py-pglite

What it is and what it does

py-pglite is a Python wrapper around PGlite that spins up an in-memory PostgreSQL instance for testing, eliminating the need for Docker containers, server configuration, or test database management. It works with any PostgreSQL client library—SQLAlchemy, Django ORM, psycopg, asyncpg—and provides pytest fixtures that give each test a fresh, isolated database. The core appeal is speed and simplicity: startup in seconds rather than minutes, and no external dependencies beyond psutil.

The package supports both Unix domain sockets (default, fastest) and TCP sockets (required for asyncpg and cloud environments). It also exposes PGlite extensions like pgvector for vector search testing. Configuration is minimal—most tests work with a single fixture—but advanced users can customize timeouts, logging, cleanup behavior, and socket modes via PGliteConfig.

Use it for:

• Unit testing SQLAlchemy or Django models without spinning up a real PostgreSQL server or Docker container.
• Testing FastAPI or other web frameworks that depend on a database, with automatic per-test isolation.
• Validating complex SQL queries, window functions, and JSON operations in CI/CD pipelines where Docker is unavailable or slow.
• Testing vector similarity search and RAG applications using pgvector extension in isolated test environments.
• Rapid local development and debugging of database-heavy code without managing a persistent PostgreSQL instance.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

py-pglite provides a zero-config PostgreSQL testing environment for Python by wrapping PGlite, letting you run real PostgreSQL queries in tests without Docker or server setup.

Yes, if you test Python code that uses PostgreSQL and want to eliminate Docker or server setup overhead. The low install friction, permissive license, and broad client library support make it a straightforward drop-in for pytest suites. Caveat: the package is in beta and aging (no release in 331 days), so verify that PGlite's feature set covers your specific PostgreSQL extensions and that performance meets your test suite's needs before committing to production CI/CD.

Install

py-pglite on PyPI

pip install py-pglite

uv add py-pglite

poetry add py-pglite

Installing py-pglite

Before you install

Low friction: pure Python wheel with only psutil as a runtime dependency. Maintenance status is aging—last release was 331 days ago, though the repo remains active with recent commits and moderate community interest (632 stars).

License in practice

Apache-2.0 permissive license allows use in commercial and open-source projects with minimal restrictions; you must include a copy of the license and note any modifications.

Quickstart

pip install py-pglite

from py_pglite import PGliteManager

with PGliteManager() as db:
    dsn = db.get_dsn()
    # Use dsn with any PostgreSQL client (SQLAlchemy, psycopg, etc.)

Requires Python 3.10 or later; asyncpg users must enable TCP mode via PGliteConfig(use_tcp=True).

Verify before relying

• Whether PGlite's feature set (JSON, arrays, window functions) covers your specific PostgreSQL extensions or advanced SQL features.
• Performance characteristics and memory footprint when running large test suites or complex queries.
• Compatibility with all Django ORM patterns and edge cases beyond the documented socket and backend patterns.

Package facts

Evidence: py_pglite-0.5.3-py3-none-any.whl

Keywords: testing, postgresql, pglite, database, pytest, sqlalchemy, sqlmodel, django, pytest-django

Tags

More Python Modules packages