Implementing the Relay Connection Specification

We naively implemented cursor based pagination in the previous tutorial . To ensure a consistent implementation of this pattern, the Relay project has a formal specification you can follow for building GraphQL APIs which use a cursor based connection pattern.

By the end of this tutorial, we should be able to return a connection of users when requested.

query getUsers {
getUsers(first: 2) {
users {
edges {
node {
id
name
occupation
age
}
}
cursor
}
pageInfo {
endCursor
hasNextPage
}
}
}
{
"data": {
"getUsers": {
"users": {
"edges": [
{
"node": {
"id": 1,
"name": "Norman Osborn",
"occupation": "Founder, Oscorp Industries",
"age": 42
},
"cursor": "dXNlcjox"
},
{
"node": {
"id": 2,
"name": "Peter Parker",
"occupation": "Freelance Photographer, The Daily Bugle",
"age": 20
},
"cursor": "dXNlcjoy"
}
]
},
"pageInfo": {
"endCursor": "dXNlcjoz",
"hasNextPage": true
}
}
}
}

Connections

A Connection represents a paginated relationship between two entities. This pattern is used when the relationship itself has attributes. For example, we might have a connection of users to represent a paginated list of users.

Let us define a Connection type which takes in a Generic ObjectType.

example.py
from typing import Generic, TypeVar
import strawberry
GenericType = TypeVar("GenericType")
@strawberry.type
class Connection(Generic[GenericType]):
page_info: "PageInfo" = strawberry.field(
description="Information to aid in pagination."
)
edges: list["Edge[GenericType]"] = strawberry.field(
description="A list of edges in this connection."
)

Connections must have atleast two fields: edges and page_info .

The page_info field contains metadata about the connection. Following the Relay specification, we can define a PageInfo type like this:

example.py
from typing import Generic, TypeVar
import strawberry
GenericType = TypeVar("GenericType")
@strawberry.type
class Connection(Generic[GenericType]):
page_info: "PageInfo" = strawberry.field(
description="Information to aid in pagination."
)
edges: list["Edge[GenericType]"] = strawberry.field(
description="A list of edges in this connection."
)
@strawberry.type
class PageInfo:
has_next_page: bool = strawberry.field(
description="When paginating forwards, are there more items?"
)
has_previous_page: bool = strawberry.field(
description="When paginating backwards, are there more items?"
)
start_cursor: Optional[str] = strawberry.field(
description="When paginating backwards, the cursor to continue."
)
end_cursor: Optional[str] = strawberry.field(
description="When paginating forwards, the cursor to continue."
)

You can read more about the PageInfo type at:

The edges field must return a list type that wraps an edge type.

Following the Relay specification, let us define an Edge that takes in a generic ObjectType.

example.py
from typing import Generic, TypeVar
import strawberry
GenericType = TypeVar("GenericType")
@strawberry.type
class Connection(Generic[GenericType]):
page_info: "PageInfo" = strawberry.field(
description="Information to aid in pagination."
)
edges: list["Edge[GenericType]"] = strawberry.field(
description="A list of edges in this connection."
)
@strawberry.type
class PageInfo:
has_next_page: bool = strawberry.field(
description="When paginating forwards, are there more items?"
)
has_previous_page: bool = strawberry.field(
description="When paginating backwards, are there more items?"
)
start_cursor: Optional[str] = strawberry.field(
description="When paginating backwards, the cursor to continue."
)
end_cursor: Optional[str] = strawberry.field(
description="When paginating forwards, the cursor to continue."
)
@strawberry.type
class Edge(Generic[GenericType]):
node: GenericType = strawberry.field(description="The item at the end of the edge.")
cursor: str = strawberry.field(description="A cursor for use in pagination.")

EdgeTypes must have atleast two fields - cursor and node . Each edge has it’s own cursor and item (represented by the node field).

Now that we have the types needed to implement pagination using Relay Connections, let us use them to paginate a list of users. For simplicity’s sake, let our dataset be a list of dictionaries.

example.py
from typing import Generic, TypeVar
import strawberry
user_data = [
{
"id": 1,
"name": "Norman Osborn",
"occupation": "Founder, Oscorp Industries",
"age": 42,
},
{
"id": 2,
"name": "Peter Parker",
"occupation": "Freelance Photographer, The Daily Bugle",
"age": 20,
},
{
"id": 3,
"name": "Harold Osborn",
"occupation": "President, Oscorp Industries",
"age": 19,
},
{
"id": 4,
"name": "Eddie Brock",
"occupation": "Journalist, The Eddie Brock Report",
"age": 20,
},
]
GenericType = TypeVar("GenericType")
@strawberry.type
class Connection(Generic[GenericType]):
page_info: "PageInfo" = strawberry.field(
description="Information to aid in pagination."
)
edges: list["Edge[GenericType]"] = strawberry.field(
description="A list of edges in this connection."
)
@strawberry.type
class PageInfo:
has_next_page: bool = strawberry.field(
description="When paginating forwards, are there more items?"
)
has_previous_page: bool = strawberry.field(
description="When paginating backwards, are there more items?"
)
start_cursor: Optional[str] = strawberry.field(
description="When paginating backwards, the cursor to continue."
)
end_cursor: Optional[str] = strawberry.field(
description="When paginating forwards, the cursor to continue."
)
@strawberry.type
class Edge(Generic[GenericType]):
node: GenericType = strawberry.field(description="The item at the end of the edge.")
cursor: str = strawberry.field(description="A cursor for use in pagination.")

Now is a good time to think of what we could use as a cursor for our dataset. Our cursor needs to be an opaque value, which doesn’t usually change over time. It makes sense to use base64 encoded IDs of users as our cursor, as they fit both criteria.

Tip

While working with Connections, it is a convention to base64-encode cursors. It provides a unified interface to the end user. API clients need not bother about the type of data to paginate, and can pass unique IDs during pagination. It also makes the cursors opaque.

Let us define a couple of helper functions to encode and decode cursors as follows:

example.py
from base64 import b64encode, b64decode
from typing import Generic, TypeVar
import strawberry
user_data = [
{
"id": 1,
"name": "Norman Osborn",
"occupation": "Founder, Oscorp Industries",
"age": 42,
},
{
"id": 2,
"name": "Peter Parker",
"occupation": "Freelance Photographer, The Daily Bugle",
"age": 20,
},
{
"id": 3,
"name": "Harold Osborn",
"occupation": "President, Oscorp Industries",
"age": 19,
},
{
"id": 4,
"name": "Eddie Brock",
"occupation": "Journalist, The Eddie Brock Report",
"age": 20,
},
]
def encode_user_cursor(id: int) -> str:
"""
Encodes the given user ID into a cursor.
:param id: The user ID to encode.
:return: The encoded cursor.
"""
return b64encode(f"user:{id}".encode("ascii")).decode("ascii")
def decode_user_cursor(cursor: str) -> int:
"""
Decodes the user ID from the given cursor.
:param cursor: The cursor to decode.
:return: The decoded user ID.
"""
cursor_data = b64decode(cursor.encode("ascii")).decode("ascii")
return int(cursor_data.split(":")[1])
GenericType = TypeVar("GenericType")
@strawberry.type
class Connection(Generic[GenericType]):
page_info: "PageInfo" = strawberry.field(
description="Information to aid in pagination."
)
edges: list["Edge[GenericType]"] = strawberry.field(
description="A list of edges in this connection."
)
@strawberry.type
class PageInfo:
has_next_page: bool = strawberry.field(
description="When paginating forwards, are there more items?"
)
has_previous_page: bool = strawberry.field(
description="When paginating backwards, are there more items?"
)
start_cursor: Optional[str] = strawberry.field(
description="When paginating backwards, the cursor to continue."
)
end_cursor: Optional[str] = strawberry.field(
description="When paginating forwards, the cursor to continue."
)
@strawberry.type
class Edge(Generic[GenericType]):
node: GenericType = strawberry.field(description="The item at the end of the edge.")
cursor: str = strawberry.field(description="A cursor for use in pagination.")

Let us define a get_users field which returns a connection of users, as well as an UserType . Let us also plug our query into a schema.

example.py
from base64 import b64encode, b64decode
from typing import List, Optional, Generic, TypeVar
import strawberry
user_data = [
{
"id": 1,
"name": "Norman Osborn",
"occupation": "Founder, Oscorp Industries",
"age": 42,
},
{
"id": 2,
"name": "Peter Parker",
"occupation": "Freelance Photographer, The Daily Bugle",
"age": 20,
},
{
"id": 3,
"name": "Harold Osborn",
"occupation": "President, Oscorp Industries",
"age": 19,
},
{
"id": 4,
"name": "Eddie Brock",
"occupation": "Journalist, The Eddie Brock Report",
"age": 20,
},
]
def encode_user_cursor(id: int) -> str:
"""
Encodes the given user ID into a cursor.
:param id: The user ID to encode.
:return: The encoded cursor.
"""
return b64encode(f"user:{id}".encode("ascii")).decode("ascii")
def decode_user_cursor(cursor: str) -> int:
"""
Decodes the user ID from the given cursor.
:param cursor: The cursor to decode.
:return: The decoded user ID.
"""
cursor_data = b64decode(cursor.encode("ascii")).decode("ascii")
return int(cursor_data.split(":")[1])
GenericType = TypeVar("GenericType")
@strawberry.type
class Connection(Generic[GenericType]):
page_info: "PageInfo" = strawberry.field(
description="Information to aid in pagination."
)
edges: list["Edge[GenericType]"] = strawberry.field(
description="A list of edges in this connection."
)
@strawberry.type
class PageInfo:
has_next_page: bool = strawberry.field(
description="When paginating forwards, are there more items?"
)
has_previous_page: bool = strawberry.field(
description="When paginating backwards, are there more items?"
)
start_cursor: Optional[str] = strawberry.field(
description="When paginating backwards, the cursor to continue."
)
end_cursor: Optional[str] = strawberry.field(
description="When paginating forwards, the cursor to continue."
)
@strawberry.type
class Edge(Generic[GenericType]):
node: GenericType = strawberry.field(description="The item at the end of the edge.")
cursor: str = strawberry.field(description="A cursor for use in pagination.")
@strawberry.type
class User:
id: int = strawberry.field(description="The id of the user.")
name: str = strawberry.field(description="The name of the user.")
occupation: str = strawberry.field(description="The occupation of the user.")
age: int = strawberry.field(description="The age of the user.")
@strawberry.type
class Query:
@strawberry.field(description="Get a list of users.")
def get_users(
self, first: int = 2, after: Optional[str] = None
) -> Connection[User]:
if after is not None:
# decode the user ID from the given cursor.
user_id = decode_user_cursor(cursor=after)
else:
# no cursor was given (this happens usually when the
# client sends a query for the first time).
user_id = 0
# filter the user data, going through the next set of results.
filtered_data = list(filter(lambda user: user["id"] > user_id, user_data))
# slice the relevant user data (Here, we also slice an
# additional user instance, to prepare the next cursor).
sliced_users = filtered_data[: first + 1]
if len(sliced_users) > first:
# calculate the client's next cursor.
last_user = sliced_users.pop(-1)
next_cursor = encode_user_cursor(id=last_user["id"])
has_next_page = True
else:
# We have reached the last page, and
# don't have the next cursor.
next_cursor = None
has_next_page = False
# We know that we have items in the
# previous page window if the initial user ID
# was not the first one.
has_previous_page = user_id > 0
# build user edges.
edges = [
Edge(
node=User(**user),
cursor=encode_user_cursor(id=user["id"]),
)
for user in sliced_users
]
if edges:
# we have atleast one edge. Get the cursor
# of the first edge we have.
start_cursor = edges[0].cursor
else:
# We have no edges to work with.
start_cursor = None
if len(edges) > 1:
# We have atleast 2 edges. Get the cursor
# of the last edge we have.
end_cursor = edges[-1].cursor
else:
# We don't have enough edges to work with.
end_cursor = None
return Connection(
edges=edges,
page_info=PageInfo(
has_next_page=has_next_page,
has_previous_page=has_previous_page,
start_cursor=start_cursor,
end_cursor=end_cursor,
),
)
schema = strawberry.Schema(query=Query)

you can start the debug server with the following command:

Terminal window
strawberry server example:schema

Here’s an example query to try out:

query getUsers {
getUsers(first: 2) {
edges {
node {
id
name
occupation
age
}
cursor
}
pageInfo {
endCursor
hasNextPage
}
}
}
Edit this page on GitHub