py2-nostalgia

(not to be confused with the PyPI project nostalgia which appears to do something useful)

Do you ever miss the ability to unpack function arguments right in the signature? Like in the good old Python 2 days?

def print_point(label, (x, y)):
    print "({0},{1}); label={2})".format(x, y, label)

print_point("origin", (420, 69))

Say no more, as this package lets you do exactly that. With the power of PEP 3107 annotations!

from nostalgia import nostalgia

@nostalgia
def print_point(label: "(", x, y: ")"):
    print(f"({x},{y}); label={label}")

print_point("origin", (420, 69))

On top of that, you can destructure dictionaries:

from nostalgia import nostalgia

@nostalgia
def print_point(_: "{", label: ", {", x, y, z: ":height }: point }"):
    print(f"label:{label}, x:{x}, y:{y}, z:{z}")

print_point({"label": "origin", "point": {"x": 420, "y": 6, "height": 9}})

If the first parameter is named _, it is ignored. This lets you add something to the start of the unpacking signature, as you can see above.

Of course, nobody is working with points or labels nowadays. Let's take a look at a more realistic example.

Leveraging the @nostalgia decorator for AI-powered workflows in the cloud

from aiosky import the_cloud

@nostalgia
async def sing_song(
    _: "{(", topic1, topic2: "):first_line, (", topic3, topic4: "):second_line, {", vol: ":loudness", device: "}: config", max_tokens: "}", api_token):
    await the_cloud.leverage(
        "chatgpt "
        f"--song {topic1},{topic2},{topic3},{topic4} "
        f"--limit {max_tokens} {vol}@{device}",
        token=api_token
    )

# Call:

prompt = {
    "first_line": ["love", "regret"],
    "second_line": ["distance", "loss"],
    "config": {
        "loudness": 11,
        "device": "obnoxious-bluetooth-speaker",
    },
    "max_tokens": 5000,
}
await sing_song(prompt, "ABCDEF")

Keeping track of work-life balance and mental health while using the py2-nostalgia library

If you are not ready to go completely insane, you can use the mild_reminiscence decorator factory instead of the nostalgia decorator.

from nostalgia import mild_reminiscence

@mild_reminiscence("label, (x, y)")
def print_point1(label, x, y):
    ...


@mild_reminiscence("{label, {x, y, z:height}:point}")
def print_point2(label, x, y, z):
    ...


@mild_reminiscence("""
  {
    (topic1, topic2):first_line,
    (topic3, topic4):second_line,
    {vol:loudness, device}:config,
    max_tokens
  },
  api_token
""")
async def sing_song(topic1, topic2, topic3, topic4, vol, device, max_tokens, api_token):
    ...

Installation

pip install git+https://github.com/decorator-factory/py2-nostalgia

(please don't actually use this)