gopantic

Practical JSON/YAML parsing with validation for Go.

Inspired by Python's Pydantic, gopantic provides type-safe parsing, coercion, and validation with idiomatic Go APIs.

Full Documentation: 1mb-dev.github.io/gopantic

Quick Start

go get github.com/1mb-dev/gopantic

package main

import (
    "fmt"
    "log"
    "github.com/1mb-dev/gopantic/pkg/model"
)

type User struct {
    ID    int    `json:"id" validate:"required,min=1"`
    Name  string `json:"name" validate:"required,min=2"`
    Email string `json:"email" validate:"required,email"`
    Age   int    `json:"age" validate:"min=18,max=120"`
}

func main() {
    raw := []byte(`{"id": "42", "name": "Alice", "email": "alice@example.com", "age": "28"}`)
    
    user, err := model.ParseInto[User](raw)
    if err != nil {
        log.Fatal(err)
    }
    
    fmt.Printf("%+v\n", user) // {ID:42 Name:Alice Email:alice@example.com Age:28}
}

Features

• JSON/YAML parsing with automatic format detection
• Type coercion ("123" → 123, "true" → true)
• Validation using struct tags (validate:"required,email,min=5")
• Standalone validation - use Validate() independently of parsing
• Cross-field validation (password confirmation, field comparisons)
• Built-in validators: required, min, max, email, alpha, alphanum, length
• Nested structs and arrays with full validation
• Time parsing (RFC3339, Unix timestamps, custom formats)
• Pointer support for optional fields (*string, *int)
• json.RawMessage support for flexible/deferred JSON parsing
• Caching support for repeated identical inputs (optional)
• Thread-safe concurrent parsing
• Zero dependencies (except optional YAML support)
• Generic API for type-safe parsing with compile-time guarantees

YAML Support

Works seamlessly with YAML - same API, automatic detection:

yamlData := []byte(`
id: 42
name: Alice
email: alice@example.com
age: 28
`)

user, err := model.ParseInto[User](yamlData) // Automatic YAML detection

json.RawMessage Support

gopantic seamlessly handles json.RawMessage fields for flexible metadata and JSONB database columns:

type Account struct {
    ID          string          `json:"id" validate:"required"`
    Name        string          `json:"name" validate:"required,min=2"`
    MetadataRaw json.RawMessage `json:"metadata,omitempty"`
}

input := []byte(`{
    "id": "acc_123",
    "name": "John Doe",
    "metadata": {"preferences": {"theme": "dark"}, "tags": ["vip"]}
}`)

account, err := model.ParseInto[Account](input)
// MetadataRaw preserves the raw JSON for later parsing

Standalone Validation

Use Validate() to validate structs from any source (database, environment, etc.):

type Config struct {
    Host string `json:"host" validate:"required"`
    Port int    `json:"port" validate:"min=1,max=65535"`
}

// Populate from environment or database
config := Config{
    Host: os.Getenv("HOST"),
    Port: 8080,
}

// Validate independently
if err := model.Validate(&config); err != nil {
    log.Fatal(err)
}

This enables flexible patterns:

// Pattern 1: Standard library unmarshal + gopantic validation
var req Request
json.Unmarshal(body, &req)  // Handles json.RawMessage correctly
model.Validate(&req)         // Apply gopantic validation rules

// Pattern 2: All-in-one (parsing + validation)
req, err := model.ParseInto[Request](body)

Validation

type Product struct {
    SKU   string  `json:"sku" validate:"required,length=8"`
    Price float64 `json:"price" validate:"required,min=0.01"`
}

// Multiple validation errors are aggregated
product, err := model.ParseInto[Product](invalidData)
// Error: "multiple errors: validation error on field 'SKU': ...; validation error on field 'Price': ..."

Cross-Field Validation

Built-in support for cross-field validation - compare fields against each other:

type UserRegistration struct {
    Password        string `json:"password" validate:"required,min=8"`
    ConfirmPassword string `json:"confirm_password" validate:"required,password_match"`
    Email           string `json:"email" validate:"required,email"`
    NotificationEmail string `json:"notification_email" validate:"email,email_different"`
}

// Register custom cross-field validators
model.RegisterGlobalCrossFieldFunc("password_match", func(fieldName string, fieldValue interface{}, structValue reflect.Value, params map[string]interface{}) error {
    confirmPassword := fieldValue.(string)
    password := structValue.FieldByName("Password").String()
    if confirmPassword != password {
        return model.NewValidationError(fieldName, fieldValue, "password_match", "passwords do not match")
    }
    return nil
})

Performance

Optimized for production use - recent improvements deliver 40-65% faster parsing with 64% less memory:

Metric	Standard Parsing	Cached Parsing
Speed vs stdlib JSON	1.7x slower	5.4x faster
Memory	435 B/op	112 B/op
Allocations	14/op	6/op

Use cached parser for repeated identical inputs (configs, retries):

parser := model.NewCachedParser[User](nil)
user, _ := parser.Parse(data) // 5.4x faster for cache hits

Note: YAML is 4x slower than JSON due to parser complexity. Use JSON for performance-critical paths.

Why Choose gopantic?

vs. Standard Library (encoding/json)

• Built-in validation - No separate validation step needed
• Type coercion - Handles "123" → 123 automatically
• Better errors - Structured error reporting with field paths
• YAML support - Automatic format detection
• Cross-field validation - Compare fields against each other

vs. Validation Libraries (go-playground/validator)

• Integrated parsing - Parse and validate in one step
• Type coercion - No manual string conversion needed
• Format agnostic - Works with JSON and YAML seamlessly
• Generics support - Type-safe with ParseInto[T]()
• Performance - Built-in caching for repeated operations

vs. Code Generation (easyjson, ffjson)

• Zero code generation - No build step or generated files
• Dynamic validation - Runtime validation rule changes
• Simpler workflow - Standard Go development process
• Faster iteration - No regeneration on struct changes
• Cross-field validation - Complex validation logic support

vs. Schema Libraries (jsonschema, gojsonschema)

• Native Go structs - Use existing struct definitions
• Compile-time safety - Type checking at compile time
• Better performance - Direct struct mapping vs. schema validation
• IDE support - Full autocompletion and refactoring
• Integrated coercion - Automatic type conversion

Development

Pre-commit Hook

A practical pre-commit hook is included that automatically:

• Checks for potential secrets in staged files
• Formats Go code with go fmt
• Runs go vet for static analysis
• Performs fast linting with golangci-lint
• Runs tests if test files are modified

Install the hook:

make hooks

Documentation

Full documentation: 1mb-dev.github.io/gopantic

Quick links:

• Getting Started - Installation and basic usage
• API Reference - Complete API documentation
• Validation Guide - All validation options
• Migration Guide - Coming from other libraries
• Examples - Runnable code examples

Resources:

• Changelog - Version history
• Security Policy - Vulnerability reporting

License

MIT License - see LICENSE for details.