test262/test/language/statements/let/syntax/let-let-declaration-split-across-two-lines.js
Mike Pennisi ade6d2e384 Remove "NotEarlyError" object
Because expectations regarding error "phase" are now expressed via test
meta-data, the test runner may now enforce this requirement on negative
tests.

Remove the "NotEarlyError" from the project source. This reduces the
amount of domain knowledge required to author tests and lessens the
potential for inconsistencies between tests.
2016-10-19 15:24:22 -04:00

32 lines
1.3 KiB
JavaScript

// Copyright (C) 2015 Mozilla Corporation. All rights reserved.
// This code is governed by the BSD license found in the LICENSE file.
/*---
author: Jeff Walden
es6id: 13.3.1.1
description: >
let: |let let| split across two lines is not subject to automatic semicolon insertion.
info: >
|let| followed by a name is a lexical declaration. This is so even if the
name is on a new line. ASI applies *only* if an offending token not allowed
by the grammar is encountered, and there's no [no LineTerminator here]
restriction in LexicalDeclaration or ForDeclaration forbidding a line break.
It's a tricky point, but this is true *even if* the name is "let", a name that
can't be bound by LexicalDeclaration or ForDeclaration. Per 5.3, static
semantics early errors are validated *after* determining productions matching
the source text.
So in this testcase, the eval text matches LexicalDeclaration. No ASI occurs,
because "let\nlet = ..." matches LexicalDeclaration before static semantics
are considered. *Then* 13.3.1.1's static semantics for the LexicalDeclaration
just chosen, per 5.3, are validated to recognize the Script as invalid. Thus
the eval script can't be evaluated, and a SyntaxError is thrown.
negative:
phase: early
type: SyntaxError
---*/
let // start of a LexicalDeclaration, *not* an ASI opportunity
let;