tree-sitter-php

0.24.1 · active · verified Thu Apr 09

tree-sitter-php provides the PHP grammar for the tree-sitter parsing system. It enables incremental parsing to build concrete syntax trees for PHP source code, facilitating advanced code analysis, linting, and editor features. The library is actively maintained, with version 0.24.1 being the latest, and receives frequent updates to support new PHP language features.

Warnings

gotcha Tree-sitter grammars are C libraries and typically require compilation. While `tree-sitter-php` provides pre-compiled binary wheels for common platforms, environments without a suitable C compiler or for non-standard architectures might encounter installation failures, requiring manual compilation.
Fix: Ensure a C compiler (e.g., GCC or Clang) is installed and available in your PATH. For specific issues, consult the `py-tree-sitter` documentation or GitHub issues. Consider using `tree-sitter-languages` for a bundled solution if facing repeated compilation issues.
breaking The PHP grammar is continuously updated to support new language features (e.g., PHP 8.1+, 8.4, 8.5). Older versions of the `tree-sitter-php` grammar might fail to correctly parse newer PHP syntax, leading to incomplete or incorrect syntax trees.
Fix: Regularly update `tree-sitter-php` to the latest version to ensure compatibility with modern PHP syntax. Check release notes for specific PHP version support.
gotcha Parsing `.php` files containing a mix of PHP and HTML can sometimes lead to issues or unexpected parsing behavior, especially when integrating with editor plugins that use multiple tree-sitter parsers (e.g., HTML and PHP). Historically, this led to a split into `php` and `php_only` parsers.
Fix: Be aware of the distinction between the full PHP grammar (which handles HTML) and `php_only` variants if your toolchain offers them. Ensure proper language injection strategies in editors or tools. Consult `nvim-treesitter` or similar editor integration documentation for specific configurations related to mixed PHP/HTML files.

Install

pip install tree-sitter-php Install with pip

Imports

language
```
import tree_sitter_php
```
The 'language' function is typically accessed via the top-level package import to get the Language object.
Language, Parser
```
from tree_sitter import Language, Parser
```
These are the core classes from the main tree-sitter Python bindings used to load and utilize the grammar.

Quickstart

This quickstart demonstrates how to load the PHP grammar, initialize a parser, and parse a sample PHP code snippet. It then shows how to access the root node of the generated syntax tree and perform a simple query to extract class and method names.

from tree_sitter import Language, Parser
import tree_sitter_php

# Load the PHP grammar
PHP_LANGUAGE = Language(tree_sitter_php.language())
parser = Parser(PHP_LANGUAGE)

php_code = b'''
<?php

class MyClass {
    public function __construct(private string $name) {}

    public function greet(): string {
        return "Hello, " . $this->name . "!";
    }
}

$obj = new MyClass("World");
echo $obj->greet();

?>
'''

# Parse the code
tree = parser.parse(php_code)

# Get the root node of the syntax tree
root_node = tree.root_node

print(f"Root node type: {root_node.type}")
print(f"Number of children: {len(root_node.children)}")
# Example: Find a class_declaration node
for child in root_node.children:
    if child.type == 'declaration_list': # often wrappers around class/function
        for decl in child.children:
            if decl.type == 'class_declaration':
                print(f"Found class: {decl.child_by_field_name('name').text.decode('utf8')}")
                break

# You can also use queries for more specific pattern matching
query = PHP_LANGUAGE.query("""
(class_declaration name: (name) @class-name)
(method_declaration name: (name) @method-name)
""")

captures = query.captures(root_node)
for node, name in captures:
    print(f"Captured: {name.replace('-', '_')}: {node.text.decode('utf8')}")

view raw JSON →