Using the Client

Loading the client

The generated APLSource/ directory is loaded into your workspace using LINK:

]LINK.Create # ./APLSource

This makes the Client class (and all generated tag namespaces) available in your workspace.

Creating a client instance

The constructor takes a config namespace. At minimum, baseUrl is required:

config ← (baseUrl: 'https://api.example.com')
client ← ⎕NEW Client config

For APIs that require authentication, add a security namespace:

config ← (
    baseUrl: 'https://api.example.com'
    security: (bearerToken: 'my-token')
)
client ← ⎕NEW Client config

Credentials can also be read from environment variables at request time using the env:[VAR_NAME] syntax¹:

config ← (
    baseUrl: 'https://api.example.com'
    security: (bearerToken: 'env:[API_TOKEN]')
)

See Client Class for the full list of config fields and authentication options.

Making requests

Operations are called as client.<tag>.<OperationId>, passing a namespace of parameters:

⍝ Operation with parameters
args ← (id: 42)
response ← client.user.GetById args

⍝ Operation with no parameters
response ← client.user.ListUsers ()

See Operation Functions for details on how to pass path, query, header, and body parameters.

Working with the response

For the purposes of this section, we assume mock is at its default value of ¯1 — see Mock mode for what is returned otherwise.

Each call returns an HttpCommand response namespace containing information about the request and the response from the server. The most commonly used fields are:

The response namespace contains many more fields. See the HttpCommand result reference for the full list.

response ← client.user.ListUsers ()
response.HttpStatus   ⍝ 200
response.Data         ⍝ parsed response body

Non-2xx responses are returned the same way — the client does not signal an error on HTTP error status codes. Check HttpStatus in your own code to handle failures. See Error Handling for more.