Bytes in Python

Binary data in Python has type bytes.

We can construct binary data by prefixing a string with b.

For example:

assert isinstance(b"a", bytes)

Each character of the string is interpreted as an ASCII character.

We can convert the binary data to an 8-bit integer:

assert int.from_bytes(b"a", byteorder="big", signed=True) == 97

And we can convert the integer back into binary data:

assert (97).to_bytes(length=1, byteorder="big", signed=True) == b"a"

We can also construct binary data by using hexadecimal characters prefixed by \x instead of ASCII characters:

assert isinstance(b"\x05", bytes)

We convert the binary data to an integer:

assert int.from_bytes(b"\x05", byteorder="big", signed=True) == 5

And can convert the integer back to binary data:

assert (5).to_bytes(length=2, byteorder="big", signed=True) == b"\x00\x05"

If we try to treat a negative integer as unsigned, we get an error:

import pytest
with pytest.raises(OverflowError):
	(-5).to_bytes(length=2, byteorder="big", signed=False)

When we index into binary data, we get an unsigned 8-bit integer, not binary data:

assert (b"hello")[0] == 104
assert [i for i in b"hello"] == [104, 101, 108, 108, 111]

When we slice into binary data, we get binary data:

assert (b"hello")[0:1] == b"h"

We can convert an integer to a binary string (for the mathematical representation rather than the storage representation):

assert bin(5) == "0b101"
assert bin(-5) == "-0b101"

And a binary string back to an integer:

assert int("0b101", 2) == 5
assert int("-0b101", 2) == -5

Similarly, we can convert an integer to a hexadecimal string:

assert hex(5) == "0x5"
assert hex(-5) == "-0x5"

And a hexadecimal string back to an integer:

assert int("0x5", 16) == 5
assert int("-0x5", 16) == -5

Note that int interprets the string (without the negative sign) as an unsigned integer.

We can also use the struct module to convert Python values to binary data and vice versa.

For example:

import struct
assert struct.pack("<B", 97) == b"a"
assert struct.unpack("<B", b"a") == (97,)

We can also create integers using the 0b or 0x prefixes directly:

assert 0b101010 == 42
assert 0x2a == 42

The bytes function creates an immutable bytes object and the bytearray function creates a mutable one:

assert bytes([97, 98, 99]) == b"abc"
assert bytearray([97, 98, 99]) == b"abc"

We can encode a string as a bytes object using encode and decode it using decode:

assert "hello".encode("utf-8") == b"hello"
assert b"hello".decode("utf-8") == "hello"

Provide feedback