Dingo

Dingo Language Patterns

Overview

Dingo is a meta-language for Go that transpiles .dingo files to .go files, providing modern language features while maintaining 100% Go ecosystem compatibility.

Repository: https://github.com/MadAppGang/dingo

When to use Dingo:

• You want concise error handling with the ? operator
• You need sum types (enums) and pattern matching
• You prefer functional patterns with lambdas
• You want Option[T] and Result[T,E] types
• You need safe navigation (?.) and null coalescing (??)

Key Philosophy: Dingo makes common Go patterns more concise and safer without departing from Go idioms. The transpiled Go code is clean and idiomatic.

Project Structure

project/
├── cmd/
│   └── api/
│       └── main.dingo         # Entry point
├── internal/
│   ├── handlers/              # HTTP handlers (.dingo files)
│   ├── services/              # Business logic
│   ├── repositories/          # Data access
│   └── models/                # Domain models
├── pkg/                       # Public packages
├── configs/                   # Configuration
├── go.mod                     # Go module file
├── go.sum                     # Go dependencies
└── .dingo/                    # Generated .go files (gitignored)

Build & Development Workflow

Commands

dingo build          # Transpile to Go and build binary
dingo run main.dingo # Transpile and run directly
dingo go             # Generate .go files only (for CI/CD)
dingo fmt            # Format Dingo files

Development Cycle

1. Write: Edit .dingo files
2. Transpile: Run dingo go to generate .go files in .dingo/
3. Type Check: gopls works on generated .go files for IDE support
4. Build: dingo build or go build on generated files

IDE Integration

gopls (Go language server) works on the generated .go files:

# After dingo go, gopls sees:
.dingo/
├── cmd/api/main.go
├── internal/handlers/user.go
└── internal/services/user.go

Configure your editor to watch .dingo/ for Go analysis.

CI/CD Pipeline

# GitHub Actions example
steps:
  - name: Install Dingo
    run: go install github.com/MadAppGang/dingo/cmd/dingo@latest

  - name: Generate Go files
    run: dingo go

  - name: Build
    run: go build -o app ./.dingo/cmd/api

  - name: Test
    run: go test ./.dingo/...

Transpilation Errors

Dingo reports errors with source locations in .dingo files:

error: mismatched types in match expression
  --> internal/handlers/user.dingo:42:5
   |
42 |     match status {
   |     ^^^^^ expected Status, found string

Fix errors in .dingo source, then re-run dingo go.

Built-in Types

Important: Option[T], Result[T,E], Some(), None(), Ok(), Err() are built-in Dingo syntax, not Go imports. The transpiler generates all necessary type definitions.

// These are built-in - no import needed
func findUser(id string) Option[User] {
    // ...
    return Some(user)  // Built-in function
    return None()      // Built-in function
}

func divide(a, b int) Result[int, string] {
    if b == 0 {
        return Err("division by zero")  // Built-in
    }
    return Ok(a / b)  // Built-in
}

Only import dgo if writing pure Go code that interoperates with Dingo-generated types.

Error Propagation (? Operator)

The ? operator provides concise error handling, similar to Rust.

Basic Propagation

// Before: verbose Go pattern
func loadUser(id string) (*User, error) {
    user, err := db.FindByID(id)
    if err != nil {
        return nil, err
    }
    return user, nil
}

// After: concise Dingo
func loadUser(id string) (*User, error) {
    user := db.FindByID(id)?
    return user, nil
}

Error Context

Add context to propagated errors:

func processOrder(id string) (*Order, error) {
    // Wrap error with message
    order := db.FindOrder(id) ? "failed to find order"

    // Validates and wraps
    validated := validateOrder(order) ? "order validation failed"

    // Process with full context
    result := processPayment(validated) ? "payment processing failed"

    return result, nil
}

Error Transform

Transform errors with closures:

func createUser(input CreateUserInput) (*User, error) {
    // Rust-style closure
    user := db.Create(input) ? |e| fmt.Errorf("db error: %w", e)

    // TypeScript-style arrow
    profile := createProfile(user.ID) ? e => AppError.Wrap(e, "profile creation")

    return user, nil
}

Error Propagation in Lambdas

When using ? inside lambda bodies, the error propagates to the enclosing function:

func processAllUsers(ids []string) (*Summary, error) {
    // Error in lambda propagates to processAllUsers
    results := map(ids, |id| {
        user := db.FindUser(id)?  // Propagates to processAllUsers, not the lambda
        return transform(user)
    })

    return summarize(results), nil
}

// For lambdas that should handle errors internally:
func processAllUsersSafe(ids []string) []Result[User, error] {
    return map(ids, |id| {
        user, err := db.FindUser(id)
        if err != nil {
            return Err(err)
        }
        return Ok(user)
    })
}

Important: The ? operator inside lambdas propagating to the enclosing function is intentional Dingo design. When the transpiler sees ? inside a lambda body, it generates special code that propagates errors to the enclosing function rather than the lambda itself. This is the most common use case for error handling in functional chains. If you need the lambda to handle errors internally (e.g., to return Result types), use explicit Go-style error checking as shown in processAllUsersSafe above.

Option[T] Type

Use Option[T] for optional values instead of nullable pointers.

Basic Usage

func findUser(id string) Option[User] {
    user, err := db.FindByID(id)
    if err != nil {
        return None()
    }
    return Some(user)
}

// Usage
user := findUser("123")
if user.IsSome() {
    fmt.Println(user.Unwrap().Name)
}

// With default
name := findUser("123").Map(|u| u.Name).UnwrapOr("Anonymous")

Option Methods

opt := Some("hello")

opt.IsSome()                    // true
opt.IsNone()                    // false
opt.Unwrap()                    // "hello" (panics if None)
opt.UnwrapOr("default")         // "hello"
opt.UnwrapOrElse(|| "computed") // "hello"
opt.Map(|s| len(s))             // Some(5)
opt.FlatMap(|s| Some(s + "!"))  // Some("hello!")

Result[T, E] Type

Use Result[T, E] for explicit error modeling.

Type Parameter Inference: Type parameters can be inferred when the context makes them clear (e.g., function return type), but explicit parameters are needed in standalone expressions like Ok[int, string](42).

Basic Usage

func divide(a, b int) Result[int, string] {
    if b == 0 {
        return Err("division by zero")
    }
    return Ok(a / b)
}

// Usage
result := divide(10, 2)
if result.IsOk() {
    fmt.Println(result.Unwrap()) // 5
}

// With default
value := divide(10, 0).UnwrapOr(0) // 0

Result Methods

ok := Ok[int, string](42)
err := Err[int, string]("failed")

ok.IsOk()              // true
ok.IsErr()             // false
ok.Unwrap()            // 42
ok.UnwrapErr()         // panics
ok.UnwrapOr(0)         // 42
ok.Map(|n| n * 2)      // Ok(84)

err.UnwrapOr(0)        // 0
err.UnwrapErr()        // "failed"

Sum Types (Enums)

Define algebraic data types with named variants.

Basic Enum

enum Status {
    Pending
    Active
    Suspended { reason: string }
    Deleted { deletedAt: time.Time, deletedBy: string }
}

func (s Status) String() string {
    match s {
        Pending => "pending",
        Active => "active",
        Suspended(reason) => fmt.Sprintf("suspended: %s", reason),
        Deleted(at, by) => fmt.Sprintf("deleted at %v by %s", at, by),
    }
}

Enum Instantiation

Use constructor syntax to create enum variants:

// Simple variants (no fields)
status := Status.Pending
active := Status.Active

// Variants with fields - constructor syntax
suspended := Status.Suspended("policy violation")
deleted := Status.Deleted(time.Now(), "admin@example.com")

Event Sourcing Example

enum UserEvent {
    Created { userID: int, email: string, createdAt: time.Time }
    EmailChanged { userID: int, oldEmail: string, newEmail: string }
    Deactivated { userID: int, reason: string }
    Reactivated { userID: int }
}

// Creating events - use constructor syntax
func recordUserCreation(id int, email string) UserEvent {
    return UserEvent.Created(id, email, time.Now())
}

func recordEmailChange(id int, old, new string) UserEvent {
    return UserEvent.EmailChanged(id, old, new)
}

func processEvent(event UserEvent) {
    match event {
        Created(id, email, _) => {
            fmt.Printf("User %d created with email %s\n", id, email)
        },
        EmailChanged(id, old, new) => {
            fmt.Printf("User %d changed email from %s to %s\n", id, old, new)
        },
        Deactivated(id, reason) => {
            fmt.Printf("User %d deactivated: %s\n", id, reason)
        },
        Reactivated(id) => {
            fmt.Printf("User %d reactivated\n", id)
        },
    }
}

Pattern Matching (match)

Exhaustive pattern matching with guards.

Basic Matching

func describe(n int) string {
    match n {
        0 => "zero",
        1 => "one",
        _ if n < 0 => "negative",
        _ if n > 100 => "large",
        _ => "other",
    }
}

With Enum Destructuring

enum Shape {
    Circle { radius: float64 }
    Rectangle { width: float64, height: float64 }
    Triangle { base: float64, height: float64 }
}

func area(shape Shape) float64 {
    match shape {
        Circle(r) => 3.14159 * r * r,
        Rectangle(w, h) => w * h,
        Triangle(b, h) => 0.5 * b * h,
    }
}

Nested Match with Guards

Complex matching scenarios with nested patterns:

enum Response {
    Success { data: Option[User], cached: bool }
    Error { code: int, message: string }
}

func handleResponse(resp Response) string {
    match resp {
        // Nested pattern: match on enum variant AND option state
        Success(Some(user), true) => {
            fmt.Sprintf("Cached user: %s", user.Name)
        },
        Success(Some(user), false) => {
            fmt.Sprintf("Fresh user: %s", user.Name)
        },
        Success(None(), _) => "No user data",

        // Guard on nested field
        Error(code, msg) if code >= 500 => {
            fmt.Sprintf("Server error %d: %s", code, msg)
        },
        Error(code, msg) if code >= 400 => {
            fmt.Sprintf("Client error %d: %s", code, msg)
        },
        Error(code, msg) => {
            fmt.Sprintf("Unknown error %d: %s", code, msg)
        },
    }
}

Wildcard Ignoring

func getOrderSummary(event OrderEvent) string {
    match event {
        OrderPlaced(id, _, amount) => fmt.Sprintf("Order %s: $%.2f", id, amount),
        OrderShipped(id, _, _) => fmt.Sprintf("Order %s shipped", id),
        _ => "Unknown event",
    }
}

Lambda Expressions

Concise function literals with two syntax styles.

Rust-Style (Pipe Syntax)

users := []User{...}

// Single argument
activeUsers := filter(users, |u| u.Active)

// Multiple arguments
sorted := sort(users, |a, b| a.Name < b.Name)

// With block body
processed := map(users, |u| {
    name := strings.ToUpper(u.Name)
    return fmt.Sprintf("%s (%d)", name, u.Age)
})

TypeScript-Style (Arrow Syntax)

// Single argument
activeUsers := filter(users, u => u.Active)

// Multiple arguments (parentheses required)
sorted := sort(users, (a, b) => a.CreatedAt.Before(b.CreatedAt))

// With block
processed := map(users, u => {
    return u.Name + " - " + u.Email
})

Practical Examples

// Filtering
adults := filter(people, |p| p.Age >= 18)

// Mapping
names := map(users, |u| u.Name)

// Reducing
total := reduce(orders, 0, |sum, o| sum + o.Amount)

// Chaining with Option
result := findUser(id)
    .Map(|u| u.Profile)
    .FlatMap(|p| p.Avatar)
    .UnwrapOr(defaultAvatar)

Safe Navigation (?.) and Null Coalescing (??)

Handle nullable chains safely.

Safe Navigation

type Config struct {
    Database *DatabaseConfig
}

type DatabaseConfig struct {
    Primary *ConnectionInfo
}

type ConnectionInfo struct {
    Host string
    Port int
}

// Safe navigation through nullable chain
host := config?.Database?.Primary?.Host ?? "localhost"
port := config?.Database?.Primary?.Port ?? 5432

Chained Safe Navigation with Null Coalescing

// Deep chain with fallback
dbHost := appConfig?.Database?.Primary?.Host
    ?? appConfig?.Database?.Fallback?.Host
    ?? envConfig?.DbHost
    ?? "localhost"

// Method chains
userName := response?.Data?.User?.Profile?.DisplayName
    ?? response?.Data?.User?.Name
    ?? "Anonymous"

// With function calls in fallback
timeout := config?.Timeout ?? getDefaultTimeout()

Null Coalescing

// Simple default
name := user.Nickname ?? user.Name ?? "Anonymous"

// With function call
timeout := config.Timeout ?? getDefaultTimeout()

// Combined with safe navigation
dbHost := appConfig?.Database?.Host ?? envConfig?.DbHost ?? "localhost"

Ternary Operator

Simple conditional expressions.

// Basic ternary
status := user.Active ? "Active" : "Inactive"

// Nested (use sparingly)
role := user.IsAdmin ? "Admin" : user.IsModerator ? "Moderator" : "User"

// In function calls
greet(user.Preferred ? user.Nickname : user.FullName)

// With expressions
discount := order.Total > 100 ? order.Total * 0.1 : 0

Tuples

Lightweight compound values.

Tuple Types

type Point2D = (float64, float64)
type Range = (int, int)
type NamedResult = (string, int, error)

Tuple Destructuring

func getCoordinates() (float64, float64) {
    return (42.5, 73.2)
}

// Destructure in assignment
(x, y) := getCoordinates()

// Ignore values with _
(lat, _) := getCoordinates()

// Multiple return handling
(user, err) := fetchUser(id)
if err != nil {
    return err
}

Tuple in Functions

func minMax(numbers []int) (int, int) {
    min := numbers[0]
    max := numbers[0]
    for _, n := range numbers {
        if n < min { min = n }
        if n > max { max = n }
    }
    return (min, max)
}

(lo, hi) := minMax([]int{3, 1, 4, 1, 5, 9})

Guard Statement

Early returns with clear error handling.

Error Binding Rules:

• Use guard x := f() else |err| { ... } when f() returns (T, error) or Result[T, E] - the error value is bound to err for custom handling
• Use guard x := f() else { ... } when f() returns Option[T] - None has no error value to bind

Basic Guard

func processUser(id string) Result[User, error] {
    // Guard with error handling
    guard user := findUser(id) else |err| {
        return Err(fmt.Errorf("user not found: %w", err))
    }

    guard profile := user.Profile else {
        return Err(errors.New("user has no profile"))
    }

    // Both user and profile are now available
    return Ok(user)
}

Guard with Option

func getDisplayName(userId string) string {
    guard user := findUser(userId) else {
        return "Unknown User"
    }

    guard nickname := user.Nickname else {
        return user.FullName
    }

    return nickname
}

Multiple Guards

func createOrder(req CreateOrderRequest) (*Order, error) {
    guard user := findUser(req.UserID) else |err| {
        return nil, fmt.Errorf("invalid user: %w", err)
    }

    guard cart := getCart(user.ID) else |err| {
        return nil, fmt.Errorf("cart not found: %w", err)
    }

    guard len(cart.Items) > 0 else {
        return nil, errors.New("cart is empty")
    }

    guard total := calculateTotal(cart) else |err| {
        return nil, fmt.Errorf("calculation failed: %w", err)
    }

    return createOrderFromCart(user, cart, total)
}

HTTP Handler Patterns

Handler with Error Propagation

// Standard Go imports apply - import net/http, encoding/json, etc.
import "github.com/go-chi/chi/v5"

func (h *Handler) GetUser(w http.ResponseWriter, r *http.Request) {
    id := chi.URLParam(r, "id")

    user := h.userService.FindByID(r.Context(), id) ? |e| {
        h.handleError(w, e)
        return
    }

    h.respond(w, http.StatusOK, user)
}

func (h *Handler) CreateUser(w http.ResponseWriter, r *http.Request) {
    var req CreateUserRequest
    json.NewDecoder(r.Body).Decode(&req) ? |e| {
        h.respondError(w, http.StatusBadRequest, "invalid request body")
        return
    }

    user := h.userService.Create(r.Context(), req) ? |e| {
        h.handleError(w, e)
        return
    }

    h.respond(w, http.StatusCreated, user)
}

Service Layer with Result Types

type UserService struct {
    repo UserRepository
}

func (s *UserService) FindByID(ctx context.Context, id string) Result[User, error] {
    guard user := s.repo.FindByID(ctx, id) else |err| {
        if errors.Is(err, sql.ErrNoRows) {
            return Err(NotFoundError("user"))
        }
        return Err(InternalError(err))
    }

    return Ok(user)
}

func (s *UserService) Create(ctx context.Context, req CreateUserRequest) Result[User, error] {
    // Check if email exists
    existing := s.repo.FindByEmail(ctx, req.Email)
    if existing.IsSome() {
        return Err(ConflictError("email already exists"))
    }

    hashedPassword := hashPassword(req.Password)?

    user := s.repo.Create(ctx, User{
        Email:        req.Email,
        PasswordHash: hashedPassword,
        Name:         req.Name,
    })?

    return Ok(user)
}

Repository Pattern

type UserRepository interface {
    FindByID(ctx context.Context, id string) Result[User, error]
    FindByEmail(ctx context.Context, email string) Option[User]
    Create(ctx context.Context, user User) Result[User, error]
    Update(ctx context.Context, user User) Result[User, error]
    Delete(ctx context.Context, id string) Result[bool, error]
}

type postgresUserRepo struct {
    db *sql.DB
}

func (r *postgresUserRepo) FindByID(ctx context.Context, id string) Result[User, error] {
    var user User
    err := r.db.QueryRowContext(ctx,
        "SELECT id, email, name, created_at FROM users WHERE id = $1",
        id,
    ).Scan(&user.ID, &user.Email, &user.Name, &user.CreatedAt)

    if err != nil {
        if errors.Is(err, sql.ErrNoRows) {
            return Err(NotFoundError("user"))
        }
        return Err(err)
    }

    return Ok(user)
}

func (r *postgresUserRepo) FindByEmail(ctx context.Context, email string) Option[User] {
    var user User
    err := r.db.QueryRowContext(ctx,
        "SELECT id, email, name FROM users WHERE email = $1",
        email,
    ).Scan(&user.ID, &user.Email, &user.Name)

    if err != nil {
        return None()
    }
    return Some(user)
}

Testing

Unit Tests with Option/Result

func TestUserService_FindByID(t *testing.T) {
    repo := mocks.NewMockUserRepository(t)
    service := NewUserService(repo)

    t.Run("returns user when found", func(t *testing.T) {
        expected := User{ID: "123", Name: "John"}
        repo.EXPECT().FindByID(mock.Anything, "123").Return(Ok(expected))

        result := service.FindByID(context.Background(), "123")

        assert.True(t, result.IsOk())
        assert.Equal(t, expected, result.Unwrap())
    })

    t.Run("returns error when not found", func(t *testing.T) {
        repo.EXPECT().FindByID(mock.Anything, "456").Return(
            Err[User, error](NotFoundError("user")),
        )

        result := service.FindByID(context.Background(), "456")

        assert.True(t, result.IsErr())
        assert.Contains(t, result.UnwrapErr().Error(), "not found")
    })
}

Testing Pattern Matching

func TestEventProcessing(t *testing.T) {
    tests := []struct {
        name     string
        event    UserEvent
        expected string
    }{
        {
            name:     "created event",
            event:    UserEvent.Created(1, "test@example.com", time.Now()),
            expected: "User 1 created",
        },
        {
            name:     "deactivated event",
            event:    UserEvent.Deactivated(1, "violation"),
            expected: "User 1 deactivated",
        },
    }

    for _, tt := range tests {
        t.Run(tt.name, func(t *testing.T) {
            result := processEvent(tt.event)
            assert.Contains(t, result, tt.expected)
        })
    }
}

Best Practices

Prefer Dingo Idioms Over Go Patterns

Error Handling Guidelines

1. Use ? for simple propagation
2. Add context with ? "message" or ? \|e\| wrap(e)
3. Use Result[T, E] when callers need error type information
4. Use guard for early returns that simplify the happy path

Type Safety Guidelines

1. Prefer Option[T] over *T for optional values
2. Use Result[T, E] for operations that can fail
3. Use enum for closed sets of variants
4. Always handle all match cases (exhaustive matching)

Lambda Guidelines

1. Use |x| expr for simple transforms
2. Use (a, b) => expr for multi-arg comparators
3. Add block { ... } for multi-line bodies
4. Prefer named functions for complex logic

Project Organization

1. Keep .dingo files in the same structure as a Go project
2. The .dingo/ directory contains generated Go files
3. Add .dingo/ to .gitignore
4. Run dingo go before go build in CI

Dingo language patterns for Go meta-programming