Skip to main content

Library for metaprogramming

Project description

Pyhandling

Library for advanced continuous handling of anything

Provides tools to extend single call logic on a nearly unlimited scale
You can even integrate the entire program logic into one call

Installation

pip install pyhandling

Usage examples

Composition

Merge your functions into one

from pyhandling import *


complemented_number = str |then>> (lambda line: line + '6') |then>> int
complemented_number(25)

to later get

256

or you can do the same but call the function immediately

25 >= str |then>> (lambda line: line + '6') |then>> int

and get the same result

256

Currying

Add additional arguments to function input arguments

formattly_sum = "{} {}{}".format

post_partial(formattly_sum, "world", '!')("Hello") 

using pseudo operators

(formattly_sum |to| "Hello")("world", '!')
(formattly_sum |to* ("Hello", "world"))('!')

(formattly_sum |by| '!')("Hello", "world")

or not necessarily now

container = close(formattly_sum)
opened_container = container("Hello")

opened_container("world", '!')

using all possible ways

post_container = close(formattly_sum, closer=post_partial)

post_container('!')("Hello", "world")

Eventually, they all return

Hello world!

Interface control

Abstract the output value

print(returnly(print)("Some input argument"))
Some input argument
Some input argument

or input values

from functools import partial


eventually(partial(print, 16))(1, 2, 3)
16

Atomic functions

Use synonyms for operators

return_(256)
raise_(Exception("Something is wrong"))
256

Traceback ...
Exception: Something is wrong

for atomic operations

execute_operation(60, '+', 4)
transform(str(), 'not')
64
True

for syntax operations

call(range, 16)

getitem_of({"some-key": "some-value"}, "some-key")
range(16)
some-value

Annotating

Use standart annotation templates from annotations package for routine cases

from pyhandling.annotations import checker_of, reformer_of, merger_of

from pyannotating import number


is_number_even: checker_of[number] = lambda number: number % 2 == 0

add_hundert_to: reformer_of[number] = lambda number: number + 100

format_lines: merger_of[str] = "{} {}!".format

or annotations themselves

from pyannotating import many_or_one

from pyhandling.annotations import handler, decorator


executing_of: Callable[[many_or_one[handler]], decorator] = ...

Function building

Create functions by describing them

total_sum: Callable[[Iterable[many_or_one[number]]], number] = documenting_by(
    """
    Function of summing numbers from the input collection or the sum of its
    subcollection.
    """
)(
    close(map |then>> tuple)(
        on_condition(isinstance |by| Iterable, sum, else_=return_)
    )
    |then>> sum
)

in several processing processes

ratio_of_square_to_full: reformer_of[number] = documenting_by(
    """
    Function of getting the ratio of the square of the input number to the
    input number to the power of its value.
    """
)(
    mergely(
        take(execute_operation),
        mergely(
            take(execute_operation),
            return_,
            take('*'),
            return_
        ),
        take('/'),
        mergely(
            take(execute_operation),
            return_,
            take('**'),
            return_
        )
    )
)

or in an indefinite number of iterative executions

from pyhandling.annotations import dirty


increase_up_to_ten: dirty[reformer_of[number]] = documenting_by(
    """
    Function that prints numbers between the input number and 10 inclusive and
    returns 10.
    """
)(
    recursively(
        returnly(print) |then>> post_partial(execute_operation, '+', 1),
        post_partial(execute_operation, '<', 10)
    )
    |then>> returnly(print)
)


increase_up_to_ten(8)
8
9
10

Chain breaking

Forcibly break the chain of actions

optionally_exponentiate: Callable[[number], number | BadResourceWrapper] = documenting_by(
    """Function of exponentiation of the input number if it is > 0."""
)(
    maybe(
        on_condition(
            post_partial(execute_operation, '<', 0),
            BadResourceWrapper,
            else_=return_
        )
        |then>> post_partial(execute_operation, '**', 2)
    )
)


optionally_exponentiate(-16)
<Wrapper of bad -16>

with the possibility of returning a "bad" resource

main: dirty[reformer_of[number]] = optionally_exponentiate |then>> optionally_get_bad_resource_from

main(8)
main(-16)
64
-16

You can also interrupt by returning an error proxy that stores the error
that occurred while processing this resource and the resource itself

from pyhandling.annotations import reformer_of


div_by_zero: reformer_of[number] = documenting_by(
    """Function for dividing an input number by zero."""
)(
    post_partial(execute_operation, '/', 0)
)


main: Callable[[number], number | BadResourceError] = (
    returnly_rollbackable(div_by_zero, take(True))
)


main(256)
BadResourceError('Resource "256" could not be handled due to ZeroDivisionError: division by zero')

with corresponding possibilities

main: reformer_of[number] = (
    partial(map |then>> maybe, returnly_rollbackable |by| take(True))(
        post_partial(execute_operation, '*', 2)
        |then>> div_by_zero
    )
    |then>> optionally_get_bad_resource_from
)


main(16)
32

Batteries

Use out-of-the-box functions to abstract from input arguments

take(256)(1, 2, 3)
event_as(execute_operation, 30, '+', 2)(1, 2, 3)
256
32

to create a collection via call

collection_from(1, 2, 3)
(1, 2, 3)

to connect collections

summed_collection_from((1, 2), (3, 4))
(1, 2, 3, 4)

to manage collection nesting

wrap_in_collection(8)
open_collection_items(((1, 2), [3], 4))
(8, )
(1, 2, 3, 4)

to represent something as a collection

as_collection(64)
as_collection([1, 2, 3])
(64, )
(1, 2, 3)

to confirm something multiple times

runner = times(3)
tuple(runner() for _ in range(8))
(True, True, True, False, True, True, True, False)

to raise only a specific error

optional_raise = optional_raising_of(ZeroDivisionError)

optional_raise(TypeError())
optional_raise(ZeroDivisionError("can't divide by zero"))
TypeError()

Traceback ...
ZeroDivisionError: can't divide by zero

to execute operations

operation_by('*', 4)(64)
callmethod(', ', 'join', ("first", "second"))
256
first, second

to decoratively create action chains

next_action_decorator_of(operation_by('**', 4))(operation_by('+', 1))(3)
previous_action_decorator_of(operation_by('+', 2))(operation_by('**', 2))(6)
256
64

to stop the chain when an error occurs

breakable_chain = chain_breaking_on_error_that(isinstance |by| ZeroDivisionError)(
    (execute_operation |by* ('+', 4)) |then>> div_by_zero
)

breakable_chain(12)
BadResourceError('Resource "16" could not be handled due to ZeroDivisionError: division by zero')

to use shortcuts of routine options

yes(1, 2, 3)
no(1, 2, 3)
True
False

Immutable classes

Create immutable classes

from typing import Iterable, Callable


@publicly_immutable
class CallingPublisher:
    name = DelegatingProperty('_name')
    followers = DelegatingProperty('_followers', getting_converter=tuple)

    def __init__(self, name: int, followers: Iterable[Callable] = tuple()):
        self._name = name
        self._followers = list(followers)

    def __repr__(self) -> str:
        return f"Publisher {self._name} with followers {self._followers}"

    def __call__(self, *args, **kwargs) -> None:
        for follower in self._followers:
            follower(*args, **kwargs)

    @to_clone
    def with_follower(self, follower: Callable) -> None:
        self._followers.append(follower)


original = CallingPublisher("Some publisher", [print])

that can't change any public attribute

original.some_attr = "some value"
Traceback ...
AttributeError: Type CallingPublisher is immutable

and automatically clone without manual creation

other = original.with_follower(operation_by('**', 4) |then>> print)

original.followers
other.followers
(<built-in function print>,)
(<built-in function print>, ActionChain(...))

what would eventually

other(4)
4
256

Debugging

Display intermediate results

showly(total_sum)([128, [100, 28]])
[128, [100, 28]]
(128, 128)
256

by different ways

logger = Logger(is_date_logging=True)

showly(total_sum, writer=logger)([[2, 10], [15, 15]])

print(*logger.logs, sep='\n')
[2023-01-24 21:38:28.791516] [[2, 10], [15, 15]]
[2023-01-24 21:38:28.791516] (12, 30)
[2023-01-24 21:38:28.791516] 42

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

Pyhandling-2.2.0.tar.gz (38.5 kB view details)

Uploaded Source

File details

Details for the file Pyhandling-2.2.0.tar.gz.

File metadata

  • Download URL: Pyhandling-2.2.0.tar.gz
  • Upload date:
  • Size: 38.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.1 CPython/3.11.0

File hashes

Hashes for Pyhandling-2.2.0.tar.gz
Algorithm Hash digest
SHA256 bd1c71d7c29b11cdc859407a9d9728b08b3194db12032cfff9e3fc57ac221cfb
MD5 6ed0b084e7700b3ce434812617cfc4c6
BLAKE2b-256 c5bd1292586b30e0af05d78dbb155652b3d4ec93eef5be3561ca49dc57dfa6c0

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page