Building a Nutritional Tracker: Part 3 - Data Validation with Zod

2025-11-10 2 min

Table of Contents

Part 3: Data Validation with Zod — Nutrition Tracker

Introduction

In the previous chapter, we set up the testing environment. Now it’s time to ensure our data is safe and correct: that is, validate everything before saving or processing.

Why use Zod for validation?

Allows you to define rules and types in one place.
Detects incorrect data at runtime before it causes errors.
Automatically generates TypeScript types from schemas (no duplication).
Validations and errors are clear and centralized.

Model structure

We start from our central model:

\text{Record} = (id, userId, userName, food, amount, unit, date, time, mealType, sweetener, notes, createdAt)

Each field requires specific rules. For example:

amount must be a number greater than zero.
date and time must follow the correct format.
unit, mealType, and sweetener must be one of the values listed in their enum.

Basic Zod schema

import { z } from 'zod'

export const RegisterSchema = z.object({
  id: z.string().uuid(),
  userId: z.string().uuid(),
  userName: z.string().min(1),
  food: z.string().min(1),
  amount: z.number().positive(),
  unit: z.enum([
    'gr',
    'ml',
    'unit',
    'portion',
    'small-portion',
    'large-portion',
  ]),
  date: z.string().regex(/^\d{4}-\d{2}-\d{2}$/), // YYYY-MM-DD
  time: z.string().regex(/^\d{2}:\d{2}$/), // HH:MM
  mealType: z.enum(['breakfast', 'lunch', 'snack', 'dinner', 'collation']),
  createdAt: z.string().regex(/^\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}Z$/),
  sweetener: z.enum(['sugar', 'sweetener']).optional().nullable(),
  notes: z.string().optional(),
})

Automatic type inference

export type Register = z.infer<typeof RegisterSchema>

This means TypeScript understands your data types directly from the schema. If you change the schema, the type updates effortlessly.

Example: Practical validation

Suppose you have user input data:

const data = {
  id: 'not-a-uuid',
  userId: '1234',
  userName: '',
  food: '',
  amount: -5,
  unit: 'litros',
  date: '11-2025-11',
  time: '8:30am',
  mealType: 'brunch',
  createdAt: '2025-11-11',
}

When validating with Zod (using .safeParse()):

const result = RegisterSchema.safeParse(data)

if (!result.success) {
  // You can iterate over errors to show messages for each field
  result.error.issues.forEach((issue) => {
    console.log(`Field: ${issue.path[0]} - ${issue.message}`)
  })
}

Validation flow diagram

graph TD
  A[User data] --> B{RegisterSchema.safeParse}
  B -- valid --> C[Process normally]
  B -- invalid --> D[Report errors]

  D --> E[Display error messages in the form]

Additional advantages

You can use enums to keep menus and selects consistent.
Error messages generated by Zod are clear and easy to display in the UI.
It’s easy to extend the schema for new fields in the future.

Unit tests (practical example)

import { describe, expect, it } from 'vitest'

describe('RegisterSchema', () => {
  it('accepts valid record', () => {
    const valid = {
      id: 'a1b2c3d4-e5f6-7890-abcd-1234567890ab',
      userId: 'b2c3d4e5-f6a7-8901-bcde-2345678901bc',
      userName: 'Juan',
      food: 'Apple',
      amount: 1,
      unit: 'unit',
      date: '2025-11-11',
      time: '08:30',
      mealType: 'breakfast',
      createdAt: '2025-11-11T08:35:00Z',
      sweetener: null,
      notes: 'Fresh',
    }
    expect(() => RegisterSchema.parse(valid)).not.toThrow()
  })

  it('rejects negative amount', () => {
    const invalid = { ...valid, amount: -2 }
    expect(() => RegisterSchema.parse(invalid)).toThrow()
  })

  it('rejects incorrect date format', () => {
    const invalid = { ...valid, date: '11-11-2025' }
    expect(() => RegisterSchema.parse(invalid)).toThrow()
  })
})

What’s next?

You now have robust validation in your model. The next step is to create the persistence layer: save and retrieve validated records, handling errors and edge cases.

Continue reading: Part 4: Persistence Layer Implementation → How to save to localStorage, retrieve data, and maintain error tolerance.

Additional Resources

nutritional-tracker data-validation zod

Ramiro N. Cosa