Metadata-Version: 2.1
Name: mkcommit
Version: 0.0.3
Summary: Dead-simple tool for enforcing clean commit messages.
Home-page: https://github.com/kjczarne/mkcommit
Author: Krzysztof J. Czarnecki
Author-email: kjczarne@gmail.com
License: MIT
Keywords: git,commit,style,templates
Platform: UNKNOWN
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.7
Requires-Python: >=3.6
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: InquirerPy
Requires-Dist: pyperclip
Provides-Extra: dev
Requires-Dist: pdoc3 ; extra == 'dev'
Requires-Dist: flake8 ; extra == 'dev'
Requires-Dist: twine ; extra == 'dev'
Requires-Dist: setuptools ; extra == 'dev'
Requires-Dist: autopep8 ; extra == 'dev'
Requires-Dist: nox ; extra == 'dev'

# mkcommit

`mkcommit` is an extremely simple tool made for commit message generation.

## Why?

### Why would anybody need this? Aren't editors enough?

When working in teams it's hard to enforce proper Git commit message style from everyone. This CLI tool asks you questions to build a commit message that you can configure for yourself and your team with a very simple Python script.

### `commitlint` exists. Why `mkcommit`?

`commitlint` is a great tool. But it has considerable learning curve if you want to leverage its full potential. This tool strives to be the exact opposite: provide bare minimum with almost no overhead for your teammates. All they will need to learn is this one command: `mkcommit`.

## Installation

If you have Python set up, you're good to go. Run `pip install mkcommit` and you're done.

## Usage

- Run `mkcommit` to generate a Git commit message and print it to standard output.
- Run `mkcommit -c` to generate a Git commmit message and copy it to your clipboard.

## Configuration

1. At the root of your repository create a Python filed named `my_repo.mkcommit.py`.
2. Compose the script:

    ```python
    from mkcommit import Keyword, CommitMessage, ask, to_stdout

    keywords = [
        Keyword(
            "feat",
            "New feature"
        ),
        Keyword(
            "fix",
            "Bugfix"
        ),
    ]

    keyword = ask("Keyword", one_of=keywords)
    ticket_number = ask("Ticket number")
    first_line = ask("Short commit message")
    extended_message = ask("Long commit message")

    c = CommitMessage(
        f"[{ticket_number}] {keyword.keyword}: {first_line}",
        extended_message
    )

    if __name__ == "__main__":
        to_stdout(c)
    ```

    The above example shows a treasure-trove of configuration options:
    - `Keyword` - can be used to create selection lists of keywords
    - `ask` - the main prompt function using `InquirerPy` at the back-end. It consists of the following modes:
      - Bare - e.g. `ask("blah")` asks for direct textual input.
      - `one_of` - asks the user to select one value from a list.
      - `one_or_more` - asks the user to select at least one value from a list.
      - `yes_no` - asks the user a yes/no question.
    - `CommitMessage` - consists of two fields, the first one is the first-line commit message that will appear directly in `git log` and the second field is an extended commit message (usually optional).
    - `to_stdout(c)` at the very end specifies the default behavior when the script is ran directly instead of with `mkcommit`. This is optional if you only intend to use the `mkcommit` command and do not wish to ever trigger the script directly.
3. Run `mkcommit`. You can either:
    - Run it as `mkcommit -f /path/to/some.mkcommit.py`
    - Or trigger it in the current working directory. `mkcommit` will search for all `*.mkcommit.py` files and will let you select the one applicable from the list (this way you can have many config files for many diverse scenarios if you so forsee).

### Input validation

Right now, we're offering the following options for input validation:

- `mkcommit.validators.is_int` - validates input as integers
- `mkcommit.validators.is_float` - validates input as floating-point numbers
- `mkcommit.validators.max_len` - raises a validation error if maximum length is exceeded
- `mkcommit.validators.matches` - matches an arbitrary regex pattern

Example of usage:

```python
from mkcommit import Keyword, CommitMessage, ask
from mkcommit.validators import is_int, matches

ticket_number = ask("Ticket number", is_int())    # integer ticket number
initials = ask("Initials", matches(r'\w\w\w\w'))  # 4-letter initials
```

#### Writing your own validators

It is possible to quickly write your own validators. You need to create a function with a signature that matches `mkcommit.model.ValidatorClosure`, so a function that returns another function, where the inner function accepts a string and returns a boolean. Quick example:

```python
from mkcommit.model import Validator
import re


def matches(pattern: str) -> Validator:
    def _v(msg: str) -> bool:
        if re.match(pattern, msg):
            return True
        else:
            return False
    return _v
```

The above is the implementation of the `matches` validator.

You can declare validators directly within your `*.mkcommit.py` file or import them from somewhere else.


