FastHTTP

FastHTTP is a lightweight wrapper for %Net.HttpRequest in InterSystems IRIS, designed to simplify HTTP requests with a concise API and built-in JSON support.

Installation

With Docker

Clone the repository and run:

docker-compose up -d

With ZPM

zpm "install fast-http"

Usage

Simple GET Request

Set response = ##class(dc.http.FastHTTP).DirectGet("url=https://httpbin.org/get")
write response.%ToJSON()

POST Request with JSON Body

Set body = {"name": "Iris", "type": "Database"}
Set response = ##class(dc.http.FastHTTP).DirectPost("url=https://httpbin.org/post", body)
write response.%ToJSON()

PUT & DELETE

// PUT
Set response = ##class(dc.http.FastHTTP).DirectPut("url=https://httpbin.org/put", {"update": 1})
// DELETE
Set response = ##class(dc.http.FastHTTP).DirectDelete("url=https://httpbin.org/delete")

Configuration String

FastHTTP uses a configuration string to set up the request. You can pass it to the FastHTTP constructor or the Direct... methods.

Format: "key=value,key2=value2"

Supported keys:

• url: The full URL to send the request to.
• Header_<Name>: Sets a request header. Ex: Header_Authorization=Bearer 123

Example with headers:

Set config = "url=https://api.example.com/data,Header_Authorization=Bearer mytoken,Header_Content-Type=application/json"
Set response = ##class(dc.http.FastHTTP).DirectGet(config)

Accessing the Client Instance

The Direct... methods also return the client instance as an output parameter if you need to access the underlying %Net.HttpRequest or response metadata.

Set response = ##class(dc.http.FastHTTP).DirectGet("url=https://httpbin.org/get", , .client)
Write "Status Code: ", client.HttpRequest.HttpResponse.StatusCode

Server-Sent Events (SSE) / AI Streaming

FastHTTP has built-in support for Server-Sent Events (SSE), making it extremely easy to consume streaming APIs like OpenAI, Anthropic, or any LLM in real-time. Instead of waiting for the full response to complete, you can process chunks on the fly.

To consume an SSE stream, follow these steps:

1. Create a dc.http.Stream which will hold the incoming bits.
2. Link it to a dc.http.SSEHandler, configured with an Adapter (a class extending dc.http.SSEAdapter).
3. Pass your stream to the FastHTTP request.

Testing with the Mock Server

If you have started the workspace using Docker, a small sse-mock service is included in the docker-compose.yml. You can use it to simulate an AI streaming response without needing a real API key or internet connection.

Once the mock server is running, you can test the SSE behavior in the IRIS terminal with the following snippet:

Set stream = ##class(dc.http.SSEChatConsoleAdapter).GetStream()
Set config = "url=http://sse-mock:5000/stream,timeout=10"
Set response = ##class(dc.http.FastHTTP).DirectGet(config, , , stream)

Testing with OpenAI

If you have an access to OpenAI API or LLM API compatible, you can use the following template:

Set stream = ##class(dc.http.SSEChatConsoleAdapter).GetStream()
Set config = "url=https://api.openai.com/v1/chat/completions,Header_Authorization=Bearer ,Header_Accept=text/event-stream"
Set body = {"model": "gpt-4", "messages": [{"role": "user", "content": "Tell me a short story."}], "stream": true }
Set response = ##class(dc.http.FastHTTP).DirectPost(config, body, .client, stream)