Data Transformation

axios_python provides two hooks specifically designed to format data before it hits the network and immediately after the response is received.

Unlike Interceptors which are meant to alter the overall configuration (like adding auth tokens or mapping full objects), Data Transformers focus strictly on the request data/headers and the raw response data.

transform_request

The transform_request pipeline modifies the payload immediately before it goes out onto the network wire. This is especially useful for custom serialization schemas.

A request transformer gets access to two parameters: data and headers.

import json
from axios_python import create

def convert_to_json(data, headers):
    if isinstance(data, dict):
        headers["Content-Type"] = "application/json"
        
        return json.dumps(data)
    
    return data

api = create({
    "transform_request": [convert_to_json]
})

response = api.post("/users", data={"username": "octocat"})

If multiple transformers are passed in the list, they execute sequentially in the exact order provided, with each step passing its result to the next.

transform_response

Conversely, transform_response allows you to modify the incoming Response.data payload directly before the instance hands control back to your business logic or response interceptors.

A response transformer takes one argument: data.

import base64
from axios_python import create

def unwrap_envelope_and_decode(data):
    if isinstance(data, dict):
        payload = data.get("internal_payload", data)
        if isinstance(payload, str):
            return base64.b64decode(payload).decode('utf-8')
        return payload

    return data

api = create({
    "base_url": "https://api.myapp.com",
    "transform_response": [unwrap_envelope_and_decode]
})

The response transformer executes after standard httpx JSON paring is attempted, meaning if the API returns valid JSON, data inside the transformer will be the initialized dictionary natively.