gistfile1.php

<?php
namespace FHP\Read;

class InvalidTokenException extends \Exception { }

class _EndOfTextException extends \Exception { }

function _chars($text) {
    while ($text !== false) {
        yield $text[0];
        $text = substr($text, 1);
    }

    throw new _EndOfTextException();
}

function _is_whitespace($c) {
    return strpos(" \n,", $c) !== false;
}

function _is_identifier_start($c) {
    return ($c >= 'a' && $c <= 'z')
        || ($c >= 'A' && $c <= 'Z')
        || strpos("+-*/&|^!<>=?", $c) !== false;
}

function _is_identifier_rest($c) {
    return _is_identifier_start($c)
        || ($c >= '0' && $c <= '9');
}

function read($text) {
    $tokens = lex($text . "\n");
    foreach ($tokens as $token) {
        var_dump($token);
    }
}

function lex($text) {
    try {
        $chars = _chars($text);

    loop:
        while (_is_whitespace($chars->current())) {
            $chars->next();
        }

        if (_is_identifier_start($chars->current())) {
            $identifier = '';
            do {
                $identifier .= $chars->current();
                $chars->next();
            } while (_is_identifier_rest($chars->current()));
            yield (object)['token' => 'identifier', 'value' => $identifier];
            goto loop;
        }

        if ($chars->current() === '"') {
            $string = '';
            for (;;) {
                $chars->next();
                if ($chars->current() === '"') {
                    break;
                } else {
                    $string .= $chars->current();
                }
            }
            yield (object)['token' => 'string', 'value' => $string];
            goto next;
        }

        switch ($chars->current()) {
            case '(': yield (object)['token' => 'lparen']; goto next;
            case ')': yield (object)['token' => 'rparen']; goto next;
        }

        throw new InvalidTokenException();

    next:
        $chars->next();
        goto loop;
    } catch (_EndOfTextException $e) {
        yield (object)['token' => 'eof'];
    }
}

read('(echo "Hello, world!")');