Skip to content

Installation

Install the package

pip install antlrope

conda install -c conda-forge antlrope

mamba install -c conda-forge antlrope

pixi workspace channel add conda-forge
pixi add antlrope

Antlrope depends on the official antlr4-python3-runtime, which your generated parser modules import, so it is installed automatically.

Antlrope ships as a pre-compiled binary wheel with the C++ engine built in, so nothing is compiled at install time. Each platform has a single CPython Stable ABI (abi3) wheel that covers CPython 3.12 and newer. Wheels are published for:

  • Linux (manylinux, x86-64 and aarch64)
  • macOS 11+ (Apple Silicon and Intel)
  • Windows (x86-64)

The minimum supported Python version is 3.12.

Install the ANTLR tool (to generate parsers)

To turn a .g4 grammar into the Python parser modules Antlrope drives, you also need the ANTLR tool itself, which is a Java program. The easiest way to get it is the antlr4-tools package. It provides the antlr4 command and downloads the ANTLR jar (and, on first use, a JDK) for you:

pip install antlr4-tools

conda install -c conda-forge antlr4-tools

mamba install -c conda-forge antlr4-tools

pixi workspace channel add conda-forge
pixi add antlr4-tools

You need the ANTLR tool only to generate or regenerate parsers, not to run them. If you already have Java and the ANTLR jar, you can use those instead; nothing here is specific to Antlrope.

See Getting started for a complete walkthrough, from generating a parser to writing and running a listener.

From source

Clone the repo from https://github.com/zuzukin/antlrope:

git clone https://github.com/zuzukin/antlrope.git

git clone git@github.com:zuzukin/antlrope.git

gh repo clone zuzukin/antlrope

The repository builds with pixi: pixi run build compiles the C++ extension and pixi run test runs the suite. The build needs a C++17 compiler, CMake 3.21 or later, and Ninja, all of which the pixi default environment provides (through its dev feature). See CONTRIBUTING.md for details.