2PPy (tuProlog in Python)

Experimental porting of 2P-Kt on Python, via JPype.

This is a work in progress.

Introduction

Object-oriented and modular ecosystem for symbolic AI and logic programming, currently featuring:

• a command-line interface, in the form of a REPL

• a module for logic terms and clauses representation, namely tuprolog.core

• a module for logic unification, namely tuprolog.unify

• a module for in-memory indexing and storing logic theories, namely tuprolog.theory

• a module for other indexed collections of logic clauses (multi-sets and queues), namely tuprolog.collections

• a module providing generic API for resolution of logic queries, namely tuprolog.solve

  • a module that implements deterministic logic programming and solving, namely tuprolog.solve.prolog
  • a module that implements probabilistic logic programming and solving, namely tuprolog.solve.problog
  • a module exposing 2P-Kt's IO library (e.g. read/1, get_char/1, open/3), namely tuprolog.solve.libs

• two parsing modules: one aimed at parsing terms, namely tuprolog.core.parsing, and the other aimed at parsing theories, namely tuprolog.theory.parsing

How to use 2ppy

Installing 2ppy

Install 2PPy from PyPI (Python 3.10 or later):

pip install 2ppy        # or: uv add 2ppy

The package ships 2P-Kt's JARs, and depends on jdk4py, which bundles a Java runtime: no separate Java installation is required. The JVM is started when tuprolog is first imported.

Use 2ppy as a Python library

1. Import tuprolog.* modules in your Python scripts
2. Use the Pythonic API to write, parse or solve your logic programs. You can find some examples in the tests folder.
3. Python values are converted where terms are expected (e.g. struct("f", "a", 1) is f(a, 1), see tuprolog.core.to_term). Methods without a Pythonic counterpart are forwarded to the underlying JVM objects, so they may return JVM types (e.g. java.lang.String).
4. Python streams (io.IOBase, e.g. files, sys.stdin, io.StringIO, io.BytesIO) are accepted wherever 2P-Kt reads or writes: parsers (read_theory(open("kb.pl"))), solvers' channels (prolog_solver(std_in=io.StringIO("..."), std_out=buffer)), and JVM methods expecting streams (binary), readers or writers (text). Binary streams are UTF-8 encoded. By default, solvers read from and write on Python's current sys.stdin, sys.stdout and sys.stderr.

Use 2PPy as an interactive REPL

1. Python shell mode: run python -m tuprolog
2. Logic solver REPL:
   1. run python -m tuprolog.repl.prolog for Prolog
   2. run python -m tuprolog.repl.problog for ProbLog

How to contribute to 2PPy

The project follows template-python-project-uv. Requirements: uv and Apache Maven (mvn on the PATH), which downloads 2P-Kt's JARs from Maven Central, as declared in pom.xml.

uv sync                      # create the environment, with the development tools
uv run poe fetch-jars        # download the JARs into tuprolog/libs/ (needed before testing or building)
uv run poe test              # run the tests
uv run poe static-checks     # compileall, ruff format --check, ruff check, mypy (what CI runs)
uv run poe static-fixes      # ruff check --fix + ruff format
uv run poe coverage && uv run poe coverage-report
uv build                     # wheel and sdist in dist/ (JARs included)

Write the code inside tuprolog/ and the tests inside tests/. Contributor notes (layout, conventions) are in AGENTS.md. Optional local Git hooks: uv run poe install-hooks [HOOK...] (pre-commit, conventional-commits, no-ai-signature, pre-push).

Releases

Commits follow Conventional Commits: on every push to master, semantic-release derives the next version from them, publishes it on PyPI, and creates the corresponding GitHub Release. pyproject.toml is the only version source: never bump it by hand. The release workflow needs the PYPI_TOKEN and RELEASE_TOKEN repository secrets.

CI runs the static checks and the tests on Ubuntu, Windows and macOS, with Python 3.10 to 3.13. Dependencies are kept up to date by Renovate.