plur

plur is a fast, parallel, drop-in test runner and watcher primarily targeting Ruby and Rails using RSpec or Minitest. Its written in Go, so just install once and use across all your projects.

Quick Start

brew install rsanheim/tap/plur
cd my-rails-project
plur -n 4 --dry-run # preview what would run (no actual test execution)
plur -n 4           # run tests across four cores
plur                # run tests with auto-detected workers (cores - 2)
plur watch          # watch for changes and run tests automatically

Supported Platforms

• macOS ARM64 (Apple Silicon)
• Linux x86_64
• Linux ARM64
• Experimental Windows x86_64

Watch mode (plur watch) will install platform-specific binaries on first use.

Key Features

Parallel Test Execution

plur -n 4                    # Run with specific worker count
plur                          # Auto-detect workers (cores-2)
plur --dry-run               # Preview execution plan

Database Management

plur db:create -n 3          # Create test databases in parallel
plur db:migrate -n 3         # Run migrations across all test DBs
plur db:setup -n 3           # Full database setup

Explicit Framework Selection

For projects where you have both rspec and minitest tests, you can explicitly select the framework you want to use.

plur --use=rspec             # Run RSpec tests explicitly
plur --use=minitest          # Run Minitest tests

If there is just one framework, omit the --use flag and plur will auto-detect the framework.

Configuration

Plur supports TOML configuration files for persistent settings:

# .plur.toml or ~/.plur.toml
workers = 4

[job.rspec]
cmd = ["bin/rspec"]

[[watch]]
name = "lib-to-spec"
source = "lib/**/*.rb"
targets = ["spec/{{match}}_spec.rb"]
jobs = ["rspec"]

Config files load in this order (later files override earlier values):

1. ~/.plur.toml
2. .plur.toml
3. PLUR_CONFIG_FILE (if set)

See docs/examples/ directory for more configuration examples.

Environment Variables

• TEST_ENV_NUMBER: Worker 0 gets "", worker N gets "N+1"
• PARALLEL_TEST_GROUPS: Total number of workers
• PARALLEL_TEST_PROCESSORS: Compatible with parallel_tests

More information in the Documentation.