Networking with URLSession, REST, JSON, GraphQL, and Concurrency

Why Networking Deserves Architecture

Networking is where many mobile apps become unreliable. Slow responses, invalid payloads, flaky connectivity, expired authentication, and poor retry strategy can make a product feel broken. Students should treat networking as a design area, not as a quick code snippet.

URLSession Basics

struct Product: Decodable, Identifiable {
    let id: Int
    let name: String
}

func fetchProducts() async throws -> [Product] {
    let url = URL(string: "https://example.com/api/products")!
    let (data, response) = try await URLSession.shared.data(from: url)
    guard let http = response as? HTTPURLResponse, http.statusCode == 200 else {
        throw URLError(.badServerResponse)
    }
    return try JSONDecoder().decode([Product].self, from: data)
}

Common API Concerns

• Request building and endpoint organization
• Authentication headers and token refresh
• Error mapping for user-friendly messages
• Timeouts and retry policies
• Pagination, caching, and deduplication
• Observability for network failures

REST, GraphQL, and Mobile Tradeoffs

REST is common and straightforward, but payloads may be over- or under-fetched. GraphQL gives clients control over the shape of requested data, but introduces schema tooling, caching complexity, and query-management concerns. Students should understand architectural tradeoffs instead of treating any style as universally better.

Example Service Layer

protocol ProductFetching {
    func fetchProducts() async throws -> [Product]
}

final class ProductAPI: ProductFetching {
    func fetchProducts() async throws -> [Product] {
        try await fetchProducts()
    }
}

Protocol-based abstraction makes testing easier because mock services can replace network calls in unit tests and previews.

Real-World Usage Snapshot

A travel booking app may call search APIs, pricing APIs, saved traveler APIs, loyalty APIs, and payment-intent APIs in one booking flow. Good networking structure is what keeps that experience responsive, debuggable, and resilient under failure conditions.