SpreadElement used in ArrayLiteral is lacking tests

[avdg]

I've noticed only 26 indirect test related to testing spread in ArrayLiterals, they are easy to pass if spread is followed only by an identifier. A assignmentExpression is allowed after the spread punctuator in current grammar.

[jugglinmike]

Thanks for the report! It's great to see parsers making use of Test262 and even better to see contributions coming back. (By the way, you may be interested in gh-196 and the proposed solution at gh-655.)

@leobalter and I recently began a coverage audit for ES2015 features, and "spread in array initializers" made our list, too.

I've submitted gh-672 to address this.

[avdg]

Thanks! And yeah, I had some doubts about these negative tests. Thanks for clarifying that error will be checked soon.

[jugglinmike]

Apologies: I'm a little confused about your comment. When you say, "some doubts
about these negative tests," are you referring to technically incorrect test
files (either in the master branch or in this pull request)? Or are you
referring to the difficulty of consuming tests that are expected to produce
runtime errors?

[avdg]

Uh, I rather thought about the error checking involved into these negative tests.

[avdg]

And sorry, this was a bit off topic

[jugglinmike]

No problem--I just wanted to be sure that we don't have any invalid tests!

[michaelficarra]

@avdg: You may also be interested in #559. I'd love to see UglifyJS start using those tests. I will get them published to https://github.com/tc39/test262-parser-tests ASAP.

[avdg]

I think we have enough error messages for now to work on, but yeah, these extra tests could be used as an extra verification tool.

[avdg]

@michaelficarra Are there es5.1 specific tests too?

[michaelficarra]

@avdg No. In my opinion, all of the tests should be kept up to date with the latest working draft. Anyway, the tests are now published at https://github.com/tc39/test262-parser-tests.

[avdg]

Should now be resolved in #672