Bats Testing Patterns

Overview

Bats (Bash Automated Testing System) provides a TAP-compliant testing framework for shell scripts. This skill documents proven patterns for writing effective, maintainable shell script tests that catch bugs early and document expected behavior.

Use this skill when:

• Writing tests for Bash or shell scripts
• Creating test fixtures and mock data for shell testing
• Setting up test infrastructure for shell-based tools
• Debugging failing shell tests
• Integrating shell tests into CI/CD pipelines

Core Testing Patterns

Basic Test Structure

Every Bats test file is a shell script with a .bats extension:

#!/usr/bin/env bats

@test "Test description goes here" {
    # Test code
    [ condition ]
}

Key Points:

• Use descriptive test names that explain what is being verified
• Each @test block is an independent test
• Tests should be focused on one specific behavior
• Use the shebang #!/usr/bin/env bats at the top

Exit Code Assertions

Test command success and failure explicitly:

#!/usr/bin/env bats

@test "Command succeeds as expected" {
    run echo "hello"
    [ "$status" -eq 0 ]
}

@test "Command fails as expected" {
    run false
    [ "$status" -ne 0 ]
}

@test "Command returns specific exit code" {
    run bash -c "exit 127"
    [ "$status" -eq 127 ]
}

@test "Can capture command result" {
    run echo "hello"
    [ $status -eq 0 ]
    [ "$output" = "hello" ]
}

Best Practice: Always use run to capture command output and exit status. The run command sets $status, $output, and $lines variables for assertions.

Output Assertions

Verify command output matches expectations:

#!/usr/bin/env bats

@test "Output matches exact string" {
    result=$(echo "hello world")
    [ "$result" = "hello world" ]
}

@test "Output contains substring" {
    result=$(echo "hello world")
    [[ "$result" == *"world"* ]]
}

@test "Output matches regex pattern" {
    result=$(date +%Y)
    [[ "$result" =~ ^[0-9]{4}$ ]]
}

@test "Multi-line output comparison" {
    run printf "line1\nline2\nline3"
    [ "$output" = "line1
line2
line3" ]
}

@test "Using lines array for output" {
    run printf "line1\nline2\nline3"
    [ "${lines[0]}" = "line1" ]
    [ "${lines[1]}" = "line2" ]
    [ "${lines[2]}" = "line3" ]
    [ "${#lines[@]}" -eq 3 ]
}

Tip: Use the $lines array when testing multi-line output - it's cleaner than string comparison.

File Assertions

Test file operations and attributes:

#!/usr/bin/env bats

setup() {
    TEST_DIR=$(mktemp -d)
    export TEST_DIR
}

teardown() {
    rm -rf "$TEST_DIR"
}

@test "File is created successfully" {
    [ ! -f "$TEST_DIR/output.txt" ]
    echo "content" > "$TEST_DIR/output.txt"
    [ -f "$TEST_DIR/output.txt" ]
}

@test "File contents match expected" {
    echo "expected content" > "$TEST_DIR/output.txt"
    [ "$(cat "$TEST_DIR/output.txt")" = "expected content" ]
}

@test "File is readable" {
    touch "$TEST_DIR/test.txt"
    [ -r "$TEST_DIR/test.txt" ]
}

@test "File has correct permissions (Linux)" {
    touch "$TEST_DIR/test.txt"
    chmod 644 "$TEST_DIR/test.txt"
    [ "$(stat -c %a "$TEST_DIR/test.txt")" = "644" ]
}

@test "File has correct permissions (macOS)" {
    touch "$TEST_DIR/test.txt"
    chmod 644 "$TEST_DIR/test.txt"
    [ "$(stat -f %OLp "$TEST_DIR/test.txt")" = "644" ]
}

@test "File size is correct" {
    echo -n "12345" > "$TEST_DIR/test.txt"
    [ "$(wc -c < "$TEST_DIR/test.txt")" -eq 5 ]
}

@test "Directory structure is created" {
    mkdir -p "$TEST_DIR/sub/nested/deep"
    [ -d "$TEST_DIR/sub/nested/deep" ]
}

Platform Note: File permission checking differs between Linux (stat -c) and macOS (stat -f). Test on your target platform or provide compatibility helpers.

Setup and Teardown Patterns

Basic Setup and Teardown

Execute code before and after each test:

#!/usr/bin/env bats

setup() {
    # Runs before EACH test
    TEST_DIR=$(mktemp -d)
    export TEST_DIR

    # Source the script under test
    source "${BATS_TEST_DIRNAME}/../bin/script.sh"
}

teardown() {
    # Runs after EACH test
    rm -rf "$TEST_DIR"
}

@test "Test using TEST_DIR" {
    touch "$TEST_DIR/file.txt"
    [ -f "$TEST_DIR/file.txt" ]
}

@test "Second test has clean TEST_DIR" {
    # TEST_DIR is recreated fresh for each test
    [ ! -f "$TEST_DIR/file.txt" ]
}

Critical: The setup() and teardown() functions run before and after EACH test, ensuring test isolation.

Setup with Test Resources

Create fixtures and test data:

#!/usr/bin/env bats

setup() {
    # Create directory structure
    TEST_DIR=$(mktemp -d)
    mkdir -p "$TEST_DIR/data/input"
    mkdir -p "$TEST_DIR/data/output"

    # Create test fixtures
    echo "line1" > "$TEST_DIR/data/input/file1.txt"
    echo "line2" > "$TEST_DIR/data/input/file2.txt"
    echo "line3" > "$TEST_DIR/data/input/file3.txt"

    # Initialize environment variables
    export DATA_DIR="$TEST_DIR/data"
    export INPUT_DIR="$DATA_DIR/input"
    export OUTPUT_DIR="$DATA_DIR/output"
    
    # Source the script being tested
    source "${BATS_TEST_DIRNAME}/../scripts/process_files.sh"
}

teardown() {
    rm -rf "$TEST_DIR"
}

@test "Processes all input files" {
    process_files "$INPUT_DIR" "$OUTPUT_DIR"
    [ -f "$OUTPUT_DIR/file1.txt" ]
    [ -f "$OUTPUT_DIR/file2.txt" ]
    [ -f "$OUTPUT_DIR/file3.txt" ]
}

@test "Handles empty input directory" {
    rm -rf "$INPUT_DIR"/*
    process_files "$INPUT_DIR" "$OUTPUT_DIR"
    [ "$(ls -A "$OUTPUT_DIR")" = "" ]
}

Global Setup/Teardown

Run expensive setup once for all tests:

#!/usr/bin/env bats

# Load shared test utilities
load test_helper

# setup_file runs ONCE before all tests in the file
setup_file() {
    export SHARED_RESOURCE=$(mktemp -d)
    export SHARED_DB="$SHARED_RESOURCE/test.db"
    
    # Expensive operation: initialize database
    echo "Creating test database..."
    sqlite3 "$SHARED_DB" < "${BATS_TEST_DIRNAME}/fixtures/schema.sql"
}

# teardown_file runs ONCE after all tests in the file
teardown_file() {
    rm -rf "$SHARED_RESOURCE"
}

# setup runs before each test (optional)
setup() {
    # Per-test setup if needed
    export TEST_ID=$(date +%s%N)
}

@test "First test uses shared resource" {
    [ -f "$SHARED_DB" ]
    sqlite3 "$SHARED_DB" "SELECT COUNT(*) FROM users;"
}

@test "Second test uses same shared resource" {
    [ -f "$SHARED_DB" ]
    # Database persists between tests
    sqlite3 "$SHARED_DB" "INSERT INTO users (name) VALUES ('test_$TEST_ID');"
}

Use Case: Global setup/teardown is perfect for expensive operations like database initialization, server startup, or large file downloads that can be shared across tests.

Mocking and Stubbing Patterns

Function Mocking

Override functions for testing:

#!/usr/bin/env bats

# Mock external command
curl() {
    echo '{"status": "success", "data": "mocked"}'
    return 0
}

@test "Function uses mocked curl" {
    export -f curl
    
    # Source script that calls curl
    source "${BATS_TEST_DIRNAME}/../scripts/api_client.sh"
    
    result=$(fetch_data "https://api.example.com/data")
    [[ "$result" == *"mocked"* ]]
}

@test "Mock can simulate failure" {
    curl() {
        echo "Connection refused"
        return 1
    }
    export -f curl
    
    source "${BATS_TEST_DIRNAME}/../scripts/api_client.sh"
    run fetch_data "https://api.example.com/data"
    [ "$status" -ne 0 ]
}

Command Stubbing with PATH Manipulation

Create stub commands that override system commands:

#!/usr/bin/env bats

setup() {
    # Create stub directory
    STUBS_DIR="$BATS_TEST_TMPDIR/stubs"
    mkdir -p "$STUBS_DIR"

    # Prepend to PATH so stubs are found first
    export PATH="$STUBS_DIR:$PATH"
}

teardown() {
    rm -rf "$STUBS_DIR"
}

create_stub() {
    local cmd="$1"
    local output="$2"
    local exit_code="${3:-0}"

    cat > "$STUBS_DIR/$cmd" <<EOF
#!/bin/bash
echo "$output"
exit $exit_code
EOF
    chmod +x "$STUBS_DIR/$cmd"
}

@test "Function works with stubbed curl" {
    create_stub curl '{"status": "ok"}' 0
    
    source "${BATS_TEST_DIRNAME}/../scripts/api_client.sh"
    run fetch_api_status
    [ "$status" -eq 0 ]
    [[ "$output" == *"ok"* ]]
}

@test "Function handles stubbed curl failure" {
    create_stub curl "curl: (6) Could not resolve host" 6
    
    source "${BATS_TEST_DIRNAME}/../scripts/api_client.sh"
    run fetch_api_status
    [ "$status" -ne 0 ]
}

@test "Can stub multiple commands" {
    create_stub git "commit abc123" 0
    create_stub docker "Container started" 0
    
    # Test code that uses both git and docker
    run git status
    [[ "$output" == *"abc123"* ]]
    
    run docker ps
    [[ "$output" == *"started"* ]]
}

Powerful Pattern: PATH manipulation allows stubbing any command without modifying the code under test.

Environment Variable Stubbing

Test different configurations:

#!/usr/bin/env bats

@test "Function uses environment override" {
    export LOG_LEVEL="DEBUG"
    export API_ENDPOINT="https://staging.example.com"
    
    source "${BATS_TEST_DIRNAME}/../scripts/config.sh"
    run get_config_value "log_level"
    [[ "$output" == *"DEBUG"* ]]
}

@test "Function uses defaults when vars unset" {
    unset LOG_LEVEL
    unset API_ENDPOINT
    
    source "${BATS_TEST_DIRNAME}/../scripts/config.sh"
    run get_config_value "log_level"
    [[ "$output" == *"INFO"* ]]  # Default value
}

@test "Function handles missing config file" {
    export CONFIG_FILE="/nonexistent/config.yaml"
    
    source "${BATS_TEST_DIRNAME}/../scripts/config.sh"
    run load_config
    [ "$status" -ne 0 ]
    [[ "$output" == *"not found"* ]]
}

Fixture Management

Using Fixture Files

Organize test data in a dedicated directory:

#!/usr/bin/env bats

# Directory structure:
# tests/
# ├── fixtures/
# │   ├── input.json
# │   ├── expected_output.json
# │   └── schema.sql
# └── script.bats

setup() {
    FIXTURES_DIR="${BATS_TEST_DIRNAME}/fixtures"
    WORK_DIR=$(mktemp -d)
    export WORK_DIR
}

teardown() {
    rm -rf "$WORK_DIR"
}

@test "Process fixture file produces expected output" {
    # Copy fixture to working directory
    cp "$FIXTURES_DIR/input.json" "$WORK_DIR/input.json"

    # Run processing function
    source "${BATS_TEST_DIRNAME}/../scripts/processor.sh"
    run process_json "$WORK_DIR/input.json" "$WORK_DIR/output.json"
    [ "$status" -eq 0 ]

    # Compare output with expected fixture
    diff "$WORK_DIR/output.json" "$FIXTURES_DIR/expected_output.json"
}

@test "Handles malformed input file" {
    echo "invalid json" > "$WORK_DIR/bad.json"
    
    source "${BATS_TEST_DIRNAME}/../scripts/processor.sh"
    run process_json "$WORK_DIR/bad.json" "$WORK_DIR/output.json"
    [ "$status" -ne 0 ]
    [[ "$output" == *"invalid"* ]]
}

Organization Tip: Keep fixtures in a fixtures/ directory alongside your test files for easy maintenance.

Dynamic Fixture Generation

Generate test data programmatically:

#!/usr/bin/env bats

generate_csv_fixture() {
    local rows="$1"
    local file="$2"

    echo "id,name,email" > "$file"
    for i in $(seq 1 "$rows"); do
        echo "$i,User$i,user$i@example.com" >> "$file"
    done
}

generate_log_fixture() {
    local lines="$1"
    local file="$2"

    for i in $(seq 1 "$lines"); do
        echo "[$(date -Iseconds)] INFO: Log entry $i" >> "$file"
    done
}

@test "Handles large CSV file" {
    generate_csv_fixture 10000 "$BATS_TEST_TMPDIR/large.csv"
    
    source "${BATS_TEST_DIRNAME}/../scripts/csv_parser.sh"
    run parse_csv "$BATS_TEST_TMPDIR/large.csv"
    [ "$status" -eq 0 ]
    [ "$(wc -l < "$BATS_TEST_TMPDIR/large.csv")" -eq 10001 ]  # Header + 10000 rows
}

@test "Handles log file with 1000 entries" {
    generate_log_fixture 1000 "$BATS_TEST_TMPDIR/app.log"
    
    source "${BATS_TEST_DIRNAME}/../scripts/log_analyzer.sh"
    run analyze_logs "$BATS_TEST_TMPDIR/app.log"
    [ "$status" -eq 0 ]
}

Benefit: Dynamic fixtures make tests more flexible and can test edge cases like performance with large datasets.

Advanced Testing Patterns

Testing Error Conditions

Ensure proper error handling:

#!/usr/bin/env bats

@test "Fails gracefully with missing required file" {
    source "${BATS_TEST_DIRNAME}/../scripts/processor.sh"
    run process_file "/nonexistent/file.txt"
    [ "$status" -ne 0 ]
    [[ "$output" == *"not found"* || "$output" == *"No such file"* ]]
}

@test "Fails with helpful message for invalid input" {
    source "${BATS_TEST_DIRNAME}/../scripts/processor.sh"
    run process_file ""
    [ "$status" -ne 0 ]
    [[ "$output" == *"Usage:"* || "$output" == *"required"* ]]
}

@test "Handles permission denied gracefully" {
    touch "$BATS_TEST_TMPDIR/readonly.txt"
    chmod 000 "$BATS_TEST_TMPDIR/readonly.txt"
    
    source "${BATS_TEST_DIRNAME}/../scripts/processor.sh"
    run process_file "$BATS_TEST_TMPDIR/readonly.txt"
    [ "$status" -ne 0 ]
    [[ "$output" == *"Permission denied"* || "$output" == *"cannot read"* ]]
    
    # Cleanup: restore permissions
    chmod 644 "$BATS_TEST_TMPDIR/readonly.txt"
}

@test "Provides usage help with invalid option" {
    source "${BATS_TEST_DIRNAME}/../scripts/processor.sh"
    run process_file --invalid-option
    [ "$status" -ne 0 ]
    [[ "$output" == *"Usage:"* ]]
}

@test "Validates input format" {
    echo "not a number" > "$BATS_TEST_TMPDIR/invalid.txt"
    
    source "${BATS_TEST_DIRNAME}/../scripts/numeric_processor.sh"
    run process_numbers "$BATS_TEST_TMPDIR/invalid.txt"
    [ "$status" -ne 0 ]
    [[ "$output" == *"invalid"* || "$output" == *"number"* ]]
}

Best Practice: Test error paths thoroughly - they're often overlooked but critical for user experience.

Testing with External Dependencies

Handle optional dependencies gracefully:

#!/usr/bin/env bats

setup() {
    # Check for required tools
    if ! command -v jq &>/dev/null; then
        skip "jq is not installed - required for JSON tests"
    fi

    source "${BATS_TEST_DIRNAME}/../scripts/json_processor.sh"
}

@test "JSON parsing works with jq" {
    run parse_json '{"name": "test", "value": 42}'
    [ "$status" -eq 0 ]
    [[ "$output" == *"test"* ]]
}

@test "Can process complex nested JSON" {
    skip_if_missing jq
    
    json='{"users": [{"name": "Alice", "age": 30}, {"name": "Bob", "age": 25}]}'
    run extract_user_names "$json"
    [ "$status" -eq 0 ]
    [[ "$output" == *"Alice"* ]]
    [[ "$output" == *"Bob"* ]]
}

skip_if_missing() {
    local tool="$1"
    if ! command -v "$tool" &>/dev/null; then
        skip "$tool is not installed"
    fi
}

Graceful Degradation: Use skip to mark tests that can't run in the current environment instead of failing them.

Testing Shell Compatibility

Ensure scripts work across different shells:

#!/usr/bin/env bats

@test "Script works in bash" {
    bash "${BATS_TEST_DIRNAME}/../scripts/portable.sh" --version
}

@test "Script works in sh (POSIX mode)" {
    sh "${BATS_TEST_DIRNAME}/../scripts/portable.sh" --version
}

@test "Script works in dash" {
    if ! command -v dash &>/dev/null; then
        skip "dash not installed"
    fi
    dash "${BATS_TEST_DIRNAME}/../scripts/portable.sh" --version
}

@test "Script uses POSIX-compliant syntax" {
    # Check for bash-specific features
    ! grep -q 'function ' "${BATS_TEST_DIRNAME}/../scripts/portable.sh"
    ! grep -q '\[\[' "${BATS_TEST_DIRNAME}/../scripts/portable.sh"
    ! grep -q '=~' "${BATS_TEST_DIRNAME}/../scripts/portable.sh"
}

Portability Tip: If your script should work on different systems, test with multiple shells to catch compatibility issues.

Parallel Test Execution

Test concurrent operations:

#!/usr/bin/env bats

@test "Multiple operations can run concurrently" {
    source "${BATS_TEST_DIRNAME}/../scripts/parallel_processor.sh"
    
    # Start multiple background processes
    for i in {1..5}; do
        process_item "$i" "$BATS_TEST_TMPDIR/output_$i.txt" &
    done
    
    # Wait for all to complete
    wait
    
    # Verify all outputs
    for i in {1..5}; do
        [ -f "$BATS_TEST_TMPDIR/output_$i.txt" ]
    done
}

@test "Concurrent file operations don't conflict" {
    source "${BATS_TEST_DIRNAME}/../scripts/file_handler.sh"
    
    # Create multiple files concurrently
    for i in {1..10}; do
        (
            echo "Content $i" > "$BATS_TEST_TMPDIR/file_$i.txt"
        ) &
    done
    wait
    
    # Verify no data corruption
    [ "$(ls "$BATS_TEST_TMPDIR"/file_*.txt | wc -l)" -eq 10 ]
}

Test Organization with Helpers

test_helper.sh Pattern

Create reusable test utilities:

#!/usr/bin/env bash
# File: tests/test_helper.bash

# Source script under test
export SCRIPT_DIR="${BATS_TEST_DIRNAME%/*}/scripts"
export BIN_DIR="${BATS_TEST_DIRNAME%/*}/bin"

# Common assertion helpers
assert_file_exists() {
    local file="$1"
    if [ ! -f "$file" ]; then
        echo "Expected file to exist: $file" >&2
        return 1
    fi
}

assert_file_contains() {
    local file="$1"
    local pattern="$2"
    
    if [ ! -f "$file" ]; then
        echo "File does not exist: $file" >&2
        return 1
    fi

    if ! grep -q "$pattern" "$file"; then
        echo "File does not contain pattern: $pattern" >&2
        echo "File contents:" >&2
        cat "$file" >&2
        return 1
    fi
}

assert_output_contains() {
    local pattern="$1"
    if [[ ! "$output" == *"$pattern"* ]]; then
        echo "Output does not contain: $pattern" >&2
        echo "Actual output:" >&2
        echo "$output" >&2
        return 1
    fi
}

# Setup helpers
create_test_workspace() {
    export TEST_WORKSPACE=$(mktemp -d)
    mkdir -p "$TEST_WORKSPACE/input"
    mkdir -p "$TEST_WORKSPACE/output"
    mkdir -p "$TEST_WORKSPACE/temp"
}

cleanup_test_workspace() {
    if [ -n "$TEST_WORKSPACE" ] && [ -d "$TEST_WORKSPACE" ]; then
        rm -rf "$TEST_WORKSPACE"
    fi
}

# Stub helpers
create_command_stub() {
    local cmd="$1"
    local output="$2"
    local exit_code="${3:-0}"
    local stub_dir="${STUBS_DIR:-$BATS_TEST_TMPDIR/stubs}"
    
    mkdir -p "$stub_dir"
    
    cat > "$stub_dir/$cmd" <<EOF
#!/bin/bash
echo "$output"
exit $exit_code
EOF
    chmod +x "$stub_dir/$cmd"
    export PATH="$stub_dir:$PATH"
}

Usage in tests:

#!/usr/bin/env bats

load test_helper

setup() {
    create_test_workspace
}

teardown() {
    cleanup_test_workspace
}

@test "Uses test helper for assertions" {
    echo "content" > "$TEST_WORKSPACE/output/result.txt"
    
    assert_file_exists "$TEST_WORKSPACE/output/result.txt"
    assert_file_contains "$TEST_WORKSPACE/output/result.txt" "content"
}

@test "Uses helper to create stubs" {
    create_command_stub "git" "commit abc123" 0
    
    run git status
    assert_output_contains "abc123"
}

Benefits:

• Reduces code duplication across test files
• Provides consistent error messages
• Makes tests more readable and maintainable

CI/CD Integration

GitHub Actions Workflow

name: Shell Script Tests

on:
  push:
    branches: [ main, develop ]
  pull_request:
    branches: [ main ]

jobs:
  test:
    runs-on: ubuntu-latest
    
    strategy:
      matrix:
        shell: [bash, dash, sh]

    steps:
      - name: Checkout repository
        uses: actions/checkout@v4

      - name: Install Bats
        run: |
          npm install --global bats

      - name: Run tests with ${{ matrix.shell }}
        run: |
          export TEST_SHELL=${{ matrix.shell }}
          bats tests/*.bats

      - name: Run tests with TAP output
        if: always()
        run: |
          bats tests/*.bats --tap | tee test_output.tap

      - name: Upload test results
        if: always()
        uses: actions/upload-artifact@v4
        with:
          name: test-results-${{ matrix.shell }}
          path: test_output.tap

Advanced workflow with parallel execution:

name: Comprehensive Shell Tests

on: [push, pull_request]

jobs:
  test:
    runs-on: ${{ matrix.os }}
    
    strategy:
      matrix:
        os: [ubuntu-latest, macos-latest]
        bats-version: ['1.11.0']
      fail-fast: false

    steps:
      - uses: actions/checkout@v4

      - name: Setup Bats
        uses: bats-core/bats-action@2.0.0

      - name: Run unit tests
        run: bats tests/unit/*.bats --timing

      - name: Run integration tests
        run: bats tests/integration/*.bats --timing

      - name: Run tests in parallel
        run: bats tests/*.bats --jobs 4 --timing

      - name: Generate coverage report
        if: matrix.os == 'ubuntu-latest'
        run: |
          # Optional: Use kcov or similar for coverage
          echo "Coverage reporting setup here"

Makefile Integration

.PHONY: test test-verbose test-tap test-unit test-integration test-parallel clean

# Default test target
test:
    @echo "Running all tests..."
    bats tests/*.bats

# Verbose output for debugging
test-verbose:
    @echo "Running tests with verbose output..."
    bats tests/*.bats --verbose

# TAP format output
test-tap:
    @echo "Running tests with TAP output..."
    bats tests/*.bats --tap

# Run only unit tests
test-unit:
    @echo "Running unit tests..."
    bats tests/unit/*.bats

# Run only integration tests
test-integration:
    @echo "Running integration tests..."
    bats tests/integration/*.bats

# Run tests in parallel for speed
test-parallel:
    @echo "Running tests in parallel..."
    bats tests/*.bats --jobs 4

# Watch mode for TDD workflow
test-watch:
    @echo "Watching for changes..."
    while true; do \
        make test; \
        inotifywait -qre close_write tests/ scripts/; \
    done

# Clean up test artifacts
clean:
    @echo "Cleaning up test artifacts..."
    rm -rf tests/tmp/
    rm -f test_output.tap
    rm -f coverage/

# Run linting on shell scripts
lint:
    @echo "Linting shell scripts..."
    shellcheck scripts/*.sh
    shellcheck tests/*.bats

# Full validation: lint + test
validate: lint test
    @echo "✓ All checks passed"

Usage:

# Run all tests
make test

# Run with verbose output for debugging
make test-verbose

# Run tests in parallel for CI
make test-parallel

# Run only integration tests
make test-integration

# Full validation before commit
make validate

Best Practices Summary

Test Quality

1. One assertion per test - Tests should verify a single behavior
2. Descriptive test names - Use clear, complete sentences
3. Test independence - Tests should not depend on execution order
4. Clean up resources - Always remove temporary files in teardown
5. Test both paths - Verify success AND failure scenarios

Test Organization

1. Group related tests - Use separate files for unit vs integration tests
2. Use fixtures - Store test data in dedicated fixtures/ directory
3. Create helpers - Extract common patterns into test_helper.bash
4. Document complex setups - Explain unusual test patterns with comments

Performance

1. Use global setup - Run expensive operations once with setup_file
2. Run in parallel - Use bats --jobs N for faster test execution
3. Mock external calls - Stub network requests and slow commands
4. Keep tests fast - Each test should complete in milliseconds

Maintainability

1. Follow conventions - Consistent naming and structure
2. Version control fixtures - Check in test data files
3. Update tests with code - Keep tests in sync with implementation
4. Review test failures - Investigate and fix flaky tests immediately

CI/CD Integration

1. Run tests automatically - On every push and pull request
2. Test multiple environments - Different OS and shell versions
3. Generate reports - Use TAP output for test dashboards
4. Fail fast - Stop CI on test failures

Common Pitfalls and Solutions

Pitfall: Tests Pass but Code is Broken

Problem: Tests don't actually verify the behavior

Solution: Always watch tests fail first (TDD approach)

# Write test first
@test "Function returns correct value" {
    run my_function "input"
    [ "$output" = "expected" ]
}

# Run and verify it FAILS (function doesn't exist yet)
# Then implement the function
# Run again and verify it PASSES

Pitfall: Tests Depend on System State

Problem: Tests pass on developer machine but fail in CI

Solution: Isolate tests with proper setup/teardown

setup() {
    # Create isolated environment
    export HOME="$BATS_TEST_TMPDIR/home"
    export XDG_CONFIG_HOME="$HOME/.config"
    mkdir -p "$XDG_CONFIG_HOME"
    
    # Set up clean PATH
    export PATH="/usr/local/bin:/usr/bin:/bin"
}

Pitfall: Flaky Tests Due to Timing

Problem: Tests occasionally fail due to race conditions

Solution: Use proper synchronization

@test "Waits for background process" {
    my_background_task &
    local pid=$!
    
    # Wait with timeout
    for i in {1..30}; do
        if [ -f "$BATS_TEST_TMPDIR/done.txt" ]; then
            break
        fi
        sleep 0.1
    done
    
    wait $pid
    [ -f "$BATS_TEST_TMPDIR/done.txt" ]
}

Pitfall: Hard to Debug Failures

Problem: Test fails but output doesn't show why

Solution: Add diagnostic output

@test "Processes file correctly" {
    run process_file "$input"
    
    # Add diagnostic output on failure
    if [ "$status" -ne 0 ]; then
        echo "Command failed with status: $status" >&2
        echo "Output:" >&2
        echo "$output" >&2
        echo "Input file contents:" >&2
        cat "$input" >&2
    fi
    
    [ "$status" -eq 0 ]
}

Additional Resources

Official Documentation

• Bats Core: https://github.com/bats-core/bats-core
• Bats Docs: https://bats-core.readthedocs.io/
• TAP Protocol: https://testanything.org/

Bats Libraries

• bats-support: https://github.com/bats-core/bats-support - Additional assertions
• bats-assert: https://github.com/bats-core/bats-assert - Helpful assertion functions
• bats-file: https://github.com/bats-core/bats-file - File system assertions

Testing Methodology

• Test-Driven Development: https://en.wikipedia.org/wiki/Test-driven_development
• Testing Best Practices: Write tests that document behavior, not implementation

Shell Testing Tools

• ShellCheck: https://www.shellcheck.net/ - Static analysis for shell scripts
• shfmt: https://github.com/mvdan/sh - Shell script formatter
• bashate: https://github.com/openstack/bashate - Bash style checker

Quick Reference

Common Bats Variables

• $status - Exit code of last run command
• $output - Combined stdout/stderr of last run command
• $lines - Array of output lines from last run command
• ${lines[0]} - First line of output
• ${#lines[@]} - Number of output lines
• $BATS_TEST_DIRNAME - Directory containing the test file
• $BATS_TEST_FILENAME - Filename of the test file
• $BATS_TEST_NAME - Name of the current test
• $BATS_TEST_TMPDIR - Temporary directory for the current test

Common Assertions

# Exit codes
[ "$status" -eq 0 ]      # Success
[ "$status" -ne 0 ]      # Failure
[ "$status" -eq 127 ]    # Specific code

# String comparison
[ "$output" = "exact" ]                # Exact match
[[ "$output" == *"substring"* ]]       # Contains
[[ "$output" =~ ^pattern$ ]]           # Regex match

# File tests
[ -f "$file" ]           # File exists
[ -d "$dir" ]            # Directory exists
[ -r "$file" ]           # Readable
[ -w "$file" ]           # Writable
[ -x "$file" ]           # Executable
[ -s "$file" ]           # Not empty

# Numeric comparison
[ "$count" -eq 5 ]       # Equal
[ "$count" -gt 0 ]       # Greater than
[ "$count" -lt 10 ]      # Less than

Useful Patterns

# Run command and capture output
run command arg1 arg2

# Skip test conditionally
skip "Reason for skipping"
skip_if_missing "jq"

# Load helper functions
load test_helper

# Create temp directory
TEST_DIR=$(mktemp -d)

# Check command exists
command -v tool &>/dev/null

# Stub a command
PATH="/path/to/stubs:$PATH"