Resolver unit tests in GraphQL - Step-by-Step Execution

Concept Flow - Resolver unit tests

Write Resolver Function

↓

Write Unit Test for Resolver

↓

Mock Dependencies (DB, APIs)

↓

Run Test

↓

Check Output vs Expected

↓

Pass?

No→Fix Resolver or Test

Yes↓

Done

This flow shows writing a resolver, creating a test with mocks, running it, and checking if output matches expected results.

Execution Sample

GraphQL

const resolver = (parent, args, context) => {
  return context.db.getUser(args.id);
};

// Test
const mockDb = { getUser: (id) => ({ id, name: 'Alice' }) };
const result = resolver(null, { id: 1 }, { db: mockDb });

This code defines a resolver that fetches a user by id from a mocked database and runs a test to check the returned user.

Execution Table

Step	Action	Input	Mock Used	Output	Check
1	Call resolver	args.id=1	mockDb.getUser	Calls mockDb.getUser(1)	N/A
2	mockDb.getUser called	id=1	N/A	{ id: 1, name: 'Alice' }	N/A
3	Resolver returns	N/A	N/A	{ id: 1, name: 'Alice' }	Output matches expected user object
4	Test assertion	Output	N/A	Pass	Output equals expected result

💡 Test passes because resolver output matches expected mocked user data

Variable Tracker

Variable	Start	After Call	Final
args	{ id: 1 }	{ id: 1 }	{ id: 1 }
context.db	mockDb object	mockDb object	mockDb object
result	undefined	{ id: 1, name: 'Alice' }	{ id: 1, name: 'Alice' }

Key Moments - 2 Insights

Why do we use a mock database instead of the real one in tests?

Using a mock database isolates the resolver logic and ensures tests run fast and reliably without depending on real database state, as shown in execution_table step 2.

What happens if the resolver returns data different from the expected output?

The test assertion in execution_table step 4 would fail, indicating a bug in the resolver or test setup that needs fixing.

Visual Quiz - 3 Questions

Look at the execution table, what is the output after the resolver calls mockDb.getUser?

A { id: 1, name: 'Alice' }

B undefined

C null

D { id: 2, name: 'Bob' }

💡 Hint

Check execution_table row 2 under Output column

At which step does the test confirm the output matches the expected result?

A Step 3

B Step 4

C Step 1

D Step 2

💡 Hint

Look at execution_table row 4 under Check column

If the mockDb.getUser returned null, how would the test result change?

A Test would pass as usual

B Test would error out immediately

C Test would fail because output differs from expected

D Test would skip assertion

💡 Hint

Consider the importance of matching output in execution_table step 4

Concept Snapshot

Resolver unit tests:
- Write resolver function accessing data
- Mock dependencies like databases
- Run resolver with test inputs
- Check output matches expected
- Fix code if test fails
- Ensures resolver logic correctness

Full Transcript

Resolver unit tests involve writing a resolver function that fetches data, then creating a test that calls this resolver with specific inputs. Instead of using a real database, a mock database is used to isolate the resolver logic and keep tests fast and reliable. The test runs the resolver, captures the output, and compares it to the expected result. If the output matches, the test passes; otherwise, it fails and requires fixing the resolver or test. This process ensures the resolver works correctly in isolation.

Practice

1. What is the main purpose of a resolver unit test in GraphQL?

easy

A. To verify the database connection settings

B. To test the entire GraphQL schema at once

C. To check if a resolver returns the correct data

D. To style the GraphQL playground interface

Solution

1. Step 1: Understand resolver role

   Resolvers are functions that fetch and return data for GraphQL queries.

2. Step 2: Purpose of unit tests

   Unit tests check small parts of code, here specifically if resolvers return correct data.

3. Final Answer:

   To check if a resolver returns the correct data -> Option C

4. Quick Check:

   Resolver unit tests = check resolver output [OK]

Hint: Resolvers return data; tests check if data is correct [OK]

Common Mistakes:

• Confusing unit tests with integration tests
• Thinking tests check UI or styling
• Assuming tests check database setup

2. Which syntax correctly defines a simple resolver unit test using Jest?

easy

A. describe('Test', () => { it('checks resolver', () => { expect(resolver()).toBe(data); }); });

B. test('Test', () => { describe('checks resolver', () => { expect(resolver()).toBe(data); }); });

C. it('Test', () => { expect(resolver()).toBe(data); }); describe('checks resolver', () => {});

D. expect('Test', () => { it('checks resolver', () => { resolver(); }); });

Solution

1. Step 1: Identify Jest test structure

   Jest uses describe to group tests and it or test for individual tests.

2. Step 2: Check correct nesting and syntax

   describe('Test', () => { it('checks resolver', () => { expect(resolver()).toBe(data); }); }); correctly nests it inside describe and uses expect properly.

3. Final Answer:

   describe('Test', () => { it('checks resolver', () => { expect(resolver()).toBe(data); }); }); -> Option A

4. Quick Check:

   describe + it + expect = correct test syntax [OK]

Hint: Use describe for groups, it for tests, expect for checks [OK]

Common Mistakes:

• Swapping describe and it blocks
• Missing expect or incorrect nesting
• Using expect without a matcher

3. Given this resolver and test code, what will the test output be?

const resolver = () => ({ id: 1, name: 'Alice' });

describe('User resolver', () => {
  it('returns correct user', () => {
    expect(resolver()).toEqual({ id: 1, name: 'Alice' });
  });
});

medium

A. Test throws runtime error

B. Test fails due to wrong object

C. Syntax error in test code

D. Test passes successfully

Solution

1. Step 1: Analyze resolver output

   The resolver returns an object { id: 1, name: 'Alice' } exactly.

2. Step 2: Check test expectation

   The test expects the same object using toEqual, which compares object values deeply.

3. Final Answer:

   Test passes successfully -> Option D

4. Quick Check:

   Exact object match = test passes [OK]

Hint: toEqual checks deep equality; objects must match exactly [OK]

Common Mistakes:

• Confusing toBe with toEqual for objects
• Expecting test to fail with correct data
• Misreading object properties

4. Identify the error in this resolver unit test code:

describe('Test resolver', () => {
  it('returns data', () => {
    expect(resolver).toBe(data);
  });
});

medium

A. No error; code is correct

B. Missing parentheses to call resolver function

C. Wrong matcher; should use toEqual instead of toBe

D. Incorrect use of describe block

Solution

1. Step 1: Check resolver usage

   The test uses resolver without parentheses, so it tests the function itself, not its return value.

2. Step 2: Correct function call

   To test the returned data, the resolver must be called as resolver().

3. Final Answer:

   Missing parentheses to call resolver function -> Option B

4. Quick Check:

   Call resolver() to get data, not resolver [OK]

Hint: Call functions with () to test their output [OK]

Common Mistakes:

• Forgetting to call the resolver function
• Confusing toBe and toEqual for objects
• Misplacing describe and it blocks

5. You want to test a resolver that fetches a user by ID from a mock database. Which approach best ensures your unit test is isolated and reliable?

hard

A. Mock the database call inside the resolver test to return fixed data

B. Connect to the real database and fetch actual user data

C. Skip testing the resolver and test only the database separately

D. Write tests that depend on network availability

Solution

1. Step 1: Understand unit test isolation

   Unit tests should test one part only, without relying on external systems like databases.

2. Step 2: Use mocking for database calls

   Mocking replaces real database calls with fixed data, making tests fast and reliable.

3. Final Answer:

   Mock the database call inside the resolver test to return fixed data -> Option A

4. Quick Check:

   Mock external calls for isolated unit tests [OK]

Hint: Mock external dependencies to isolate resolver tests [OK]

Common Mistakes:

• Using real database in unit tests
• Skipping resolver tests entirely
• Writing flaky tests depending on network