Chapter 6: Functional Programming

"Functional programming imposes discipline upon assignment."

For New Learners

The Oldest Paradigm, Adopted Last

Functional programming is the oldest of the three paradigms — rooted in Alonzo Church's λ-calculus (1936) — yet the last to gain mainstream adoption. Its core idea: variables don't vary. Once a value is bound, it never changes. No assignment. No mutation.

This sounds restrictive. It is. But it eliminates an entire universe of bugs.

Why Immutability Matters for Architecture

Consider every race condition, concurrency bug, or unexpected state corruption you've debugged. What do they share?

Multiple threads competing to update the same mutable state.

If state never changes, there is nothing to compete over. Race conditions become impossible by design.

"All race conditions, deadlock conditions, and concurrent update problems are due to mutable variables."

Segregation of Mutability

Pure immutability is impractical — you eventually write to databases, accept input, publish events. The solution is segregation:

Keep the core of the system pure (no side effects)
Isolate mutation at the boundaries
Data crossing boundaries travels as immutable structures

Architectural Zone	Mutability State	Operational Behavior	Examples & Boundaries
Pure Functional Core	Strictly Immutable	Pure transformation functions ( $f(input) o output$ ), deterministic, zero side-effects.	Domain entities, business calculations, rule validators.
Mutable Boundary	Controlled Mutable	Handles side-effects, manages I/O, persists state, interacts with external network sockets.	Repository writes, REST controllers, Kafka consumers, DB connections.

Event Sourcing

Instead of storing current state (mutable), store the history of events (immutable). Current state is derived by replaying events.

Only appends — no updates or deletes
Complete audit trail preserved
State at any past moment is reproducible
CRUD becomes CR

Senior Deep Dive

Immutable Value Objects in Java

// Mutable — dangerous to share
public class Money {
    private BigDecimal amount;
    public void setAmount(BigDecimal a) { this.amount = a; }
}

// Immutable record — safe to share across threads and boundaries
public record Money(BigDecimal amount, Currency currency) {
    public Money add(Money other) {
        return new Money(this.amount.add(other.amount), this.currency);
    }
}

Functional Pipelines vs Mutable Loops

// Mutable — shared mutable list, harder to parallelize safely
List<OrderSummary> results = new ArrayList<>();
for (Order o : orders) {
    if (o.isActive()) results.add(new OrderSummary(o.getId(), o.getTotal()));
}

// Immutable pipeline — pure transformations, parallelizable
List<OrderSummary> results = orders.stream()
    .filter(Order::isActive)
    .map(o -> new OrderSummary(o.getId(), o.getTotal()))
    .toList();

Event Sourcing with Spring + Axon

// Immutable event — a fact that happened
public record OrderPlacedEvent(String orderId, List<LineItem> items, Instant at) {}

@Aggregate
public class Order {
    @AggregateIdentifier private String orderId;
    private OrderStatus status;

    @CommandHandler
    public Order(PlaceOrderCommand cmd) {
        apply(new OrderPlacedEvent(cmd.orderId(), cmd.items(), Instant.now()));
    }

    @EventSourcingHandler
    public void on(OrderPlacedEvent event) {
        this.orderId = event.orderId();
        this.status = OrderStatus.PLACED;
    }
}

The aggregate only publishes immutable events; state is always reconstructed from the event log — never stored directly.

The Architectural Segregation Pattern

Spring's @Transactional boundary is exactly the "mutable state fence":

Domain / Use Cases       ← Pure: no side effects, fully unit-testable
       ↕  (immutable DTOs / commands / events)
Adapters / Repositories  ← @Transactional: where mutation is committed
       ↕
Database / Message Bus   ← Where state actually persists

Everything inside the use case should be a pure function of its inputs. Spring infrastructure handles the commit.

Summary

Concept	Key Point
Immutability	No assignment → no race conditions, no concurrent update bugs
Segregation	Isolate mutation at boundaries; keep core logic pure
Event Sourcing	Store events (immutable facts), derive state by replay
Java tools	Records, stream pipelines, Axon Framework, CQRS
Architecture lesson	Immutable data crossing boundaries prevents hidden coupling and state corruption

🎓 For New Learners