Refresh string literal representation

[microsoft]

Refresh string literal representation

[roblourens]

Honestly I'm not sure what else we had in mind. Would have to look back at old notes.

Have you considered also making escape characters like \n its own token appearing as a children node in a string?

Probably would not parse it at that granularity unless there was a good reason for it.

[PythooonUser]

Hey,

here in #11 and in #168 you talk about that strings and template strings are still not implemented perfectly. Can you iterate on what exactly is still missing or not working?

I had a look at the parse output of e.g.

<?php "Hello $world";

Which gives

{
    "SourceFileNode": {
        "statementList": [
            [...]
            {
                "ExpressionStatement": {
                    "expression": {
                        "StringLiteral": {
                            "startQuote": {
                                "kind": "DoubleQuoteToken",
                                "fullStart": 6,
                                "start": 6,
                                "length": 1
                            },
                            "children": [
                                {
                                    "kind": "EncapsedAndWhitespace",
                                    "fullStart": 7,
                                    "start": 7,
                                    "length": 6
                                },
                                {
                                    "Variable": {
                                        "dollar": null,
                                        "name": {
                                            "kind": "VariableName",
                                            "fullStart": 13,
                                            "start": 13,
                                            "length": 6
                                        }
                                    }
                                }
                            ],
                            "endQuote": {
                                "kind": "DoubleQuoteToken",
                                "fullStart": 19,
                                "start": 19,
                                "length": 1
                            }
                        }
                    },
                    "semicolon": {
                        "kind": "SemicolonToken",
                        "fullStart": 20,
                        "start": 20,
                        "length": 1
                    }
                }
            }
        ]
        [...]
    }
}

And it returns an instance of StringLiteral which is essentially a non-delimited list of whitespace/characters and expression nodes. That sounds good to me and seems to be flexible enough.

Have you considered also making escape characters like \n its own token appearing as a children node in a string? Could that be helpful? I know that this distinction is made in language grammars (syntax highlighting) at least.

Cheers

[PythooonUser]

Yes, I see. The current implementation seems to be in a good state :)

[roblourens]

Thanks for taking a look though. I will close this unless there is a reason to reopen it.