A very efficient object enumeration method~

[qop1832]

This method has a qualitative leap compared to traditional object traversal, even in languages like Python that are far less efficient than C, traversing 100 objects only takes about 0.002 seconds

import pymem


def enum_object():
    obj_dict = {}
    chain_entry = pm.read_longlong(pm.base_address + 0x31c5600) + 0x120
    cur_chain_ptr = pm.read_longlong(chain_entry)
    if cur_chain_ptr == chain_entry:
        return obj_dict
    while True:
        object_entry = cur_chain_ptr - 0x68
        object_type = pm.read_bytes(object_entry + 0x10, 1)[0]
        guid = int.from_bytes(pm.read_bytes(object_entry + 0x18, 0x10), "little", signed=False)
        obj_dict[guid] = (object_type, hex(object_entry))
        cur_chain_ptr = pm.read_longlong(cur_chain_ptr)
        if cur_chain_ptr == chain_entry or cur_chain_ptr == 0:
            return obj_dict


pm = pymem.Pymem("WowClassic.exe")
object_dict = enum_object()

obj_list.png

[scizzydo]

Isn't what you're doing there already the traditional way? I'm confused

[qop1832]

Maybe I'm lagging behind,,, before this I saw that almost all (including myself) object traversal requires the use of Count, and then
(Array + 0x8 * i) to traverse, this often requires nested loops, and does not jump out until enough Count has been read. This occasionally causes some problems and is very time-consuming. However, this does not require Count, and the code looks more concise and More efficient.

[Razzue]

Edit** (foot in mouth moment)

You're just iterating over the enum visible objects TsList While quicker, is still not the fastest.

internal Entity Search(Guid input)
{
    var result = new Entity();
    try
    {
        if (input.IsEmpty()) return result;


        var hash = unchecked((0xA2AA033B * input.Value2) + (0xD6D018F5 * input.Value1));


        var index = hash & (tsArray.Maximum - 1);
        var pointer = tsArray.Entry<IntPtr>(index);


        if (pointer.ToInt64() == 0) return default;
        var wowObject = Memory.Read<Entity>(pointer);
        do
        {
            if (wowObject.Guid.Equals(input))
            {
                result = wowObject;
                break;
            }


            if (wowObject.Next.ToInt64() == 0) break;
            Memory.Read(wowObject.Next, ref wowObject);
        } while (true);
    }
    catch (Exception e) { Console.WriteLine(e); }
    return result;
}

Is a variation on chases Guid search method they posted a few months back, Is still the quickest call I've used for fetching entities at a whim.
Using this I've all but abolished most of my entity caching (no point when I can grab specific entity in < 0.001 ms)

Similar hashes also happen for WDB entry access, and DB2 field access internally (though different keys)

[qop1832]

Does this method pass in the GUID to return the object pointer? Or am I understanding it wrong?~ If this is the case, is there any efficient way to get the GUID of all visible objects?