惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

Y
Y Combinator Blog
GbyAI
GbyAI
U
Unit 42
WordPress大学
WordPress大学
Last Week in AI
Last Week in AI
P
Proofpoint News Feed
D
DataBreaches.Net
N
Netflix TechBlog - Medium
H
Hackread – Cybersecurity News, Data Breaches, AI and More
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
C
Check Point Blog
Martin Fowler
Martin Fowler
月光博客
月光博客
MongoDB | Blog
MongoDB | Blog
MyScale Blog
MyScale Blog
The Cloudflare Blog
Apple Machine Learning Research
Apple Machine Learning Research
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
M
MIT News - Artificial intelligence
云风的 BLOG
云风的 BLOG
罗磊的独立博客
B
Blog RSS Feed
J
Java Code Geeks
The GitHub Blog
The GitHub Blog

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant
Finishing a Payments Flow in NestJS: Events, Config, and ...
Rhico · 2026-04-28 · via DEV Community

I'm building RunHop in public, a social + event platform for running races. Today was a cleanup-and-integration day for the payments module.

The core problem wasn't creating payments. That part already existed. The real work was making payment review actually affect the rest of the system in a clean way.

The Product-Level Rule

There are two different registration paths:

  • Free race: registration should be confirmed immediately
  • Paid race: registration should stay pending until payment review is approved

That sounds simple, but it creates a design question: when a payment is approved, which service owns the registration state change?

Keep the State Transition with the Owning Context

My first instinct was the obvious one: if PaymentService.review() approves the payment, just update the registration there too.

That would work, but it would make the payment module responsible for registration lifecycle rules. I didn't want that coupling.

So I kept the payment review logic in PaymentService and emitted domain events:

this.eventEmitter.emit(NotificationEventTypes.PAYMENT_APPROVED, {
    paymentId: payment.id,
    registrationId: payment.registration.id,
});

Enter fullscreen mode Exit fullscreen mode

and

this.eventEmitter.emit(NotificationEventTypes.PAYMENT_REJECTED, {
    paymentId: payment.id,
    registrationId: payment.registration.id,
    rejectionCount: rejectionCount,
});

Enter fullscreen mode Exit fullscreen mode

Then RegistrationService listens:

@OnEvent(NotificationEventTypes.PAYMENT_APPROVED)
async handlePaymentApprovedEvent(event: {
    paymentId: string;
    registrationId: string;
}) {
    await this.prisma.registration.update({
        where: { id: event.registrationId },
        data: { status: 'CONFIRMED' },
    });
}

Enter fullscreen mode Exit fullscreen mode

and:

@OnEvent(NotificationEventTypes.PAYMENT_REJECTED)
async handlePaymentRejectedEvent(event: {
    paymentId: string;
    registrationId: string;
    rejectionCount: number;
}) {
    if (
        event.rejectionCount >=
        this.configService.get<number>('MAX_PAYMENT_ATTEMPTS', 3)
    ) {
        await this.prisma.registration.update({
            where: { id: event.registrationId },
            data: { status: 'CANCELLED' },
        });
    }
}

Enter fullscreen mode Exit fullscreen mode

That split turned out cleaner than I expected. Payment owns payment review. Registration owns registration status.

Free Races Should Skip the Pending State
I also adjusted registration creation to handle free races directly:

return await this.prisma.registration.create({
    data: {
        userId,
        raceId,
        ...(race.price === 0
            ? { status: RegistrationStatus.CONFIRMED }
            : {}),
    },
});

Enter fullscreen mode Exit fullscreen mode

This removed the need for a separate confirmation step for zero-price races. The service can decide the initial state up front from the race price.

P2002 Is the Real Duplicate Guard

One of the more useful reminders today was about duplicate registrations.

At the application level, it's tempting to do:

  1. check if the registration exists
  2. if not, create it
  3. That works in the happy path, but it's not enough under concurrency. Two requests can pass the pre-check before either insert commits.

The actual protection comes from the Prisma schema:

@@unique([userId, raceId])
and the create call needs to treat P2002 as the authoritative duplicate signal:

} catch (error) {
    if (
        error instanceof Prisma.PrismaClientKnownRequestError &&
        error.code === 'P2002'
    ) {
        throw new ConflictException('You are already registered for this race.');
    }

    throw error;
}

Enter fullscreen mode Exit fullscreen mode

That changed how I think about the flow.

Pre-checks improve messaging and readability
Unique constraints enforce reality
They're not interchangeable.

Config Instead of Magic Numbers

The payment module originally had:

const MAX_PAYMENT_ATTEMPTS = 3;
I moved that behind ConfigService so the service reads from env instead:

this.configService.get<number>('MAX_PAYMENT_ATTEMPTS', 3)
Small change, but it matters. Retry limits are a policy decision, and policy is usually config, not hardcoded behavior.

Endpoint Work

I also finished the controller side so the flow is reachable in the app:

GET /events/:eventId/payments
PATCH /payments/:id/review
GET /registrations/:id
The review path uses existing org membership checks so admins can review event payments without opening up the endpoint to everyone.

Verification

Fresh verification for the session:

npm run build passed
targeted unit tests passed: 4 suites, 57 tests
e2e is partially blocked right now because the local test database on localhost:5433 was not running during the session
That last part matters. The code is in much better shape, but I don't want to blur “designed and unit-tested” with “fully e2e-verified.”

Takeaway

The useful lesson from today wasn't “how to add a payment review endpoint.” It was that a flow starts feeling solid when the owning modules each handle their own state transitions.

Payment review should not secretly become registration business logic.

Events were the clean line here.