Example programs demonstrating AILANG features, organized by purpose and runnability.

examples/
├── runnable/          # ✅ Full working programs (21 files) - CI verified
├── snippets/          # 📝 Documentation code snippets (21 files)
├── tests/             # 🧪 Test cases and edge cases (37 files)
├── experimental/      # 🔬 Future features (9 files) - Not working yet
└── plans/             # Design plan examples

# Run a simple program
ailang run --caps IO examples/runnable/simple.ail --entry main

# Run with multiple capabilities
ailang run --caps IO,FS,Net examples/runnable/effects_basic.ail --entry main

# Run JSON parsing example
ailang run --caps IO examples/runnable/json_basic_decode.ail --entry main

ailang repl
> 2 + 3
5
> let x = 10 in x * 2
20

Copy expressions from snippets/ files and paste into the REPL.

Best for: Learning how to write complete AILANG programs

• ✅ All files have proper module structure with export func main()
• ✅ Verified by CI (make verify-examples)
• ✅ Expected to run without errors
• ✅ Demonstrate real-world usage patterns

Examples: ADTs, recursion, effects, blocks, JSON, I/O

Best for: Understanding AILANG syntax and features

• 📝 Single expressions and code fragments for tutorials
• 📝 Not meant to be run directly (no module wrapper)
• 📝 Can be copied into REPL for experimentation
• 📝 Show syntax and language features clearly

Examples: hello world, arithmetic, lambdas, records, type classes

Best for: Understanding how specific features should behave

• 🧪 Verify language behavior and edge cases
• 🧪 Some intentionally fail (testing error handling)
• 🧪 May require specific capabilities or external resources
• 🧪 Not all are verified by CI

Examples: effect system, pattern guards, module imports, builtins

Best for: Seeing what's coming in future releases

• ⚠️ DO NOT WORK with current version
• 🔬 Show intended future API and syntax
• 🔬 Will become working examples when features ship

Examples: AI agent integration, concurrency, HTTP servers

Basic Language

• Simple expressions: runnable/simple.ail
• Arithmetic: snippets/arithmetic.ail
• Functions: snippets/func_expressions.ail
• Lambdas: snippets/lambda_expressions.ail, snippets/showcase/02_lambdas.ail

Data Types

• Records: snippets/records.ail, runnable/micro_record_person.ail
• Lists: snippets/list_patterns.ail, snippets/showcase/03_lists.ail
• ADTs: runnable/adt_simple.ail, runnable/adt_option.ail
• Option type: runnable/micro_option_map.ail

Control Flow

• Pattern matching: runnable/patterns.ail
• Guards: runnable/guards_basic.ail, tests/test_guard_bool.ail
• Blocks: runnable/micro_block_if.ail, runnable/micro_block_seq.ail

Recursion

• Factorial: runnable/recursion_factorial.ail
• Fibonacci: runnable/recursion_fibonacci.ail
• Mutual recursion: runnable/recursion_mutual.ail
• Quicksort: runnable/recursion_quicksort.ail

Effects

• Basic I/O: runnable/effects_basic.ail, runnable/demos/hello_io.ail
• Pure vs effectful: runnable/effects_pure.ail
• Effect capabilities: tests/test_effect_capability.ail

Type System

• Type inference: snippets/showcase/01_type_inference.ail
• Type classes: snippets/typeclasses.ail, snippets/showcase/03_type_classes.ail
• Closures: snippets/showcase/04_closures.ail

JSON

• JSON parsing: runnable/json_basic_decode.ail

Modules

• Minimal module: tests/test_module_minimal.ail
• Imports: snippets/v3_3/imports.ail, tests/test_import_func.ail

Only runnable/ examples are verified by continuous integration:

make verify-examples  # Runs all runnable examples

Expected pass rate: 95-100% (21 examples, all should work)

1. Create file in examples/runnable/
2. Use proper module structure:

module examples/my_example

export func main() ! {IO} -> string {
  _io_println("Hello!");
  "ok"
}

3. Test it works: ailang run --caps IO examples/runnable/my_example.ail --entry main
4. CI will automatically verify it

1. Create file in examples/snippets/
2. Use simple expressions (no module wrapper needed):

-- Demonstrates feature X
let x = 10 in
x * 2  -- Returns 20

3. Add comments showing expected output
4. Test in REPL before committing

1. Create file in examples/tests/ with test_ prefix
2. Document expected behavior (pass/fail, error message)
3. Test manually with required capabilities

Before:

• All examples in flat examples/ directory
• CI verified everything (high failure rate)
• Unclear which examples should work

After (v0.3.14+):

• Organized by purpose and runnability
• CI only verifies runnable/ examples
• Clear expectations for each category

This improves:

• Clarity: Users know which examples should work
• CI reliability: Only test runnable programs
• Documentation: Snippets are clearly not full programs

• AILANG Documentation - Full language reference
• CHANGELOG.md - Version history
• CONTRIBUTING.md - Development guide
• Each subdirectory has its own README with details