Repository navigation
Releases: mhamilt/max-byte-cast
Releases · mhamilt/max-byte-cast
Release list
Compile for Apple Silicon
Long Type modifier and Byte Casting
This version adds the ability to cast data back to bytes under b mode. It also adds the l modifier for dealing with long types (e.g. long int and double
| Format | Data Type |
|---|---|
| i | 32-bit int |
| h | 16-bit int |
| f | 32-bit float |
| s | ASCII string (greedy) |
| b | byte output |
| Modifier | Effect |
|---|---|
u |
unsigned (i and h only) |
l |
long type, i: int64 f: double |
> |
Big endian (append) |
< |
Little endian (append) |
Windows and macOS: All data types
This release covers all the standard data types:
| Format | Data Type |
|---|---|
| i | 32-bit int |
| h | 16-bit int |
| f | 32-bit float |
| s | string until end |
Modifiers:
u: unsigned (iandhonly)>: Big endian (append)<: Little endian (append)
A toggle has been added for easy endian switch.
Minimal Working macOS version
A simple implementation of a type casting max object. Currently implements
- A single 32-bit int
- A single 32-bit float
- A greedy string