saspytest

pytest support library for SAS testing with saspy.

saspytest provides utility functions and pytest-style assertions for testing SAS programs through the saspy Python interface. It is inspired by SASUnit assertion macros.

Features

• Session management – Shared or isolated SAS sessions with automatic setup and teardown
• Log assertions – Validate SAS logs for errors, warnings, or custom patterns
• Dataset assertions – Check dataset existence, record counts, columns, and equality
• Macro assertions – Verify SAS macro variable values and existence
• File utilities – Upload/download files and submit local .sas scripts

Installation

pip install saspytest

With uv, use uv add saspytest in a uv-managed project, or uv pip install saspytest in an existing environment.

Prerequisites

• Python 3.11+
• A working saspy configuration that can connect to your SAS deployment
• Java (required by saspy only for IOM connections)
• pandas 3.0+ (required by saspytest's get_dataset_as_df() helper)

Note: saspytest does not bundle proprietary SAS JARs. You must provide your own SAS deployment and configure saspy accordingly. sascfg_personal.py is saspy's configuration file; saspytest_config.py is saspytest's separate auto-discovered project configuration convention. See Getting Started.

Getting Started

The normal saspytest workflow is to write pytest tests that execute SAS code against a live SAS session. See the Getting Started guide.

Run a user test project with:

pytest -v

The repository's mock-based unit tests are internal contributor tests. They do not require SAS and are documented in Contributing.

Quickstart

When saspytest is installed, pytest auto-loads the package's fixture plugin, so the sas_session fixture is available in your tests without importing it. The example requires a live SAS connection. If a project also contains SAS-free Python tests, use the optional live-SAS test marker.

from saspytest import (
    assert_no_errors,
    assert_dataset_exists,
    assert_record_count,
)


def test_sas_program(sas_session):
    result = sas_session.submit("""
        data work.SASPYTEST_EXAMPLE;
            do i = 1 to 5;
                output;
            end;
        run;
    """)

    assert_no_errors(result["LOG"])
    assert_dataset_exists(sas_session, "SASPYTEST_EXAMPLE")
    assert_record_count(sas_session, "SASPYTEST_EXAMPLE", 5)

Documentation

• Getting Started – Configure SAS execution environment and write your first live-SAS test
• User Guide – How to use saspytest: session management, assertions, file utilities, and examples
• API Reference – Complete function and fixture reference
• Support Policy – Supported Python, saspy, and SAS configurations
• Development Setup – Getting started as a contributor
• Development Guidelines – Code style and conventions
• AI Skills – Agent skills available in this repository (e.g. sas-tester)

License

MIT License. See LICENSE for details.