Getting the untransformed version of a string about php-parser HOT 3 CLOSED

Yup, the parser currently tries to present the AST "as seen by PHP", so it also parses strings and numbers.

One way to retain the original value would be by defining a custom lexer which adds the originalValue as an attribute:

<?php

class KeepOriginalValueLexer extends PHPParser_Lexer // or PHPParser_Lexer_Emulative
{
    public function getNextToken(&$value = null, &$startAttributes = null, &$endAttributes = null) {
        $tokenId = parent::getNextToken($value, $startAttributes, $endAttributes);

        if ($tokenId == PHPParser_Parser::T_CONSTANT_ENCAPSED_STRING
            || $tokenId == PHPParser_Parser::T_LNUMBER
            || $tokenId == PHPParser_Parser::T_DNUMBER
        ) {
            // could also use $startAttributes, doesn't really matter here
            $endAttributes['originalValue'] = $value;
        }

        return $tokenId;
    }
}

This should make the original value available through $node->getAttribute('originalValue').

Would that solution work for you?

from php-parser.

It will certainly help. The only downside for me is that the regular expression is still used, costing performance

from php-parser.

I doubt that the escape sequence parsing really takes much time (at least compared to all the other stuff), but I think that it wouldn't be much of a problem to outsource the scalar value parsing into another class (which is injected into the parser). If you would prefer it that way I'll gladly implement it :)

from php-parser.