feat: top level functions

[dmadisetti]

📝 Summary

Adds top level functions

• @app.fn (but maybe app.function is better) to wrap functions, register them but leave them directly exposed
• Serialize to @app.fn if there are no dependencies

Just an intro PR for #2293

For completion of 2293, I think out standing changes are:

• some UI aspect (maybe a new language toggle?)
• Top level imports
• Looser requirements for serialization (can use imports and other app.fns)

But this is hidden under App(_toplevel_fn=True)

🔍 Description of Changes

see the smoke test or:

@app.fn
def self_ref_fib(n: int) -> int:
    if n == 0:
        return 0
    if n == 1:
        return 1
    return self_ref_fib(n - 1) + self_ref_fib(n - 2)


@app.fn(disabled=True, hide_code=True)
def divide(x, y):
    return y / x


@app.fn
def subtraction(a: "int", b: "int") -> "int":
    return a - b


@app.fn
# Comments inbetween
def multiply(a, b) -> "int":
    return a * b


@app.fn
def addition(a: int, b: int) -> int:
    # int is considered no good, re-eval
    return a + b

📋 Checklist

• I have read the contributor guidelines.
• For large changes, or changes that affect the public API: this change was discussed or approved through an issue, on Discord, or the community discussions (Please provide a link if applicable).
• I have added tests for the changes made.
• I have run the code and verified that it works as expected.

📜 Reviewers

@akshayka OR @mscolnick

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name	Status	Preview	Comments	Updated (UTC)
marimo-docs	✅ Ready (Inspect)	Visit Preview	💬 Add feedback	Feb 12, 2025 4:23am
marimo-storybook	✅ Ready (Inspect)	Visit Preview	💬 Add feedback	Feb 12, 2025 4:23am

[akshayka]

This is excellent. Saw the attention to detail with preserving comments in various locations — between the decorator invocation and function def, in the function body ... and really appreciate that. My only note is what you already called out, to rename app.fn to app.function.

[akshayka]

Awesome! LGTM modulo failing tests, looks like just snapshots that need to be updated.

[akshayka]

Weird, no idea what these failures are. They don't seem related ...

FAILED tests/_server/api/endpoints/test_execution.py::TestExecutionRoutes_EditMode::test_app_meta_request - KeyError: 'test-1'
FAILED tests/_server/test_sessions.py::test_kernel_manager_interrupt - assert False
 +  where False = empty()
 +    where empty = <multiprocessing.queues.Queue object at 0x0000028EAF598A60>.empty
 +      where <multiprocessing.queues.Queue object at 0x0000028EAF598A60> = <marimo._server.sessions.QueueManager object at 0x0000028EAF5983D0>.control_queue
= 2 failed, 2157 passed, 217 skipped, 82 deselected, 18 xfailed, 4 xpassed, 7 warnings in 298.97s (0:04:58) =

[akshayka]

I'm also unable to reproduce the test_export_markdown failure, and we're also seeing it in other branches, so that seems unrelated.

On the surface it feels like it would be related to this PR, but I also see it failing in another PR ... https://github.com/marimo-team/marimo/actions/runs/13278229043/job/37071591905?pr=3760