QBox

A lazily awaited container for async operations.

QBox solves the “colored functions” problem by wrapping async operations in a transparent proxy that defers evaluation until the value is “observed.”

The Problem

Async/await creates a viral pattern where every function that uses async code must also be async:

# The traditional way - async spreads everywhere
async def get_user_data():
    user = await fetch_user()
    profile = await fetch_profile(user.id)
    return await process_data(user, profile)

# Every caller must also be async
async def main():
    data = await get_user_data()
    if data.score > 100:
        print("High score!")

The Solution

QBox lets you write normal synchronous code that transparently handles async operations:

from qbox import QBox

# Wrap async calls in QBox - think of the variable as your data
user = QBox(fetch_user())
profile = QBox(fetch_profile(user.id))  # Lazy - chains without blocking
score = (user.score + profile.bonus) * 2  # Still lazy!

# Only blocks when you actually need the value
if score > 100:  # Observation happens here
    print("High score!")

The Quantum Metaphor

Like Schrödinger’s cat, a QBox value exists in superposition until observed:

import asyncio
import random
from qbox import QBox, observe

class Cat:
    pass

class AliveCat(Cat):
    def speak(self) -> str:
        return "Meow!"

class DeadCat(Cat):
    def speak(self) -> str:
        return "..."

async def infernal_device() -> Cat:
    """The cat's fate is determined only when observed."""
    await asyncio.sleep(0.1)  # Quantum uncertainty...
    return AliveCat() if random.random() > 0.5 else DeadCat()

# The cat exists in superposition
cat = QBox(infernal_device())  # cat is a QBox[Cat], fate undetermined

# Observation collapses the wave function
observed_cat = observe(cat)  # NOW the cat is either alive or dead
print(observed_cat.speak())  # "Meow!" or "..."

# After observation, 'cat' IS the actual Cat object (not a QBox)
print(type(cat))  # <class 'AliveCat'> or <class 'DeadCat'>

Installation

pip install qbox

Quick Start

import asyncio
from qbox import QBox

async def fetch_number() -> int:
    await asyncio.sleep(0.1)  # Simulate async work
    return 42

# Create a QBox - starts the async operation immediately (default)
number = QBox(fetch_number())

# Chain operations - all lazy, no blocking
result = (number + 8) * 2 - 10  # Returns QBox, not int

# Value is computed only when needed
if result > 50:  # Blocks here, evaluates entire chain
    print(f"Result: {result}")  # Prints "Result: 90"

When Does Code Execute?

Default behavior (start=’soon’):

async def log_and_fetch():
    print("EXECUTING")
    return await fetch_data()

data = QBox(log_and_fetch())  # Coroutine submitted NOW, starts running
print("Continuing...")        # "EXECUTING" may print during this line
if data:                      # Blocks until result ready (may already be done)
    print(data)

Deferred behavior (start=’observed’):

data = QBox(log_and_fetch(), start='observed')
# "EXECUTING" has NOT printed - coroutine not submitted yet

print("Doing other work...")  # Nothing happening in background
if data:                      # NOW coroutine submitted, blocks for result
    print(data)
# Output:
# Doing other work...
# EXECUTING
# <the data>

When to use each:

start=	Use When
'soon' (default)	You’ll likely need the value; want parallelism with other sync work
'observed'	Might not need the value; building lazy chains; deferring expensive work

API Reference

Creating a QBox

from qbox import QBox
from collections.abc import Mapping

# Basic creation (starts immediately by default)
user = QBox(some_coroutine())

# Deferred execution
config = QBox(load_config(), start='observed')

# With type hint for isinstance support
data = QBox(fetch_dict(), mimic_type=Mapping)

# With custom observation scope
result = QBox(fetch_data(), scope='locals')  # 'locals', 'stack', or 'globals'

# With repr that triggers observation
debug_value = QBox(fetch_value(), repr_observes=True)

Constructor Parameters

QBox(
    awaitable,            # The coroutine or awaitable to wrap
    mimic_type=None,      # Type for isinstance mimicry (e.g., Mapping)
    scope='stack',        # Reference replacement scope
    start='soon',         # When to start: 'soon' or 'observed'
    repr_observes=False,  # Whether repr() triggers observation
    cancel_on_delete=True,  # Whether to cancel pending work on deletion
)

Type Checking with QBox

QBox is transparent to static type checkers. For runtime checks, use QBox._qbox_is_qbox() and prefer ABCs via mimic_type to keep laziness. See Type Checking Guide and Static Typing for details.

Accessing the Value

from qbox import QBox, observe

# Explicit observation (recommended - also replaces references)
value = observe(data)

# Safe observation (works on non-QBox values too)
value = observe(maybe_a_qbox)

# Or use in boolean/comparison context (auto-evaluates)
if data > 10:
    print("Large!")

Lazy Operations (return new QBox)

• Arithmetic: +, -, *, /, //, %, **

• Bitwise: &, |, ^, <<, >>

• Unary: -data, +data, abs(data), ~data

• Item access: data[key], data[start:end]

• Attribute access: data.attribute

• Calling: data(args)

• repr() with default repr_observes=False

Force Evaluation (return concrete values)

• Comparisons: <, <=, ==, !=, >=, >

• Type conversions: bool(), str(), int(), float()

• Container operations: len(), in, iteration

• Math functions: round(), math.floor(), math.ceil()

• repr() with repr_observes=True

Side Effects and Exceptions

Since execution is lazy by default with start='observed', side effects and exceptions occur during observation, not during QBox creation.

Side Effects

async def write_to_database(record):
    await db.insert(record)  # Side effect!
    return record.id

record_id = QBox(write_to_database(user_record), start='observed')
# Database write has NOT happened yet!

# To ensure side effects have occurred:
observe(record_id)  # Write happens HERE

Exceptions

async def might_fail():
    raise ValueError("Something went wrong")

result = QBox(might_fail())
# No exception raised yet - exceptions surface on observation, not creation

try:
    if result:  # Exception raised HERE on observation
        print(result)
except ValueError:
    print("Caught during observation")

# Exceptions are cached - subsequent access re-raises the same exception

Type Mimicry

QBox aims to be invisible - when you observe a value, the box disappears:

user = QBox(fetch_user())
print(user.name)  # After this, `user` IS the User object

isinstance Support

Use ABCs for lazy type checking without forcing evaluation:

from collections.abc import Mapping
from qbox import QBox

data = QBox(fetch_dict(), mimic_type=Mapping)
isinstance(data, Mapping)  # True, stays lazy!

Or opt-in to full transparency:

from qbox import enable_qbox_isinstance

enable_qbox_isinstance()
isinstance(data, dict)  # Works! Forces observation automatically

Or use the context manager for scoped transparency:

from qbox import qbox_isinstance

with qbox_isinstance():
    isinstance(data, dict)  # Works within this block
# Original isinstance restored after block

Explicit Observation

Use the observe() function for explicit control:

from qbox import observe

value = observe(data)  # Force evaluation, returns unwrapped value
value = observe(non_qbox)  # Safe for non-QBox values (returns unchanged)

See Type Checking Guide for details.

Error Handling

Exceptions from the wrapped coroutine are cached and re-raised on access:

async def failing():
    raise ValueError("oops")

result = QBox(failing())

try:
    observe(result)  # Raises ValueError
except ValueError:
    pass

observe(result)  # Raises same ValueError (cached)

Thread Safety

QBox is designed for thread-safe access:

• A singleton background event loop runs on a daemon thread

• Value caching uses proper locking

• Multiple threads can safely access the same QBox

Requirements

• Python 3.10+

• No runtime dependencies (pure stdlib)

Contributing

Contributions are welcome! Please see CONTRIBUTING.rst for guidelines.

License

Distributed under the MIT License. See LICENSE.rst for details.