documented 0.1.1

Templated docstrings for Python classes.

documented

Templated docstrings for Python classes.

Features

• Describe your business logic in docstrings of your classes and exceptions;
• When printing an object or an exception, the library will substitute the placeholders in the docstring text with runtime values,
• And you (or your user) will see a human readable text.

Installation

pip install documented

Example

from dataclasses import dataclass
from documented import DocumentedError


@dataclass
class InsufficientWizardryLevel(DocumentedError):
    """
    🧙 Your level of wizardry is insufficient ☹

        Spell: {self.spell}
        Minimum level required: {self.required_level}
        Actual level: {self.actual_level} {self.comment}

    Unseen University will be happy to assist in your training! 🎓
    """

    spell: str
    required_level: int
    actual_level: int

    @property
    def comment(self) -> str:
        if self.actual_level <= 0:
            return '(You are Rincewind, right? Hi!)'
        else:
            return ''


raise InsufficientWizardryLevel(
    spell='Animal transformation',
    required_level=8,
    actual_level=0,
)

which prints:

---------------------------------------------------------------------
InsufficientWizardryLevel           Traceback (most recent call last)
<ipython-input-1-d8ccdb953cf6> in <module>
     27 
     28 
---> 29 raise InsufficientWizardryLevel(
     30     spell='Animal transformation',
     31     required_level=8,

InsufficientWizardryLevel: 
🧙 Your level of wizardry is insufficient ☹

    Spell: Animal transformation
    Minimum level required: 8
    Actual level: 0 (You are Rincewind, right? Hi!)

Unseen University will be happy to assist in your training! 🎓

Usage

• Template rendering is done using str.format().
• That function receives the object instance as self keyword argument.
• From template, you can't call methods of the object, but you can access its fields and properties.
• textwrap.dedent() is applied to the result, thus Python indentation rules do not corrupt the resulting message.

Dynamically computed pieces of content may be introduced using:

• @property
• or, @cached_property for performance.

You can also access elements of lists and dicts by index, for example: {self.countries[US]}.

Making your exceptions sane

• Create your own exception classes in terms of your domain, to play a part in your business logic.

• Do not use the word Exception or Error in their names. Your code should raise things like:

  • BalanceInsufficient
  • PlanetNotFound
  • TetOffline
  • OrderDeclined

  And should not:

  • ValueError
  • Exception
  • CatastrophicalError

• Store meaningful properties of your errors in fields of the exception classes.

• Use dataclasses, attrs or pydantic to save yourself from boilerplate in __init__() — and to get IDE support.

• Maintain docstrings of your exceptions to contain up-to-date, human readable descriptions of what they mean.

• You will be stimulated to do this by documented: when an exception happens, the docstring becomes actually useful.

Links

• About naming and abstracting things: Kevlin Henney. Seven Ineffective Coding Habits of Many Programmers
• Python: Better Typed Than You Think summarizes a number of ways to handle errors in Python programs
• dry-python/returns proposes to replace exceptions with monadic Result container, which works great in Scala, Haskell, and Rust, — but arguably not everyone would want to adopt this approach in their Python codebase.

Which actually explains the meaning of this little helper: if we're stuck with exceptions in Python, why not at least make them friendlier?

This project was generated with wemake-python-package. Current template version is: 5840464a31423422d7523897d854e92408eee6b8. See what is updated since then.