m-o

A command line tool for parsing Python dataclass reprs and pretty-printing them. The name is based on the WALL-E character M-O, a robot who liked to clean things up.

Installation

$ cargo install m-o

Example

# my_data.py

@dataclass
class Dog:
    name: str
    age: int
    friends: List[str]

pip = Dog("Pip", 7, ["Quincy", "Digger"])
print(pip)

$ # without m-o:
$ python my_data.py
Dog(name="Pip", age=7, friends=["Quincy", "Digger"])

$ # with m-o:
$ python my_data.py | m-o --columns 20
Dog(
    name="Pip",
    age=7,
    friends=[
        "Quincy",
        "Digger",
    ],
)

CLI Options

USAGE:
    m-o [OPTIONS]

FLAGS:
    -h, --help       Prints help information
    -V, --version    Prints version information

OPTIONS:
    -c, --columns <columns>    Specifies the width of the terminal or file that
                               the results will be printed to. If unspecified,
                               `m-o` will try to use the width of the current
                               terminal window. Defaults to 80 columns.
    -i, --indent <indent>      The number of spaces used for a single
                               indentation in the output. [default: 4]

Motivation

Python's pprint.pprint function allows common values (tuples, lists, dicts, etc.) to be pretty-printed, but does not know how to format Python 3.7's dataclasses. I use dataclasses pretty frequently, and often need to debug deeply nested trees of dataclasses. The __repr__ method on dataclasses displays everything on one line which is difficult to read.

Rather than implementing pprint.pprint for your dataclasses (who even knows how to do this anyway?), just print out the value and pipe it into this tool. The data structure will be pretty-printed to your terminal.

(Planned/Current) Features

• Parse string escape characters (0.1.5)
• Use "Wadler-style" pretty-printing algorithm (0.1.6)
• Use stable version of pretty.rs (1.0.0)
• Add command-line options for:
  • Indentation level (currently 4 spaces) (0.1.7)
  • Target width (number of columns) (0.1.7)
• Better error messages when parsing fails
• Allow non-keyword arguments in constructors (ex: Dog('Pip', age=7)) (0.1.7)
• Support multi-identifier paths in symbols (ex: datetime.datetime) (0.1.7)