Let's streamline this:

1. If json.loads(request.content) works, go with that. json.loads accepts raw bytes, so we don't have to decode for that.
2. If https://docs.mitmproxy.org/stable/api/mitmproxy/http.html#Message.text does not raise, assign body_str to repr(text)
3. As a fallback, assign body_str to repr(request.content).

Technically steps 1-2 are optional, but definitely nice to have.

Not sure how to do step 2.

I was thinking of something like this:

from pprint import pformat

try:
    body_val = json.loads(request.content)
except ValueError:
    try:
        body_val = request.text
    except ValueError:
        body_val = request.content

body_str = f"body = {pformat(body_val, indent=4, sort_dicts=False)}"

And this, will create weird code format like this

body = {   'string': 'Hello, world!',
    'number': 42,
    'float': 3.14,
    'boolean': True,
    'nullValue': None,
    'object': {'name': 'John', 'age': 30},
    'array': [1, 2, 3, 4]}

Can we use pformat here instead of manually constructing stuff?

No. This will create something like this

cookies = MultiDictView[['cookie', 'chocolate_chip'], ['session_id', 'abc123'], ['user_id', '987654321']]

Can we use pformat here instead of manually constructing stuff?

Similar to the above, this will create

headers = Headers[(b'header', b'qvalue')]

Let's use repr() here, then we don't need any custom escaping functions