System Design (LLD)Setter Injection

Setter Injection: Providing Dependencies Through Mutator Methods

LevelIntermediate

Duration50 mins

TopicSetter Injection

1 / 4

Passing Dependencies Through Setters

Beyond the Constructor: An Alternative Path to Dependency Wiring

In the previous module, we explored constructor injection—the technique where dependencies are passed at object creation time through constructor parameters. Constructor injection is widely regarded as the gold standard for mandatory dependencies because it enforces completeness: an object cannot exist without its required collaborators.

But what happens when a dependency isn't mandatory? What if a component can function with a default behavior, but should be reconfigurable after instantiation? What if the dependency isn't known at construction time?

Enter setter injection—an alternative dependency injection mechanism where dependencies are provided through mutator methods (setters) after the object has been constructed. This technique opens doors that constructor injection cannot, but it also introduces complexities that demand careful consideration.

What You Will Learn

By the end of this page, you will understand the fundamental mechanics of setter injection, how it differs from constructor injection architecturally, the method signature patterns that enable clean setter injection, and the critical role of null-state handling in setter-injected components. You'll see complete code examples demonstrating proper implementation.

The Mechanics of Setter Injection

Setter injection is conceptually straightforward: rather than receiving dependencies at construction time, an object exposes setter methods that external code (typically a container or composition root) calls to provide dependencies after the object exists.

The core pattern involves three elements:

A private field to hold the dependency reference
A setter method that accepts the dependency and assigns it to the field
Usage code that checks for the dependency's presence before use

Let's examine the simplest possible implementation:

EmailService.ts
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
// Basic setter injection pattern
 
interface ILogger {
    log(message: string): void;
}
 
class EmailService {
    // Private field to hold the optional dependency
    private logger: ILogger | null = null;
 
    // Setter method - the injection point
    public setLogger(logger: ILogger): void {
        this.logger = logger;
    }
 
    public sendEmail(to: string, subject: string, body: string): void {
        // Check if logger is available before using
        if (this.logger) {
            this.logger.log(`Sending email to ${to}: ${subject}`);
        }
 
        // Core email sending logic
        this.performEmailDelivery(to, subject, body);
 
        if (this.logger) {
            this.logger.log(`Email sent successfully to ${to}`);
        }
    }
 
    private performEmailDelivery(to: string, subject: string, body: string): void {
        // Actual email sending implementation
        console.log(`Delivering email to ${to}`);
    }
}
 
// Usage: Dependency is injected AFTER construction
const emailService = new EmailService();
 
// The service works without a logger (graceful degradation)
emailService.sendEmail("user@example.com", "Hello", "World");
 
// Later, inject a logger
const consoleLogger: ILogger = {
    log: (message) => console.log(`[LOG] ${message}`)
};
emailService.setLogger(consoleLogger);
 
// Now the service logs its operations
emailService.sendEmail("user@example.com", "Hello Again", "World");

The Two-Phase Lifecycle

Notice the key difference from constructor injection: the object goes through a two-phase lifecycle. First, it's constructed (potentially in an incomplete state). Second, dependencies are injected. This two-phase approach is both the power and the peril of setter injection—it enables flexibility but requires careful handling of the 'incomplete' state.

Architectural Comparison with Constructor Injection

Understanding setter injection requires understanding how it fundamentally differs from constructor injection. These aren't just syntactic variations—they represent different architectural decisions with cascading implications.

Object Invariants and State Validity

Constructor injection enforces what we call a strong invariant: the object is never in an invalid state because all required dependencies are provided at construction time. The constructor acts as a gatekeeper—if you have a reference to the object, you know it's fully configured.

Setter injection, by contrast, produces a weak invariant: the object can exist without its dependencies. This means the class must be designed to handle the "unconfigured" state gracefully, or external code must guarantee proper initialization sequences.

Constructor Injection vs Setter Injection: Architectural Comparison
Aspect	Constructor Injection	Setter Injection
Dependency Availability	Guaranteed from creation	Must be null-checked before use
Object State	Always valid	Potentially incomplete
Immutability	Dependencies can be final/readonly	Dependencies are mutable by design
Testing	Mock at construction	Mock via setter before test
Framework Integration	May require complex factories	Simpler framework wiring
Circular Dependencies	Causes construction deadlock	Can be resolved post-construction
Configuration Ordering	Implicit (construction order)	Explicit (setter call order matters)
API Clarity	Dependencies visible in constructor	Dependencies scattered across setters
Fail-Fast	Fails immediately on missing dependency	May fail later at runtime

Visualizing the Difference: Object Creation Flow

Consider how objects come into existence with each approach:

Constructor Injection Flow:

[Resolve Dependencies] → [Create Object with Dependencies] → [Ready to Use]

Setter Injection Flow:

[Create Empty Object] → [Call Setter 1] → [Call Setter 2] → ... → [Ready to Use]

The setter injection flow has more steps and, crucially, the "Ready to Use" state isn't enforced by the language—it depends on all setters being called correctly. This is why we say setter injection trades compile-time safety for runtime flexibility.

ObjectLifecycleComparison.ts
TypeScript
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
// CONSTRUCTOR INJECTION: Strong invariant
 
interface IDatabase {
    query(sql: string): any[];
}
 
interface ICache {
    get(key: string): any;
    set(key: string, value: any): void;
}
 
class UserRepositoryConstructor {
    private readonly database: IDatabase;
    private readonly cache: ICache;
 
    // Object cannot exist without these dependencies
    constructor(database: IDatabase, cache: ICache) {
        if (!database) throw new Error("database is required");
        if (!cache) throw new Error("cache is required");
        this.database = database;
        this.cache = cache;
    }
 
    // Methods can assume dependencies exist - no null checks needed
    findUser(id: string): any {
        const cached = this.cache.get(id);
        if (cached) return cached;
        
        const user = this.database.query(`SELECT * FROM users WHERE id = '${id}'`)[0];
        this.cache.set(id, user);
        return user;
    }
}
 
// SETTER INJECTION: Weak invariant
 
class UserRepositorySetter {
    private database: IDatabase | null = null;
    private cache: ICache | null = null;
 
    // Object exists without dependencies
    constructor() {}
 
    setDatabase(database: IDatabase): void {
        this.database = database;
    }
 
    setCache(cache: ICache): void {
        this.cache = cache;
    }
 
    // Methods must handle missing dependencies
    findUser(id: string): any {
        // Must check every time - defensive programming required
        if (!this.database) {
            throw new Error("Database not configured - call setDatabase first");
        }
 
        // Cache is optional - check before use
        if (this.cache) {
            const cached = this.cache.get(id);
            if (cached) return cached;
        }
 
        const user = this.database.query(`SELECT * FROM users WHERE id = '${id}'`)[0];
        
        // Only cache if cache is available
        if (this.cache) {
            this.cache.set(id, user);
        }
        
        return user;
    }
}
 
// Usage comparison
 
// Constructor: One-shot, guaranteed complete
const repoConstructor = new UserRepositoryConstructor(
    mockDatabase,
    mockCache
);
// Immediately usable
repoConstructor.findUser("123");
 
// Setter: Multi-step, must verify completeness
const repoSetter = new UserRepositorySetter();
// NOT usable yet - would throw error
repoSetter.setDatabase(mockDatabase);
// Still missing cache, but might work (cache is optional)
repoSetter.setCache(mockCache);
// Now fully configured
repoSetter.findUser("123");

The Hidden Cost of Weak Invariants

Notice how the setter injection version requires null checks throughout the class. These checks add cognitive overhead, increase code volume, and create potential failure points. Every method that uses a setter-injected dependency must decide: throw an error if missing, silently skip the operation, or use a default behavior. This decision complexity is the tax you pay for setter injection's flexibility.

Method Signature Patterns for Setter Injection

The design of setter methods significantly impacts usability, safety, and maintainability. Let's explore common patterns for designing setter injection points, from basic setters to fluent interfaces and validation-enhanced variants.

Pattern 1: Simple Setter

The most basic pattern—a void method that accepts and stores the dependency:

SimpleSetterPattern.ts
TypeScript
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
// Pattern 1: Simple Setter
// Pros: Simple, familiar, works with most frameworks
// Cons: No validation, no fluent chaining
 
class NotificationService {
    private emailSender: IEmailSender | null = null;
    private smsSender: ISmsSender | null = null;
 
    // Simple setter - just stores the reference
    public setEmailSender(sender: IEmailSender): void {
        this.emailSender = sender;
    }
 
    public setSmsSender(sender: ISmsSender): void {
        this.smsSender = sender;
    }
}
 
// Usage
const service = new NotificationService();
service.setEmailSender(new SmtpEmailSender());
service.setSmsSender(new TwilioSmsSender());

Pattern 2: Fluent Setter (Method Chaining)

Return this from setters to enable fluent configuration:

FluentSetterPattern.ts
TypeScript
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
// Pattern 2: Fluent Setter (Method Chaining)
// Pros: Cleaner configuration code, self-documenting
// Cons: May not be recognized by all DI frameworks
 
class NotificationService {
    private emailSender: IEmailSender | null = null;
    private smsSender: ISmsSender | null = null;
    private logger: ILogger | null = null;
 
    // Fluent setters return 'this' for chaining
    public setEmailSender(sender: IEmailSender): this {
        this.emailSender = sender;
        return this;
    }
 
    public setSmsSender(sender: ISmsSender): this {
        this.smsSender = sender;
        return this;
    }
 
    public setLogger(logger: ILogger): this {
        this.logger = logger;
        return this;
    }
}
 
// Fluent usage - reads like a configuration DSL
const service = new NotificationService()
    .setEmailSender(new SmtpEmailSender())
    .setSmsSender(new TwilioSmsSender())
    .setLogger(new ConsoleLogger());
 
// Compare to non-fluent equivalent:
const service2 = new NotificationService();
service2.setEmailSender(new SmtpEmailSender());
service2.setSmsSender(new TwilioSmsSender());
service2.setLogger(new ConsoleLogger());

Pattern 3: Validating Setter

Add validation logic to prevent invalid configurations:

ValidatingSetterPattern.ts
TypeScript
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
// Pattern 3: Validating Setter
// Pros: Fail-fast on invalid configuration, clearer error messages
// Cons: More code, may throw during framework wiring
 
class NotificationService {
    private emailSender: IEmailSender | null = null;
    private maxRetries: number = 3;
    private _initialized: boolean = false;
 
    // Validating setter - checks preconditions
    public setEmailSender(sender: IEmailSender): void {
        if (!sender) {
            throw new Error("Email sender cannot be null");
        }
        
        // Prevent reconfiguration after initialization
        if (this._initialized) {
            throw new Error("Cannot change email sender after initialization");
        }
 
        this.emailSender = sender;
    }
 
    // Setter with business logic validation
    public setMaxRetries(retries: number): void {
        if (retries < 0) {
            throw new Error("Max retries cannot be negative");
        }
        if (retries > 10) {
            throw new Error("Max retries cannot exceed 10 (safety limit)");
        }
        
        this.maxRetries = retries;
    }
 
    // Lifecycle method that locks configuration
    public initialize(): void {
        if (!this.emailSender) {
            throw new Error("Email sender must be configured before initialization");
        }
        
        this._initialized = true;
    }
}
 
// Usage with validation
const service = new NotificationService();
 
try {
    service.setEmailSender(null!); // Throws immediately
} catch (e) {
    console.error("Caught: " + e.message);
}
 
service.setEmailSender(new SmtpEmailSender());
service.setMaxRetries(5);
service.initialize();
 
try {
    service.setEmailSender(new DifferentSender()); // Throws - already initialized
} catch (e) {
    console.error("Caught: " + e.message);
}

Pattern 4: Property-Based Injection (Language-Specific)

Some languages provide property syntax that offers a cleaner alternative to explicit setters:

PropertyBasedInjection.cs
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
// Pattern 4: Property-Based Injection (C#)
// Pros: Idiomatic, can add validation in setter
// Cons: Harder to distinguish injection points from regular properties
 
public class NotificationService
{
    private IEmailSender? _emailSender;
    private ILogger? _logger;
 
    // Property with simple setter - common for optional dependencies
    public ILogger? Logger { get; set; }
 
    // Property with validation in setter
    public IEmailSender? EmailSender
    {
        get => _emailSender;
        set
        {
            if (value == null)
                throw new ArgumentNullException(nameof(value));
            
            if (_emailSender != null)
                throw new InvalidOperationException("Email sender already configured");
            
            _emailSender = value;
        }
    }
    
    // Required property (C# 11+) - must be set before use
    public required IConfiguration Configuration { get; set; }
}
 
// Usage
var service = new NotificationService
{
    Configuration = config,  // Required - must be set
    Logger = logger,         // Optional
    EmailSender = sender     // Optional with validation
};

Choosing the Right Setter Pattern

•Simple setters when working with DI frameworks that expect standard Java Bean conventions
•Fluent setters when configuring objects manually and readability is paramount
•Validating setters when invalid configuration would cause hard-to-debug runtime failures
•Property-based injection when the language supports it and you want idiomatic code

Handling the Null State: Defensive Programming for Setter Injection

The defining characteristic of setter injection is that dependencies might not be present when methods are called. This creates a fundamental design decision: How should the class behave when a setter-injected dependency is null?

There are four primary strategies, each with distinct trade-offs:

Strategy 1: Throw an Exception

Fail loudly and immediately when a required dependency is missing. This is the fail-fast approach—bugs are discovered early rather than manifesting as subtle misbehavior.

When to Use:

The dependency is required for correct operation
The null state represents a programming error, not a valid configuration
You want maximum visibility into configuration mistakes

TypeScript
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
class PaymentProcessor {
    private paymentGateway: IPaymentGateway | null = null;
 
    setPaymentGateway(gateway: IPaymentGateway): void {
        this.paymentGateway = gateway;
    }
 
    processPayment(amount: number): PaymentResult {
        // Fail-fast: Cannot process payments without a gateway
        if (!this.paymentGateway) {
            throw new Error(
                "PaymentProcessor not properly configured: " +
                "setPaymentGateway() must be called before processPayment()"
            );
        }
 
        return this.paymentGateway.charge(amount);
    }
}
 
// This pattern makes misconfiguration obvious
const processor = new PaymentProcessor();
processor.processPayment(100); // Throws immediately with clear message

Initialization Lifecycle Patterns

One of the challenges with setter injection is ensuring all required dependencies are injected before the object is used. Several patterns address this:

The Initialization Method Pattern

Add an explicit initialize() method that validates configuration and transitions the object to a ready state:

InitializationMethodPattern.ts
TypeScript
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
enum ServiceState {
    UNCONFIGURED = "unconfigured",
    CONFIGURED = "configured",
    INITIALIZED = "initialized",
    DISPOSED = "disposed"
}
 
class DatabaseConnectionPool {
    private state: ServiceState = ServiceState.UNCONFIGURED;
    private connectionString: string | null = null;
    private maxConnections: number = 10;
    private connections: Connection[] = [];
 
    // Setters for configuration
    setConnectionString(connectionString: string): void {
        this.validateState([ServiceState.UNCONFIGURED, ServiceState.CONFIGURED]);
        this.connectionString = connectionString;
        this.state = ServiceState.CONFIGURED;
    }
 
    setMaxConnections(max: number): void {
        this.validateState([ServiceState.UNCONFIGURED, ServiceState.CONFIGURED]);
        if (max < 1 || max > 100) {
            throw new Error("Max connections must be between 1 and 100");
        }
        this.maxConnections = max;
        this.state = ServiceState.CONFIGURED;
    }
 
    // Explicit initialization - validates and activates
    async initialize(): Promise<void> {
        this.validateState([ServiceState.CONFIGURED]);
 
        if (!this.connectionString) {
            throw new Error("Connection string is required");
        }
 
        // Actually create the pool
        for (let i = 0; i < this.maxConnections; i++) {
            const conn = await this.createConnection(this.connectionString);
            this.connections.push(conn);
        }
 
        this.state = ServiceState.INITIALIZED;
    }
 
    // Public methods check for initialized state
    getConnection(): Connection {
        this.validateState([ServiceState.INITIALIZED]);
        // Return a connection from the pool
        return this.connections.pop() || this.createConnection(this.connectionString!);
    }
 
    releaseConnection(conn: Connection): void {
        this.validateState([ServiceState.INITIALIZED]);
        this.connections.push(conn);
    }
 
    // Cleanup
    async dispose(): Promise<void> {
        for (const conn of this.connections) {
            await conn.close();
        }
        this.connections = [];
        this.state = ServiceState.DISPOSED;
    }
 
    // State validation helper
    private validateState(allowedStates: ServiceState[]): void {
        if (!allowedStates.includes(this.state)) {
            throw new Error(
                `Invalid state: ${this.state}. ` +
                                        `Expected one of: ${allowedStates.join(", ")}`
            );
        }
    }
 
    private async createConnection(connectionString: string): Promise<Connection> {
        // Create actual connection
        return new Connection(connectionString);
    }
}
 
// Usage with lifecycle
const pool = new DatabaseConnectionPool();
 
// Configure (can be in any order)
pool.setMaxConnections(20);
pool.setConnectionString("postgresql://localhost/mydb");
 
// Must initialize before use
await pool.initialize();
 
// Now safe to use
const conn = pool.getConnection();
// ... use connection ...
pool.releaseConnection(conn);
 
// Cleanup when done
await pool.dispose();

Framework Integration

Many DI frameworks support lifecycle callbacks like @PostConstruct (Java), or IInitializable/IDisposable (C#). These hooks call your initialization method automatically after all setters have been invoked, bridging the gap between setter injection flexibility and initialization safety.

Summary: The Mechanics of Setter Injection

We've explored the foundational mechanics of setter injection—how dependencies are provided through mutator methods after object construction. Let's consolidate the key insights:

Key Takeaways

•Setter injection uses mutator methods rather than constructor parameters to provide dependencies after object creation
•Two-phase lifecycle — objects are first constructed, then configured, which enables flexibility but requires careful state management
•Weak invariants — unlike constructor injection, setter-injected dependencies may be null, requiring explicit null handling throughout the class
•Multiple setter patterns exist — from simple setters to fluent builders to validating setters with lifecycle enforcement
•Four null-handling strategies — throw exceptions, skip silently, use defaults, or alter behavior based on configuration
•Initialization methods can bridge the gap, validating configuration before transitioning to a usable state

What's Next:

Now that we understand the mechanics of setter injection, the next page explores its primary use case: optional dependencies. We'll examine when dependencies truly are optional, how to design for graceful degradation, and the patterns that make optional dependency injection clean and maintainable.

Page Complete

You now understand how setter injection works mechanically—the method signatures, the field storage, and the null-state handling. Next, we'll explore the Optional Dependencies Pattern, where setter injection truly shines.

1 / 4

Loading learning content...

System Design (LLD)Setter Injection

Setter Injection: Providing Dependencies Through Mutator Methods

LevelIntermediate

Duration50 mins

TopicSetter Injection

1 / 4

Passing Dependencies Through Setters

Beyond the Constructor: An Alternative Path to Dependency Wiring

What You Will Learn

The Mechanics of Setter Injection

The core pattern involves three elements:

A private field to hold the dependency reference
A setter method that accepts the dependency and assigns it to the field
Usage code that checks for the dependency's presence before use

Let's examine the simplest possible implementation:

EmailService.ts
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
// Basic setter injection pattern
 
interface ILogger {
    log(message: string): void;
}
 
class EmailService {
    // Private field to hold the optional dependency
    private logger: ILogger | null = null;
 
    // Setter method - the injection point
    public setLogger(logger: ILogger): void {
        this.logger = logger;
    }
 
    public sendEmail(to: string, subject: string, body: string): void {
        // Check if logger is available before using
        if (this.logger) {
            this.logger.log(`Sending email to ${to}: ${subject}`);
        }
 
        // Core email sending logic
        this.performEmailDelivery(to, subject, body);
 
        if (this.logger) {
            this.logger.log(`Email sent successfully to ${to}`);
        }
    }
 
    private performEmailDelivery(to: string, subject: string, body: string): void {
        // Actual email sending implementation
        console.log(`Delivering email to ${to}`);
    }
}
 
// Usage: Dependency is injected AFTER construction
const emailService = new EmailService();
 
// The service works without a logger (graceful degradation)
emailService.sendEmail("user@example.com", "Hello", "World");
 
// Later, inject a logger
const consoleLogger: ILogger = {
    log: (message) => console.log(`[LOG] ${message}`)
};
emailService.setLogger(consoleLogger);
 
// Now the service logs its operations
emailService.sendEmail("user@example.com", "Hello Again", "World");

The Two-Phase Lifecycle

Architectural Comparison with Constructor Injection

Object Invariants and State Validity

Constructor Injection vs Setter Injection: Architectural Comparison
Aspect	Constructor Injection	Setter Injection
Dependency Availability	Guaranteed from creation	Must be null-checked before use
Object State	Always valid	Potentially incomplete
Immutability	Dependencies can be final/readonly	Dependencies are mutable by design
Testing	Mock at construction	Mock via setter before test
Framework Integration	May require complex factories	Simpler framework wiring
Circular Dependencies	Causes construction deadlock	Can be resolved post-construction
Configuration Ordering	Implicit (construction order)	Explicit (setter call order matters)
API Clarity	Dependencies visible in constructor	Dependencies scattered across setters
Fail-Fast	Fails immediately on missing dependency	May fail later at runtime

Visualizing the Difference: Object Creation Flow

Consider how objects come into existence with each approach:

Constructor Injection Flow:

[Resolve Dependencies] → [Create Object with Dependencies] → [Ready to Use]

Setter Injection Flow:

[Create Empty Object] → [Call Setter 1] → [Call Setter 2] → ... → [Ready to Use]

ObjectLifecycleComparison.ts
TypeScript
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
// CONSTRUCTOR INJECTION: Strong invariant
 
interface IDatabase {
    query(sql: string): any[];
}
 
interface ICache {
    get(key: string): any;
    set(key: string, value: any): void;
}
 
class UserRepositoryConstructor {
    private readonly database: IDatabase;
    private readonly cache: ICache;
 
    // Object cannot exist without these dependencies
    constructor(database: IDatabase, cache: ICache) {
        if (!database) throw new Error("database is required");
        if (!cache) throw new Error("cache is required");
        this.database = database;
        this.cache = cache;
    }
 
    // Methods can assume dependencies exist - no null checks needed
    findUser(id: string): any {
        const cached = this.cache.get(id);
        if (cached) return cached;
        
        const user = this.database.query(`SELECT * FROM users WHERE id = '${id}'`)[0];
        this.cache.set(id, user);
        return user;
    }
}
 
// SETTER INJECTION: Weak invariant
 
class UserRepositorySetter {
    private database: IDatabase | null = null;
    private cache: ICache | null = null;
 
    // Object exists without dependencies
    constructor() {}
 
    setDatabase(database: IDatabase): void {
        this.database = database;
    }
 
    setCache(cache: ICache): void {
        this.cache = cache;
    }
 
    // Methods must handle missing dependencies
    findUser(id: string): any {
        // Must check every time - defensive programming required
        if (!this.database) {
            throw new Error("Database not configured - call setDatabase first");
        }
 
        // Cache is optional - check before use
        if (this.cache) {
            const cached = this.cache.get(id);
            if (cached) return cached;
        }
 
        const user = this.database.query(`SELECT * FROM users WHERE id = '${id}'`)[0];
        
        // Only cache if cache is available
        if (this.cache) {
            this.cache.set(id, user);
        }
        
        return user;
    }
}
 
// Usage comparison
 
// Constructor: One-shot, guaranteed complete
const repoConstructor = new UserRepositoryConstructor(
    mockDatabase,
    mockCache
);
// Immediately usable
repoConstructor.findUser("123");
 
// Setter: Multi-step, must verify completeness
const repoSetter = new UserRepositorySetter();
// NOT usable yet - would throw error
repoSetter.setDatabase(mockDatabase);
// Still missing cache, but might work (cache is optional)
repoSetter.setCache(mockCache);
// Now fully configured
repoSetter.findUser("123");

The Hidden Cost of Weak Invariants

Method Signature Patterns for Setter Injection

Pattern 1: Simple Setter

The most basic pattern—a void method that accepts and stores the dependency:

SimpleSetterPattern.ts
TypeScript
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
// Pattern 1: Simple Setter
// Pros: Simple, familiar, works with most frameworks
// Cons: No validation, no fluent chaining
 
class NotificationService {
    private emailSender: IEmailSender | null = null;
    private smsSender: ISmsSender | null = null;
 
    // Simple setter - just stores the reference
    public setEmailSender(sender: IEmailSender): void {
        this.emailSender = sender;
    }
 
    public setSmsSender(sender: ISmsSender): void {
        this.smsSender = sender;
    }
}
 
// Usage
const service = new NotificationService();
service.setEmailSender(new SmtpEmailSender());
service.setSmsSender(new TwilioSmsSender());

Pattern 2: Fluent Setter (Method Chaining)

Return this from setters to enable fluent configuration:

FluentSetterPattern.ts
TypeScript
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
// Pattern 2: Fluent Setter (Method Chaining)
// Pros: Cleaner configuration code, self-documenting
// Cons: May not be recognized by all DI frameworks
 
class NotificationService {
    private emailSender: IEmailSender | null = null;
    private smsSender: ISmsSender | null = null;
    private logger: ILogger | null = null;
 
    // Fluent setters return 'this' for chaining
    public setEmailSender(sender: IEmailSender): this {
        this.emailSender = sender;
        return this;
    }
 
    public setSmsSender(sender: ISmsSender): this {
        this.smsSender = sender;
        return this;
    }
 
    public setLogger(logger: ILogger): this {
        this.logger = logger;
        return this;
    }
}
 
// Fluent usage - reads like a configuration DSL
const service = new NotificationService()
    .setEmailSender(new SmtpEmailSender())
    .setSmsSender(new TwilioSmsSender())
    .setLogger(new ConsoleLogger());
 
// Compare to non-fluent equivalent:
const service2 = new NotificationService();
service2.setEmailSender(new SmtpEmailSender());
service2.setSmsSender(new TwilioSmsSender());
service2.setLogger(new ConsoleLogger());

Pattern 3: Validating Setter

Add validation logic to prevent invalid configurations:

ValidatingSetterPattern.ts
TypeScript
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
// Pattern 3: Validating Setter
// Pros: Fail-fast on invalid configuration, clearer error messages
// Cons: More code, may throw during framework wiring
 
class NotificationService {
    private emailSender: IEmailSender | null = null;
    private maxRetries: number = 3;
    private _initialized: boolean = false;
 
    // Validating setter - checks preconditions
    public setEmailSender(sender: IEmailSender): void {
        if (!sender) {
            throw new Error("Email sender cannot be null");
        }
        
        // Prevent reconfiguration after initialization
        if (this._initialized) {
            throw new Error("Cannot change email sender after initialization");
        }
 
        this.emailSender = sender;
    }
 
    // Setter with business logic validation
    public setMaxRetries(retries: number): void {
        if (retries < 0) {
            throw new Error("Max retries cannot be negative");
        }
        if (retries > 10) {
            throw new Error("Max retries cannot exceed 10 (safety limit)");
        }
        
        this.maxRetries = retries;
    }
 
    // Lifecycle method that locks configuration
    public initialize(): void {
        if (!this.emailSender) {
            throw new Error("Email sender must be configured before initialization");
        }
        
        this._initialized = true;
    }
}
 
// Usage with validation
const service = new NotificationService();
 
try {
    service.setEmailSender(null!); // Throws immediately
} catch (e) {
    console.error("Caught: " + e.message);
}
 
service.setEmailSender(new SmtpEmailSender());
service.setMaxRetries(5);
service.initialize();
 
try {
    service.setEmailSender(new DifferentSender()); // Throws - already initialized
} catch (e) {
    console.error("Caught: " + e.message);
}

Pattern 4: Property-Based Injection (Language-Specific)

Some languages provide property syntax that offers a cleaner alternative to explicit setters:

PropertyBasedInjection.cs
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
// Pattern 4: Property-Based Injection (C#)
// Pros: Idiomatic, can add validation in setter
// Cons: Harder to distinguish injection points from regular properties
 
public class NotificationService
{
    private IEmailSender? _emailSender;
    private ILogger? _logger;
 
    // Property with simple setter - common for optional dependencies
    public ILogger? Logger { get; set; }
 
    // Property with validation in setter
    public IEmailSender? EmailSender
    {
        get => _emailSender;
        set
        {
            if (value == null)
                throw new ArgumentNullException(nameof(value));
            
            if (_emailSender != null)
                throw new InvalidOperationException("Email sender already configured");
            
            _emailSender = value;
        }
    }
    
    // Required property (C# 11+) - must be set before use
    public required IConfiguration Configuration { get; set; }
}
 
// Usage
var service = new NotificationService
{
    Configuration = config,  // Required - must be set
    Logger = logger,         // Optional
    EmailSender = sender     // Optional with validation
};

Choosing the Right Setter Pattern

•Simple setters when working with DI frameworks that expect standard Java Bean conventions
•Fluent setters when configuring objects manually and readability is paramount
•Validating setters when invalid configuration would cause hard-to-debug runtime failures
•Property-based injection when the language supports it and you want idiomatic code

Handling the Null State: Defensive Programming for Setter Injection

There are four primary strategies, each with distinct trade-offs:

Strategy 1: Throw an Exception

Fail loudly and immediately when a required dependency is missing. This is the fail-fast approach—bugs are discovered early rather than manifesting as subtle misbehavior.

When to Use:

The dependency is required for correct operation
The null state represents a programming error, not a valid configuration
You want maximum visibility into configuration mistakes

TypeScript
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
class PaymentProcessor {
    private paymentGateway: IPaymentGateway | null = null;
 
    setPaymentGateway(gateway: IPaymentGateway): void {
        this.paymentGateway = gateway;
    }
 
    processPayment(amount: number): PaymentResult {
        // Fail-fast: Cannot process payments without a gateway
        if (!this.paymentGateway) {
            throw new Error(
                "PaymentProcessor not properly configured: " +
                "setPaymentGateway() must be called before processPayment()"
            );
        }
 
        return this.paymentGateway.charge(amount);
    }
}
 
// This pattern makes misconfiguration obvious
const processor = new PaymentProcessor();
processor.processPayment(100); // Throws immediately with clear message

Initialization Lifecycle Patterns

One of the challenges with setter injection is ensuring all required dependencies are injected before the object is used. Several patterns address this:

The Initialization Method Pattern

Add an explicit initialize() method that validates configuration and transitions the object to a ready state:

InitializationMethodPattern.ts
TypeScript
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
enum ServiceState {
    UNCONFIGURED = "unconfigured",
    CONFIGURED = "configured",
    INITIALIZED = "initialized",
    DISPOSED = "disposed"
}
 
class DatabaseConnectionPool {
    private state: ServiceState = ServiceState.UNCONFIGURED;
    private connectionString: string | null = null;
    private maxConnections: number = 10;
    private connections: Connection[] = [];
 
    // Setters for configuration
    setConnectionString(connectionString: string): void {
        this.validateState([ServiceState.UNCONFIGURED, ServiceState.CONFIGURED]);
        this.connectionString = connectionString;
        this.state = ServiceState.CONFIGURED;
    }
 
    setMaxConnections(max: number): void {
        this.validateState([ServiceState.UNCONFIGURED, ServiceState.CONFIGURED]);
        if (max < 1 || max > 100) {
            throw new Error("Max connections must be between 1 and 100");
        }
        this.maxConnections = max;
        this.state = ServiceState.CONFIGURED;
    }
 
    // Explicit initialization - validates and activates
    async initialize(): Promise<void> {
        this.validateState([ServiceState.CONFIGURED]);
 
        if (!this.connectionString) {
            throw new Error("Connection string is required");
        }
 
        // Actually create the pool
        for (let i = 0; i < this.maxConnections; i++) {
            const conn = await this.createConnection(this.connectionString);
            this.connections.push(conn);
        }
 
        this.state = ServiceState.INITIALIZED;
    }
 
    // Public methods check for initialized state
    getConnection(): Connection {
        this.validateState([ServiceState.INITIALIZED]);
        // Return a connection from the pool
        return this.connections.pop() || this.createConnection(this.connectionString!);
    }
 
    releaseConnection(conn: Connection): void {
        this.validateState([ServiceState.INITIALIZED]);
        this.connections.push(conn);
    }
 
    // Cleanup
    async dispose(): Promise<void> {
        for (const conn of this.connections) {
            await conn.close();
        }
        this.connections = [];
        this.state = ServiceState.DISPOSED;
    }
 
    // State validation helper
    private validateState(allowedStates: ServiceState[]): void {
        if (!allowedStates.includes(this.state)) {
            throw new Error(
                `Invalid state: ${this.state}. ` +
                                        `Expected one of: ${allowedStates.join(", ")}`
            );
        }
    }
 
    private async createConnection(connectionString: string): Promise<Connection> {
        // Create actual connection
        return new Connection(connectionString);
    }
}
 
// Usage with lifecycle
const pool = new DatabaseConnectionPool();
 
// Configure (can be in any order)
pool.setMaxConnections(20);
pool.setConnectionString("postgresql://localhost/mydb");
 
// Must initialize before use
await pool.initialize();
 
// Now safe to use
const conn = pool.getConnection();
// ... use connection ...
pool.releaseConnection(conn);
 
// Cleanup when done
await pool.dispose();

Framework Integration

Summary: The Mechanics of Setter Injection

We've explored the foundational mechanics of setter injection—how dependencies are provided through mutator methods after object construction. Let's consolidate the key insights:

Key Takeaways

•Setter injection uses mutator methods rather than constructor parameters to provide dependencies after object creation
•Two-phase lifecycle — objects are first constructed, then configured, which enables flexibility but requires careful state management
•Weak invariants — unlike constructor injection, setter-injected dependencies may be null, requiring explicit null handling throughout the class
•Multiple setter patterns exist — from simple setters to fluent builders to validating setters with lifecycle enforcement
•Four null-handling strategies — throw exceptions, skip silently, use defaults, or alter behavior based on configuration
•Initialization methods can bridge the gap, validating configuration before transitioning to a usable state

What's Next:

Page Complete

1 / 4