Detailed introduction to the struct.pack() function and struct.unpack() function in python-Python Tutorial-php.cn

Detailed introduction to the struct.pack() function and struct.unpack() function in python

高洛峰

Release： 2017-03-06 14:40:02

Original

3575 people have browsed it

The struct in python is mainly used to process C structure data. When reading, it is first converted to Python's string type, and then converted to Python's structured type, such as tuple. Generally, the input channels come from files or network binary streams.

　 1.struct.pack() and struct.unpack()

During the conversion process, a format strings is mainly used. Use To specify the conversion method and format.

Let’s talk about the main methods:

1.1 struct.pack(fmt,v1,v2,...)

Put v1, v2 and other parameters The value is wrapped in one layer, and the wrapping method is specified by fmt. The wrapped parameters must strictly conform to fmt. Finally returns a wrapped string .

　1.2 struct.unpack(fmt,string)

　As the name suggests, unpack. For example, pack is packaged, and then unpacked can be used to unpack. Returns a tuple (tuple) obtained by unpacking data (string), even if there is only one data, it will be unpacked into a tuple. Among them, len(string) must be equal to calcsize(fmt), which involves a calcsize function. struct.calcsize(fmt): This is used to calculate the size of the structure described in the fmt format.

The format string consists of one or more format characters. For the description of these format characters, refer to the Python manual as follows

##xpad byte no valueccharstring of length 1bsignedcharintegerBunsignedchar integer?_Boolbool(1)hshortintegerHunsignedshortintegeriintintegerIunsignedintinteger or longllongintegerLunsignedlonglongqlonglonglong(2)Qunsignedlonglonglong(2)ffloatfloatddoublefloatschar[]stringpchar[]stringPvoid*long

Format	c Type	Python	#Note

　2. Code example

import struct 

# native byteorder 
buffer = struct.pack("ihb", 1, 2, 3) 
print repr(buffer) 
print struct.unpack("ihb", buffer) 

# data from a sequence, network byteorder 
data = [1, 2, 3] 
buffer = struct.pack("!ihb", *data)
print repr(buffer) 
print struct.unpack("!ihb", buffer)

 

 

Output:

'\x01\x00\x00\x00\x02\x00\x03'
(1, 2, 3)
'\x00\x00\x00\x01\x00\x02\x03'
(1, 2, 3)

Copy after login

View Code

First, pack the parameters 1, 2, and 3. Before packing, 1, 2, and 3 obviously belong to the integer in the python data type. After packing, they become a C-structured binary string and are converted into the python string type for display. '\x01\x00\x00\x00\x02\x00\x03'. Since this machine is little endian ('little-endian', please refer to here for the difference between big endian and little endian, so the high bits are placed in the low address segment. i represents the int type in the C struct, so this machine occupies 4 bits, 1 Expressed as 01000000; h represents the short type in the C struct, occupying 2 bits, so it is expressed as 0200; similarly, b represents the signed char type in the C struct, occupying 1 bit, so it is expressed as 03.

Others. The conversion of the structure is also similar. For some special ones, you can refer to the Manual of the official document.

At the beginning of the Format string, there is an optional character to determine the big endian and little endian. The list is as follows:

##@=<##>big-endianstandard!network (= big-endian)standard

If not appended, the default is @. Even if the native character order (big endian or little endian) is used, the size of the C structure and the alignment in memory are consistent with the native machine (native). For example, The integer of some machines is 2 bits, while some machines have four bits; some machine memories have four bits aligned, and some have n bits aligned (n is unknown, and I don't know how many).

There is also a standard option, which is described as: If standard is used, there will be no memory alignment for any type.

For example, in the second half of the applet just now, the first bit in the format string used is! , which is the standard alignment of big endian mode, so the output is '\x00\x00\x00\x01\x00\x02\x03', in which the high bit itself is placed in the high address bit of the memory.

For more detailed introductions to the struct.pack() function and struct.unpack() function in Python, please pay attention to the PHP Chinese website!


native	native
native	standard
little-endian	standard