Python deep copy, shallow copy

In Python, there is a difference between the assignment of an object and the copy (deep/shallow copy), which can produce unexpected results if used without notice.

First of all, we should have the following knowledge about the assignment operation:

1. Assignment is to assign the address of an object to a variable, so that the variable points to that address (old bottles of old wine).
2. Modify immutable Objects ( str , tuple ) need to open up new space
3. modifying mutable objects ( list etc.) no need to open up new space

Shallow copy

? A shallow copy copies only the addresses of the elements in the container, copies the references, and does not copy the contents.

In [1]: a = [11,22,33]In [2]: id(a)Out[2]: 22132096In [3]: b = aIn [4]: id(b)Out[4]: 22132096 # 引用相同In [5]: a.append(44)In [6]: aOut[6]: [11, 22, 33, 44]In [7]: bOut[7]: [11, 22, 33, 44] # b的值也变化

Deep copy

A deep copy is a copy (recursive) for all levels of an object

in [8]:ImportCopyIn [9]: A=[ One, A, -]in [Ten]:ID(a) out[Ten]:74598640in [ One]: b=Copy.deepcopy (b) in [ A]:ID(b) out[ A]:79142624in [ -]: A.append ( -) in [ -]: aout[ -]: [ One, A, -, -]in [ the]: bout[ the]: [ One, A, -, -]

Further understanding

in [ at]: A=[ One, A, -]in [ -]: b=[ -, -, the]in [ -]: C=(A, B) in [ -]: E=Copy.deepcopy (c) in [ -]: A.append ( the) in [ -]: aout[ -]: [ One, A, -, the]in [ in]: bout[ in]: [ -, -, the]in [ -]: cout[ -]: ([ One, A, -, the], [ -, -, the]) in [ to]: eout[ to]: ([ One, A, -], [ -, -, the])# Copy.copy () methodin [ +]: F=Copy.copy (c) in [ -]: A.append ( the) in [ the]: aout[ the]: [ One, A, -, the, the]in [ *]: bout[ *]: [ -, -, the]in [ $]: cout[ $]: ([ One, A, -, the, the], [ -, -, the]) in [Panax Notoginseng]: eout[Panax Notoginseng]: ([ One, A, -], [ -, -, the]) in [ -]: fout[ -]: ([ One, A, -, the, the], [ -, -, the])

Other ways to copy a shallow copy differs from a copy of a immutable type and a mutable type

in [ the]: A=[ One, A, -]in [ the]: b=Copy.copy (a) in [ -]:ID(a) out[ -]:59275144in [ the]:ID(b) out[ the]:59525600in [ the]: A.append ( -) in [ the]: aout[ the]: [ One, A, -, -]in [94]: bout[94]: [ One, A, -]in [ the]: A=( One, A, -) in [ the]: b=Copy.copy (a) in [ the]:ID(a) out[ the]:58890680in [98]:ID(b) out[98]:58890680

• A shard expression can assign a sequence of values

    ="abc"    = a[:]

• The copy method of the dictionary can copy a dictionary

    =dict(name="zhangsan", age=27)    = d.copy()

• Some built-in functions can generate copies (list)

    =list(range(10))    =list(a)

• copy function in the copy module

    import copy    = (1,2,3)    = copy.copy(a)

Python deep copy, shallow copy