half-done stuff is one of the reasons why we introduced
the
latter).
That said, the parser should still spit out something well-formed, and not
make later phases crash. So that you can e.g. try it in d8, and write JS
test
cases for parsing (which would be nice to have, btw). For example, for now
just
create a dummy FunctionLiteral for all arrow functions.
Sure, I am adding a dummy FunctionLiteral here so d8 won't crash. That will
also
allow me to uncomment some test cases which rely on having a non-NULL
expression
to be able to call increase_parenthesization_level() on them.
https://codereview.chromium.org/160073006/
--
--
v8-dev mailing list
[email protected]
http://groups.google.com/group/v8-dev
---
You received this message because you are subscribed to the Google Groups "v8-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
