Python Pocket Reference, 3rd Edition

26

|

Python Pocket Reference

Dictionaries

Dictionaries are mutable (i.e., changeable) tables of object

references, accessed by key, not position. They are unor-

dered collections, implemented internally as dynamically

expandable hash tables.

Literals and creation

Dictionaries are written as comma-separated series of key:

value

pairs inside curly braces. Assigning to new keys gener-

ates new entries. Any immutable object can be a key (e.g.,

string, number, tuple), and class instances can be keys if they

inherit hashing protocol methods. Tuple keys support com-

pound values (e.g.,

adict[(M,D,Y)], with parentheses

optional).

{}

An empty dictionary.

{'spam': 2, 'eggs': 3}

A two-item dictionary: keys 'spam' and 'eggs'.

adict = { 'info': { 42: 1, type("): 2 }, 'spam': [] }

Nested dictionaries: adict['info'][42] fetches 1.

adict = dict(zip('abc', [1, 2, 3]))

Creates a dictionary by passing the key/value tuples list

to the type constructor.

adict = dict(name='Bob', age=42, job=('mgr', 'dev'))

Creates a dictionary by passing keyword arguments to

the type constructor.

Operations

Operations comprise all mapping operations (see Table 5,

earlier in this book), plus the following dictionary-specific

methods.

Specific Built-in Types

|

27

adict.has_key(k)

k in adict

Returns True if adict has a key k,orFalse otherwise.

Since Python