Skip to content

dataclasses.md

Latest commit

 

History

History
125 lines (88 loc) · 4.19 KB

dataclasses.md

File metadata and controls

125 lines (88 loc) · 4.19 KB

Dataclasses

Has this ever happened to you? Creating a class with a lot of parameters, and you need to initialize one by one manually.

class Product(object):
    def __init__(self, id, author_id, category_id, brand_id, spu_id, 
                 title, item_id, n_comments, creation_time, update_time, 
                 source='', parent_id=0, ancestor_id=0):
        
        self.id = id
        self.author_id = author_id
        ...

Sometimes you also have to write a __repr__ or __str__ to express the classes with these variables, or __eq__, __gt__ to compare the classes. This is very annoying and tedious, so python provides a dataclasses module in version 3.7 to simplify the process in building classes.

Table of Contents

Basic

dataclasses will automatically add __repr__, __eq__ and other magic methods for us. Just install and import the dataclasses module and add the dataclass decorator to the class.

from dataclasses import dataclass

@dataclass
class InventoryItem:
    name: str
    unit_price: float
    quantity_on_hand: int = 0

    def total_cost(self) -> float:
        return self.unit_price * self.quantity_on_hand

item = InventoryItem("product", 100, 1)
print(item)  # InventoryItem(name='product', unit_price=100, quantity_on_hand=1)

You can see that the class of each variable is marked with variable: type, which is python 3 function annotations. The typing module that helps annotations will also be used later on.

Post Init

There are some variables that we want to change in __init__, so dataclasses provides a __post_init__ function to do that.

from dataclasses import dataclass

@dataclass
class InventoryItem:
    name: str
    unit_price: float
    quantity_on_hand: int = 0

    def __post_init__(self):
        self.unit_price *= 1.1

Field

dataclasses provides a field() method to initialize each variable differently. The parameters to field() are:

  • default: default value
  • default_factory: zero-argument callable used to define default value
  • init: bool, indicates whether this variable should be put into __init__ or not.
  • compare: bool, indicates whether this variable should be put into a comparable magic method such as __eq__, __gt__, etc.
from dataclasses import dataclass, field
from typing import List

@dataclass
class InventoryItem:
    name: str
    unit_price: float = field(default=0.0)
    quantity_on_hand: int = field(default=0, repr=False)
    parts: List[str] = field(default_factory=list)


item = InventoryItem("product")
print(item)  # InventoryItem(name='product', unit_price=0.0, parts=[])

InitVar

dataclasses also provides variables that only appear in __post_init__ and don't really become class or instance variables. Just define the variable as InitVar and pass it as __post_init__.

from dataclasses import InitVar, dataclass, field
from typing import List


@dataclass
class InventoryItem:
    name: str
    unit_price: float = field(default=0.0)
    quantity_on_hand: int = field(default=0, repr=False)
    parts: List[str] = field(default_factory=list)
    parts_number: InitVar[int] = 0

    def __post_init__(self, parts_number):
        self.parts.extend([f"part{i}" for i in range(1, parts_number + 1)])


item = InventoryItem("product", parts_number=2)
print(item)  # InventoryItem(name='product', unit_price=0.0, parts=['part1', 'part2'])

Related Articles

Article Link
attrs 和 Python3.7 的 dataclasses https://zhuanlan.zhihu.com/p/34963159
Using Python dataclasses to simplify managing class objects https://www.youtube.com/watch?v=FcVCfGJrkUQ