Skip to content

Developer-friendly value range checks with user-friendly error messages

License

Notifications You must be signed in to change notification settings

TheMatjaz/Rangeforce

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

35 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Rangeforce: developer-friendly range checks with user-friendly error messages

Ever had to write the same old if value < 0 bit of Python code to validate a user input? Tired of not understandable error messages like "Illegal value"?

Rangeforce is a very simple module offering multiple functions checking the range of values, including integers or floats, or lengths of structures like strings or lists. It does so in one line of code while providing an understandable error message that can be displayed directly to the user.

def classic_approach():
    value = int(input('How many hours do you sleep per day? '))
    if value < 0:
        raise ValueError('Hours of sleep must be positive')
    elif value > 24:
        raise ValueError('Hours of sleep must be max 24')
    else:
        return value

# Same code, but simplified using Rangeforce
import rangeforce as rf

def with_rangeforce():
    value = int(input('How many hours do you sleep per day? '))
    return rf.limited(value, 0, 24, 'Hours of sleep')  # Magically in 1 line
    # In case of error it raises a RangeError with a simple readable message:
    # rangeforce.RangeError: Hours of sleep must be in range [0, 24]. 42 found
    # instead.

Features

  • Validate that a value is within a [min, max] interval or equal to another one
    • Including case when min is -infinity and max is +infinity
    • Optional enforcing of the type (e.g. the value must be a float)
  • Functions to validate that a value is an unsigned/signed integer that fits into 8/16/32/64 bits, as the C data types
  • Functions to validate positive, negative, non-positive, non-negative integers
  • Validation of the length of an object, either within a [min, max] interval or exact length
  • Utility function to clip (limit) a value to fit within a range
  • Customisable name of the variable under validation for the error message
  • Customisable exception type raised (defaults to RangeError)

Installation

pip install Rangeforce

or just include the rangeforce.py file in your project (copy-paste). It's self-contained and has no dependencies other than the standard Python library (specifically math).

Example usage

import rangeforce as rf

value = rf.limited(8000, 20, 5000, dtype=int)
# If successful, value will held 8000, otherwise (as it would happen in this
# example) raises a rangeforce.RangeError (or another custom exception)
# with a useful message:
# "Value must be in range [20, 5000]. 8000 found instead."
# 
# The message is ready to be shown directly to the user:

try:
    value = rf.limited(8000, 20, 5000, dtype=int)
except rf.RangeError as error:
    print(str(error))

# A missing bound (min or max) means unbounded
value = rf.limited(2000.0, None, 5000.0, dtype=float)
# Value must be <= 5000 but can be as small as it gets, including negative


# Especially useful in setters to validate the input in one line
class FullHdPicturesPixel(object):
    def __init__(self, x, y):
        self._x = None
        self._y = None
        self.x = x
        self.y = y
    
    @property
    def x(self):
        return self._x
    
    @x.setter
    def x(self, new_x):
        self._x = rf.limited(new_x, 0, 1920, 'The X pixel coordinate', int)
    
    @property
    def y(self):
        return self._y
    
    @y.setter
    def y(self, new_y):
        self._y = rf.limited(new_y, 0, 1080, 'The Y pixel coordinate', int)
        
pixel = FullHdPicturesPixel(10, 2000)
# This raises a RangeError with the message:
# "The Y pixel coordinate must be in range [0, 1080]. 2000 found instead."


# Operating with C-like data structures and binary data?
# These functions might come in handy!
value = rf.int8(20) 
value = rf.int16(20) 
value = rf.int32(20) 
value = rf.int64(20)
value = rf.uint8(20) 
value = rf.uint16(20) 
value = rf.uint32(20) 
value = rf.uint64(20)

# Customize the name of the variable:
distance = rf.uint8(-3, 'Distance')
# This raises a RangeError with the message:
# "Distance must be in range [0, 255]. -3 found instead."

# Customize the exacption type raised:
distance = rf.uint8(-3, 'Distance', ex=OverflowError)
# This raises an OverflowError with the message:
# "Distance must be in range [0, 255]. -3 found instead."

# Compare two values for equality
answer = rf.exactly(42, 42, 'The answer')

# To check the length range of anything (e.g. a list or bytes):
value = rf.limited_len([1, 2, 3], 2, 7) # 'value' will hold [1, 2, 3]

# If you need an exact length, do it like this
pair = rf.exact_len([10, 20, 30], 2, name='pair of values')
# This raises a RangeError with the message:
# "Length of pair of values must be exactly 2. 3 found instead.

About

Developer-friendly value range checks with user-friendly error messages

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages