miguelcast / php-parser

A PHP parser written in PHP

PHP Parser

This is a PHP 5.2 to PHP 7.0 parser written in PHP. Its purpose is to simplify static code analysis and manipulation.

Documentation for version 2.x (stable; for running on PHP >= 5.4; for parsing PHP 5.2 to PHP 7.0).

Documentation for version 1.x (unsupported; for running on PHP >= 5.3; for parsing PHP 5.2 to PHP 5.6).

In a Nutshell

The parser turns PHP source code into an abstract syntax tree. For example, if you pass the following code into the parser:

<?php
echo 'Hi', 'World';
hello\world('foo', 'bar' . 'baz');

You'll get a syntax tree looking roughly like this:

array(
    0: Stmt_Echo(
        exprs: array(
            0: Scalar_String(
                value: Hi
            )
            1: Scalar_String(
                value: World
            )
        )
    )
    1: Expr_FuncCall(
        name: Name(
            parts: array(
                0: hello
                1: world
            )
        )
        args: array(
            0: Arg(
                value: Scalar_String(
                    value: foo
                )
                byRef: false
            )
            1: Arg(
                value: Expr_Concat(
                    left: Scalar_String(
                        value: bar
                    )
                    right: Scalar_String(
                        value: baz
                    )
                )
                byRef: false
            )
        )
    )
)

You can then work with this syntax tree, for example to statically analyze the code (e.g. to find programming errors or security issues).

Additionally, you can convert a syntax tree back to PHP code. This allows you to do code preprocessing (like automatedly porting code to older PHP versions).

Installation

The preferred installation method is composer:

php composer.phar require nikic/php-parser

Documentation

1. Introduction
2. Usage of basic components
3. Other node tree representations
4. Code generation

Component documentation:

1. Error
2. Lexer