Product Archetype (Archetyp produktu)

Based on "Enterprise Patterns and MDA" by Jim Arlow & Ila Neustadt

Problem

Product w katalogu mają złożone struktury:

• Kategorie ze wspólnymi atrybutami: Wszystkie laptopy mają RAM i pamięć masową; wszystkie koszulki mają rozmiar i kolor
• Katalog a fizyczne przedmioty: „iPhone 15 Pro 256GB Czarny" (katalog) vs „konkretny telefon z numerem seryjnym #XYZ" (egzemplarz)
• Warianty: Ten sam produkt w różnych rozmiarach/kolorach
• Złożoność cenowa: Cena bazowa, rabaty, ceny specyficzne dla klienta
• Hierarchiczne kategorie: Elektronika → Komputery → Laptopy → Laptopy gamingowe

Koncepcja

Archetyp Product ma trzy warstwy:

1. ProductType — Kategoria/szablon definiujący strukturę (atrybuty, reguły)
2. Product — Pozycja katalogowa, którą można zamówić (SKU, cena, opis)
3. ProductInstance — Fizyczny/serializowany przedmiot w magazynie

Można to sobie wyobrazić jako:

• ProductType = „Laptop" (definiuje jakie atrybuty istnieją)
• Product = „ThinkPad X1 Carbon Gen 11" (to, co zamawiasz)
• ProductInstance = „Nr seryjny #ABC123" (to, co otrzymujesz)

Struktura

Atrybuty

ProductType

Product

ProductInstance

AttributeDefinition

Zachowania

Entity ProductType:
    
    function isLeaf(): Boolean
        return children is empty
    
    function isRoot(): Boolean
        return parent is None
    
    function getPath(): String[]
        // Returns ["Electronics", "Computers", "Laptops"]
        if parent is None:
            return [name]
        return parent.getPath() + [name]
    
    function getAllAttributeDefinitions(): AttributeDefinition[]
        // Inherit from ancestors + own
        inherited = parent?.getAllAttributeDefinitions() or []
        return inherited + attributeDefinitions
    
    function canHaveProducts(): Boolean
        // Only leaf types can have products directly
        return isLeaf()
    
    function getDescendants(): ProductType[]
        result = children.copy()
        for child in children:
            result += child.getDescendants()
        return result


Entity Product:
    
    function isAvailable(): Boolean
        return status == ProductStatus.Active
    
    function getAttribute(code: String): Any
        return attributes.get(code)
    
    function setAttribute(code: String, value: Any):
        definition = productType.getAllAttributeDefinitions()
            .find(d => d.code == code)
        
        require definition is not None  // Attribute must be defined
        require definition.validate(value)  // Value must be valid
        
        attributes[code] = value
    
    function validate(): ValidationResult
        errors = []
        
        for definition in productType.getAllAttributeDefinitions():
            value = attributes.get(definition.code)
            
            if definition.required and value is None:
                errors.add("Missing required attribute: " + definition.code)
            
            if value is not None and not definition.validate(value):
                errors.add("Invalid value for " + definition.code)
        
        return ValidationResult(isValid: errors.isEmpty(), errors: errors)
    
    function calculatePrice(customer: Customer): Money
        price = basePrice
        
        // Apply customer-specific pricing
        if customer.loyaltyTier == "Gold":
            price = price * 0.95  // 5% discount
        
        // Apply quantity discounts, promotions, etc.
        return price
    
    function discontinue():
        require status == ProductStatus.Active
        status = ProductStatus.Discontinued
        emit ProductDiscontinued(this)


Entity ProductInstance:
    
    function isAvailable(): Boolean
        return status == InstanceStatus.Available
    
    function isExpired(): Boolean
        return expiryDate is not None and expiryDate < today()
    
    function sell(order: Order):
        require isAvailable()
        require not isExpired()
        
        status = InstanceStatus.Sold
        soldTo = order
        soldAt = now()
        
        emit InstanceSold(this, order)
    
    function markDamaged(reason: String):
        status = InstanceStatus.Damaged
        damageReason = reason
        damagedAt = now()


Entity AttributeDefinition:
    
    function validate(value: Any): Boolean
        if value is None:
            return not required
        
        // Type check
        match dataType:
            case DataType.String:
                if not (value is String): return false
            case DataType.Integer:
                if not (value is Integer): return false
            case DataType.Decimal:
                if not (value is Number): return false
            case DataType.Boolean:
                if not (value is Boolean): return false
            case DataType.Enum:
                if value not in options: return false
        
        // Range check
        if minValue is not None and value < minValue:
            return false
        if maxValue is not None and value > maxValue:
            return false
        
        return true

Przykłady

Konfiguracja katalogu produktów

// ProductType hierarchy: Electronics > Laptops
electronics = new ProductType(code: "ELECTRONICS", name: "Electronics")
laptops = new ProductType(code: "LAPTOP", name: "Laptops", parent: electronics)

// Define attributes for Laptops
laptops.attributeDefinitions.add(new AttributeDefinition(
    code: "RAM_GB", name: "RAM", dataType: DataType.Integer,
    required: true, minValue: 4, maxValue: 128, unit: "GB"
))
laptops.attributeDefinitions.add(new AttributeDefinition(
    code: "STORAGE_GB", name: "Storage", dataType: DataType.Integer,
    required: true, unit: "GB"
))
laptops.attributeDefinitions.add(new AttributeDefinition(
    code: "SCREEN_SIZE", name: "Screen Size", dataType: DataType.Decimal,
    required: true, unit: "inches"
))

Tworzenie Product i Product Instance

// Product (catalog item)
thinkpad = new Product(
    sku: "LEN-X1C-G11-16-512",
    name: "ThinkPad X1 Carbon Gen 11 (16GB/512GB)",
    productType: laptops,
    basePrice: Money(649900, "PLN"),  // 6499.00 PLN
    status: ProductStatus.Draft
)
thinkpad.setAttribute("RAM_GB", 16)
thinkpad.setAttribute("STORAGE_GB", 512)
thinkpad.setAttribute("SCREEN_SIZE", 14.0)

thinkpad.validate()  // isValid: true
thinkpad.publish()   // status = Active

// ProductInstance (specific unit in warehouse)
unit1 = new ProductInstance(
    product: thinkpad,
    serialNumber: "PF4ABC123",
    manufacturedDate: 2025-01-15,
    status: InstanceStatus.Available
)

Maszyny stanów

Status Product

Status Product Instance

Niezmienniki

1. Unikalny SKU — Żadne dwa Product nie mogą mieć tego samego SKU
2. Unikalny numer seryjny — Żadne dwa Product Instance nie mogą mieć tego samego numeru seryjnego
3. Acykliczna hierarchia typów — ProductType nie może być swoim własnym przodkiem
4. Product na typach liściowych — Product mogą należeć tylko do liściowych ProductType
5. Wymagane atrybuty ustawione — Aktywne Product muszą mieć ustawione wszystkie wymagane atrybuty
6. Poprawne wartości atrybutów — Wartości atrybutów muszą odpowiadać ich definicjom

Obsługa błędów

Schemat SQL

sql-- Product type hierarchy
CREATE TABLE product_type (
    id              UUID PRIMARY KEY,
    code            VARCHAR(50) NOT NULL UNIQUE,
    name            VARCHAR(100) NOT NULL,
    description     TEXT,
    parent_id       UUID REFERENCES product_type(id),
    active          BOOLEAN NOT NULL DEFAULT TRUE,
    created_at      TIMESTAMP NOT NULL DEFAULT CURRENT_TIMESTAMP
);

-- Attribute definitions for product types
CREATE TABLE attribute_definition (
    id              UUID PRIMARY KEY,
    product_type_id UUID NOT NULL REFERENCES product_type(id) ON DELETE CASCADE,
    code            VARCHAR(50) NOT NULL,
    name            VARCHAR(100) NOT NULL,
    data_type       VARCHAR(20) NOT NULL,  -- 'String', 'Integer', 'Decimal', 'Boolean', 'Enum'
    required        BOOLEAN NOT NULL DEFAULT FALSE,
    options         TEXT,  -- JSON array for Enum type: '["S", "M", "L", "XL"]'
    unit            VARCHAR(20),
    min_value       DECIMAL,
    max_value       DECIMAL,
    sort_order      INTEGER NOT NULL DEFAULT 0,
    
    UNIQUE(product_type_id, code)
);

-- Products (catalog items)
CREATE TABLE product (
    id              UUID PRIMARY KEY,
    sku             VARCHAR(50) NOT NULL UNIQUE,
    name            VARCHAR(255) NOT NULL,
    description     TEXT,
    product_type_id UUID NOT NULL REFERENCES product_type(id),
    base_price_cents INTEGER NOT NULL,
    base_price_currency CHAR(3) NOT NULL DEFAULT 'PLN',
    cost_price_cents INTEGER,
    cost_price_currency CHAR(3),
    status          VARCHAR(20) NOT NULL DEFAULT 'Draft',
    weight_kg       DECIMAL(10,3),
    length_cm       DECIMAL(10,2),
    width_cm        DECIMAL(10,2),
    height_cm       DECIMAL(10,2),
    created_at      TIMESTAMP NOT NULL DEFAULT CURRENT_TIMESTAMP,
    updated_at      TIMESTAMP NOT NULL DEFAULT CURRENT_TIMESTAMP,
    
    CONSTRAINT chk_product_status CHECK (status IN ('Draft', 'Active', 'Discontinued', 'Deleted'))
);

-- Product attribute values (EAV pattern)
CREATE TABLE product_attribute (
    product_id      UUID NOT NULL REFERENCES product(id) ON DELETE CASCADE,
    attribute_code  VARCHAR(50) NOT NULL,
    value_string    VARCHAR(255),
    value_integer   INTEGER,
    value_decimal   DECIMAL(15,4),
    value_boolean   BOOLEAN,
    value_date      DATE,
    
    PRIMARY KEY (product_id, attribute_code)
);

-- Product instances (serialized items)
CREATE TABLE product_instance (
    id              UUID PRIMARY KEY,
    product_id      UUID NOT NULL REFERENCES product(id),
    serial_number   VARCHAR(100) UNIQUE,
    batch_number    VARCHAR(50),
    manufactured_date DATE,
    expiry_date     DATE,
    status          VARCHAR(20) NOT NULL DEFAULT 'Available',
    created_at      TIMESTAMP NOT NULL DEFAULT CURRENT_TIMESTAMP,
    
    CONSTRAINT chk_instance_status CHECK (
        status IN ('Available', 'Reserved', 'Sold', 'Damaged', 'Expired', 'Returned')
    )
);

-- Indexes
CREATE INDEX idx_product_type ON product(product_type_id);
CREATE INDEX idx_product_status ON product(status);
CREATE INDEX idx_product_type_parent ON product_type(parent_id);
CREATE INDEX idx_instance_product ON product_instance(product_id);
CREATE INDEX idx_instance_status ON product_instance(status);
CREATE INDEX idx_instance_expiry ON product_instance(expiry_date) WHERE expiry_date IS NOT NULL;

Typowe zapytania

sql-- Get product with all attributes
SELECT 
    p.*,
    json_object_agg(pa.attribute_code, 
        COALESCE(pa.value_string, pa.value_integer::text, 
                 pa.value_decimal::text, pa.value_boolean::text, 
                 pa.value_date::text)
    ) as attributes
FROM product p
LEFT JOIN product_attribute pa ON pa.product_id = p.id
WHERE p.id = :productId
GROUP BY p.id;

-- Get products by type (including subtypes)
WITH RECURSIVE type_tree AS (
    SELECT id FROM product_type WHERE id = :typeId
    UNION ALL
    SELECT pt.id FROM product_type pt
    JOIN type_tree tt ON pt.parent_id = tt.id
)
SELECT p.* FROM product p
WHERE p.product_type_id IN (SELECT id FROM type_tree)
  AND p.status = 'Active';

-- Get type hierarchy path
WITH RECURSIVE type_path AS (
    SELECT id, name, parent_id, 1 as level
    FROM product_type WHERE id = :typeId
    UNION ALL
    SELECT pt.id, pt.name, pt.parent_id, tp.level + 1
    FROM product_type pt
    JOIN type_path tp ON pt.id = tp.parent_id
)
SELECT name FROM type_path ORDER BY level DESC;

-- Find available instances with earliest expiry (FEFO)
SELECT pi.* FROM product_instance pi
WHERE pi.product_id = :productId
  AND pi.status = 'Available'
  AND (pi.expiry_date IS NULL OR pi.expiry_date > CURRENT_DATE)
ORDER BY pi.expiry_date NULLS LAST
LIMIT :quantity;

Zagadnienia projektowe

EAV vs JSON vs szerokie tabele

Zalecenie:

• EAV dla naprawdę dynamicznych atrybutów
• JSON dla danych semi-strukturalnych
• Dedykowane kolumny dla często wyszukiwanych/filtrowanych atrybutów

Kiedy stosować Product Instance

Strategie cenowe

Archetyp Product obsługuje różne modele cenowe:

• Cena bazowa — W encji Product
• Ceny według poziomu klienta — Obliczane w czasie rzeczywistym
• Progi ilościowe — Osobna encja PriceBreak
• Promocje czasowe — Encja Promotion z datami

Related Patterns

• Order — Product są zamawiane
• Inventory — Product Instance śledzony w Inventory
• Party Role (Supplier) — Kto dostarcza Product