joy-format: a python formatter for aligning adjacent assignment expressions

---

I like like the aesthetic of assignment statements.

Seeing code formatted in this way brings me joy, so I'm writing a formatter in the hope that I can use it in some projects of mine. Ideally, it will extend something like black's formatting, which I by-and-large find to be great.

Examples:

Simple assignment example:

# Before
foobar = baz
foo = foobar

# After
foobar = baz
foo    = foobar

More realistic example:

# Before
self.first_linear = nn.Linear(config.hidden_size, config.hidden_size, bias=False)
self.second_linear = nn.Linear(config.hidden_size, config.num_labels, bias=False)
self.third_linear = nn.Linear(config.hidden_size, config.num_labels)

# After
self.first_linear  = nn.Linear(config.hidden_size, config.hidden_size, bias=False)
self.second_linear = nn.Linear(config.hidden_size, config.num_labels, bias=False)
self.third_linear  = nn.Linear(config.hidden_size, config.num_labels)

Method definition with default values:

# Before
def forward(
    self,
    input_ids: torch.Tensor = None,
    attention_mask=None,
    token_type_ids=None,
    position_ids=None,
    head_mask=None,
    inputs_embeds=None,
    labels=None,
    output_attentions=None,
    output_hidden_states=None,
    return_dict=None,
): ...

# After
def forward(
    self,
    input_ids : torch.Tensor = None,
    attention_mask           = None,
    token_type_ids           = None,
    position_ids             = None,
    head_mask                = None,
    inputs_embeds            = None,
    labels                   = None,
    output_attentions        = None,
    output_hidden_states     = None,
    return_dict              = None,
): ...

Type hinted function arguments:

# Before
def f(
    attention: torch.Tensor,
    labels: np.ndarray,
    logger: logging.Logger,    
) -> tuple[torch.Tensor, bool]: ...

# After
def f(
    attention : torch.Tensor,
    labels    : np.ndarray,
    logger    : logging.Logger,
) -> tuple[torch.Tensor, bool]: ...

Pre-commit installation

Add the following to your .pre-commit-config.yaml file. Note: compatibility with other linters / formatters is not guaranteed, please create an issue for any conflicts you encounter.

repos:
-   repo: https://github.com/JosephSBoyle/joy_formatter.git
    rev: latest
    hooks:
    -   id: align-assignments

VScode Installation

1. Git clone this repo.
2. Copy the full path to joy.py
3. Install the 'custom local formatters extension'
4. Open the settings.json file by pressing ctrl+shift+p and selecting "Open user settings json".
5. Add these key-value pairs.

    "python.autoComplete.extraPaths": [],
    "python.analysis.extraPaths": [],
    "customLocalFormatters.formatters": [
        {
          "command": "python <PATH TO YOUR joy.py SCRIPT HERE>",
          "languages": ["python"]
        }
    ],

6. Replace <PATH TO YOUR joy.py SCRIPT HERE> with the path from step #2!

Dev-Notes

The ./rt command is used as a shorthand to run the tests, passing any flags to pytest. Sort of how some people use a test command in their makefiles and invoke make test to run their tests.

If any code is malformatted by this tool, please open an issue with the input and output that you expected / received and I'll do my best to take a look.