struct.pack fmt Documentation

When working with hex protocols in Python, you always end up using struct.pack. There are so many format variations that I simply can't remember them all. For example:

struct.pack('>H', checksum)
data = struct.pack('>I', id) 

1. Byte Order and Alignment (First Character)

The first character of the format string usually indicates byte order and alignment:

2. Data Types and Sizes

3. Special Symbols

4. Common Packing Examples

1. Pack a big-endian 4-byte unsigned integer plus a double-precision float:

   PythonCopy

   data = struct.pack('>Id', 0x1234, 3.14)
   

2. Pack a little-endian struct with padding (e.g., int + char, 4-byte alignment):

   PythonCopy

   data = struct.pack('<ic', 42, b'A')  # 自动填充 1 字节
   

3. Pack a fixed-length string:

   PythonCopy

   data = struct.pack('10s', b'hello')  # 补零到 10 字节
   

4. Pack a boolean plus an unsigned short (network order):

   PythonCopy

   data = struct.pack('!?H', True, 0xABCD)
   

5. Notes

1. Value range validation:

   • For example, B (0–255) and I (0–0xFFFFFFFF) raise an error when out of range.

   • It's a good idea to check the value range before packing:

     PythonCopy

     if not (0 <= value <= 0xFFFF):
     raise ValueError("数值超出范围")
     

2. Platform differences:

   • l and L are 4 bytes on 32-bit systems, and may be 8 bytes on 64-bit systems.
   • Use i or I to ensure a fixed 4 bytes.

3. String handling:

   • The s format requires an explicit length (e.g., 10s).
   • If you need a dynamic-length string, combine it with len(): python s = b'hello' data = struct.pack(f'I{len(s)}s', len(s), s)