Libmem

Added APIs related to demangled symbols:

LM_DemangleSymbol
LM_FreeDemangleSymbol
LM_EnumSymbolsDemangled
LM_FindSymbolAddressDemangled

Instead of working with normal symbols, these APIs first resolve the symbols.
For example, the symbol `_Z15_LM_EnumSymbolsP11lm_module_tPFiP11lm_symbol_tPvES3_` gets resolved to ` _LM_EnumSymbols(lm_module_t*, int (*)(lm_symbol_t*, void*), void*)`, which is significantly more readable.

This version does not have many improvements. The main thing that changed is that some Rust APIs are now declared as `unsafe` (they've always been `unsafe`, but they were not marked as such). Is this a good reason for a version bump? I think so, because it wouldn't make sense for libmem version `4.0.0` not to be compatible with libmem version `4.0.1`, for example. So I think `4.1.0` was necessary.
If you need Windows binaries, use version `4.0.0`. Nothing has changed in the C API, only in Rust.

Finally, version 4.0 has arrived.
This version has significant improvements when compared to the last release.
To name a few:
- Added Assembler/Disassembler
- Added VMT Hooking API
- Modulated libmem
- Added Rust support
- Simplified API
- Removed overkill features
- Fixed bugs
And much more stuff.

I believe this is as good as libmem has ever been, there are still going to be a few more updates but this project is getting to a happy ending.

4.0.0-p4
Backup before readding `lm_process_t` and storing the path, bits and name on it.
The same thing with `lm_module_t`.

4.0.0-p3
libmem has been modulated and cleaned up, APIs renamed etc.
Python support is broken in this version, because I haven't adapted the APIs yet.
This is a backup before switching to PID only calls, no more `lm_process_t`, just like it was before.
For more information, see: https://github.com/rdbo/libmem/issues/67

4.0.0-p2
Added a disassembler (capstone) and an assembler (keystone)
This is a backup before libmem's modulation

4.0.0-p1
Backup of the current state of libmem

This version is a backup. I plan on re-re-remaking libmem and make it as good as I possibly can on a 4.X version.
There are a few problems in the current version of libmem, which I think can be avoided. Some examples:
1. Forcing developers to free memory that they didn't allocate
2. Bad prefix MEM (already being used by Windows. For example, MEM_COMMIT)
3. Useless data and types, such as MEM_PAYLOADS, mem_asm_t, mem_load_t, etc.

The list goes on.

I need a memory library that comes close to perfection, and the current state of libmem is far from that.
Hopefully on the next version, there will be some major improvements.

Fixed multiple warnings.

Fixed signature scanning