qirun-chen/fluent-python

Data Model

case class

ClassName = collection.namedtuple('ClassName', [attr1, attr2, ...])

len()

The implementation of __len__(self)

Note that len() for built-in types is super fast. Because there is no method call, it will read the value from the CPython's field.

len(ClassName)

iteration and slicing

The implementation of __getitem__(self)

deck[0], deck[:3], deck[1::2]  # slicing
  
for card in deck:  # iteration
	print(card)

sorting

Define the key function to return the rank value. Note that the key in dict() cannot be string type(because it is intepreted as a symbol?).

suit_values = dict(spades=1, diamonds=2, clubs=3, hearts=4)
print(suit_values)
 
def spade_high(card):
	rank_index = FrenchDeck.ranks.index(card.rank)
	print(rank_index)
	return rank_index * len(suit_values) + suit_values[card.suit]

Special methods (speical due to they are invoked by Python interpret.)

repr

String representation, like toString in Scala

bool

If __bool__ is not implemented, len() will be invoked for an object. If its result is 0, then bool(0) will return False. True otherwise.

add , mul , abs ...

List comprehension and generator expressions

Line breaks are ignored inside pairs of [], {} or (), so you can build multi-line listcomps and genexps without using the ugly \ line continuation escape.

These expressions have their own local scope like functions.

	Pros	Cons
listcomp	objects are stored for late use	create each object in memory
genexps	yield items one by one using the iterator protocol. Especially for huge cartesian product	items are not stored