Compact Variable Length Integer
A variable-width data format for unsigned integers allowing a more compact representation for smaller values. This is sometimes referred to as "compact size" or "var int".
Numbers are encoded with the first rule that applies of the following:
- If the number < 253 (0xFD), store it in 1 byte, left-padded with zeros.
- If the number fits in 16 bits (but is greater than 252), store it in 3 bytes: a 1-byte value 253 (0xFD) followed by the 2 byte little-endian number.
| Byte Index |
C-Style Calculation |
| 0 |
0xFD |
| 1 |
value & 255 |
| 2 |
value >> 8 |
- If the number fits in 32 bits (but not 8 or 16), store it in 5 bytes: a 1-byte value 254 (0xFE) followed by the 4 byte little-endian number
| Byte Index |
C-Style Calculation |
| 0 |
0xFE |
| 1 |
value & 255 |
| 2 |
(value >> 8) & 255 |
| 3 |
(value >> 16) & 255 |
| 4 |
(value >> 24) & 255 |
- If the number fits in 64 bits (but not 8, 16, or 32), store it in 9 bytes: a 1-byte value 255 (0xFF) followed by the 8 byte little-endian number
| Byte Index |
C-Style Calculation |
| 0 |
0xFF |
| 1 |
value & 255 |
| 2 |
(value >> 8) & 255 |
| 3 |
(value >> 16) & 255 |
| 4 |
(value >> 24) & 255 |
| 5 |
(value >> 32) & 255 |
| 6 |
(value >> 40) & 255 |
| 7 |
(value >> 48) & 255 |
| 8 |
(value >> 56) & 255 |