"requireSpaceAfterKeywords" should ignore object keys

[jscs-dev]

"requireSpaceAfterKeywords" should ignore object keys

[keithamus]

The problem here lies within the use of Esprima's "tokens" vs the full "body" tree object. For a simple test-case we can see:

test.js

x = { for: 'bar' };

The "for" part in the tokens array produces:

{
    "type": "Keyword",
    "value": "for",
    "range": [
        6,
        9
    ],
    "loc": {
        "start": {
            "line": 1,
            "column": 6
        },
        "end": {
            "line": 1,
            "column": 9
        }
    }
},

but in the "body" property, it produces this ouput:

"key": {
    "type": "Identifier",
    "name": "for",
    "range": [
        6,
        9
    ],
    "loc": {
        "start": {
            "line": 1,
            "column": 6
        },
        "end": {
            "line": 1,
            "column": 9
        }
    }
},

Because the tokens array (perhaps incorrectly?) puts the "type" as "Keyword", which is no different to a real for token, JSCS cannot tell it is a Keyword.

Possible Solutions

• Move JSCS to use the "body" tree object, so it can more accurately summise what each node type is
• Continue using the "tokens" array, but add a lookbehind in the rule, to check if the "Keyword" node is inside an object's "Punctuation" node
• Talk to Esprima devs or raise an issue about the "tokens" array labelling object properties as "Keyword"s instead of "Identifier"s