Node interface pattern in GraphQL - Step-by-Step Execution

Concept Flow - Node interface pattern

Client sends query with node ID

↓

Server receives node ID

↓

Resolve node by ID

↓

Find User

↓

Return User

↓

Return Node with __typename

↓

Client receives node

The Node interface pattern lets clients fetch any object by a global ID. The server resolves the ID to the correct type and returns the object with its type info.

Execution Sample

GraphQL

query {
  node(id: "user-1") {
    __typename
    id
    ... on User {
      name
    }
  }
}

This query asks for a node by ID 'user-1', requesting its type, id, and if it's a User, its name.

Execution Table

Step	Action	Input	Process	Output
1	Receive query	node(id: "user-1")	Parse query and extract ID	ID extracted: 'user-1'
2	Resolve node	'user-1'	Look up object by ID	Found User object with id 'user-1'
3	Determine type	User object	Set __typename to 'User'	__typename: 'User'
4	Fetch fields	User object	Retrieve id and name fields	id: 'user-1', name: 'Alice'
5	Return response	Collected data	Package data with __typename	{ __typename: 'User', id: 'user-1', name: 'Alice' }
6	End	Response sent	No more steps	Query complete

💡 Query ends after returning the resolved node with its type and requested fields.

Variable Tracker

Variable	Start	After Step 1	After Step 2	After Step 3	After Step 4	Final
queryID	null	"user-1"	"user-1"	"user-1"	"user-1"	"user-1"
resolvedNode	null	null	{type: 'User', id: 'user-1', name: 'Alice'}	{type: 'User', id: 'user-1', name: 'Alice'}	{type: 'User', id: 'user-1', name: 'Alice'}	{type: 'User', id: 'user-1', name: 'Alice'}
__typename	null	null	null	"User"	"User"	"User"
response	null	null	null	null	{ __typename: 'User', id: 'user-1', name: 'Alice' }	{ __typename: 'User', id: 'user-1', name: 'Alice' }

Key Moments - 3 Insights

Why do we need the __typename field in the response?

The __typename tells the client what type of object was returned (User, Post, etc.). This helps the client know which fields to expect and how to handle the data, as shown in step 3 of the execution_table.

What happens if the ID does not match any object?

If no object matches the ID, the resolve step (step 2) returns null, and the node field in the response will be null, indicating no object found.

How does the server know which type to return for a given ID?

The server uses the ID format or a lookup table to find the object and its type, as shown in step 2 where 'user-1' resolves to a User object.

Visual Quiz - 3 Questions

Look at the execution_table, what is the __typename value at step 3?

A null

B "User"

C "Post"

D "Node"

💡 Hint

Check the 'Determine type' action in step 3 of the execution_table.

At which step does the server fetch the specific fields like 'name' for the User?

A Step 4

B Step 3

C Step 2

D Step 5

💡 Hint

Look for the step labeled 'Fetch fields' in the execution_table.

If the query asked for a Post node instead of a User, which step would change?

A Step 1 - Receive query

B Step 3 - Determine type

C Step 2 - Resolve node

D Step 6 - End

💡 Hint

Refer to step 2 where the object is found by ID; the type depends on the resolved object.

Concept Snapshot

Node interface pattern in GraphQL:
- Use a global ID to fetch any object.
- Server resolves ID to specific type (User, Post, etc.).
- Response includes __typename for client to identify type.
- Enables flexible, unified querying of different object types.
- Commonly used for Relay-compliant APIs.

Full Transcript

The Node interface pattern allows clients to request any object by a global ID. The server receives the ID, looks up the object, determines its type, and returns the object with a __typename field. This helps clients know what type of object they received and what fields to expect. For example, a query asking for node(id: "user-1") returns a User object with its id and name. The execution steps include receiving the query, resolving the node by ID, determining the type, fetching requested fields, and returning the response. If the ID does not match any object, the node field returns null. This pattern simplifies fetching different types of objects through a single interface.

Practice

1. What is the main purpose of the Node interface in GraphQL?

easy

A. To provide a unique ID for all object types

B. To define custom mutations for each type

C. To restrict queries to only one type

D. To automatically generate database schemas

Solution

1. Step 1: Understand the Node interface role

   The Node interface is designed to give every object type a unique identifier.

2. Step 2: Identify its main use

   This unique ID allows fetching any object by ID in a single query.

3. Final Answer:

   To provide a unique ID for all object types -> Option A

4. Quick Check:

   Node interface = unique ID for all types [OK]

Hint: Node interface always provides unique IDs for all types [OK]

Common Mistakes:

• Thinking Node defines mutations
• Believing Node restricts queries to one type
• Assuming Node auto-generates database schemas

2. Which of the following is the correct way to declare the Node interface in GraphQL SDL?

easy

A. interface Node { id: Boolean! }

B. interface Node { id: String }

C. interface Node { id: Int! }

D. interface Node { id: ID! }

Solution

1. Step 1: Recall the Node interface ID type

   The Node interface requires an id field of type ID!, which is a non-null unique identifier.

2. Step 2: Check each option's ID type

   interface Node { id: ID! } uses ID!, which is correct. Others use wrong types like String, Int, or Boolean.

3. Final Answer:

   interface Node { id: ID! } -> Option D

4. Quick Check:

   ID field in Node = ID! type [OK]

Hint: Node interface ID must be non-null ID type [OK]

Common Mistakes:

• Using String or Int instead of ID type
• Making ID nullable (missing !)
• Using Boolean as ID type

3. Given this query using the Node interface:

query { node(id: "123") { id ... on User { name } ... on Post { title } } }

What fields will be returned if the node with ID "123" is a Post?

medium

A. { "id": "123", "name": "Alice" }

B. { "id": "123", "title": "GraphQL Basics" }

C. { "id": "123" }

D. Error: Cannot query fragment on Post

Solution

1. Step 1: Understand the query structure

   The query fetches a node by ID and requests the id field plus fragments for User and Post types.

2. Step 2: Determine the node type and returned fields

   If the node is a Post, the title field from the Post fragment is returned along with id. The User fragment is ignored.

3. Final Answer:

   { "id": "123", "title": "GraphQL Basics" } -> Option B

4. Quick Check:

   Node query returns fields for actual type fragment [OK]

Hint: Fragments return fields only for matching node type [OK]

Common Mistakes:

• Expecting fields from non-matching fragments
• Ignoring the id field
• Assuming query causes error

4. You wrote this schema snippet:

interface Node { id: ID! }
type User implements Node { id: ID name: String }

Why will this schema cause an error?

medium

A. User's id field must be non-null (ID!) to match Node interface

B. User type cannot implement Node interface

C. id field type must be String, not ID

D. User type must not have extra fields like name

Solution

1. Step 1: Compare interface and type field definitions

   The Node interface requires id as ID! (non-null). The User type declares id as ID (nullable).

2. Step 2: Understand GraphQL type compatibility rules

   Implementing types must match or be more strict. Here, User's id is less strict (nullable), causing an error.

3. Final Answer:

   User's id field must be non-null (ID!) to match Node interface -> Option A

4. Quick Check:

   Implementing type fields must match interface exactly [OK]

Hint: Implementing type fields must be equal or stricter than interface [OK]

Common Mistakes:

• Thinking User can't implement Node
• Changing id type to String
• Removing extra fields like name

5. You want to fetch a list of mixed objects (Users and Posts) by their IDs using the Node interface. Which approach correctly fetches their specific fields in one query?

hard

A. Use a union type instead of Node interface

B. Query Users and Posts separately with two queries

C. Use a nodes(ids: [ID!]!) query returning [Node], then use inline fragments for User and Post fields

D. Fetch only the id field without fragments

Solution

1. Step 1: Understand Node interface usage for mixed types

   The Node interface allows fetching any object by ID in one query, returning a list of Nodes.

2. Step 2: Use inline fragments to get type-specific fields

   To get fields specific to Users and Posts, use inline fragments ... on User and ... on Post inside the query.

3. Final Answer:

   Use a nodes(ids: [ID!]!) query returning [Node], then use inline fragments for User and Post fields -> Option C

4. Quick Check:

   Node interface + fragments fetch mixed types in one query [OK]

Hint: Use nodes query with fragments for mixed type fields [OK]

Common Mistakes:

• Querying types separately instead of one query
• Using union instead of interface for this pattern
• Fetching only id without needed fields