Share Lineage 2 GF AI Decompiler

[verbrannt]

BTW tree splitting produces too long paths in Windows, so it can't handle it properly =/
Max path length in Windows is 260 or something, so I'll limit tree depth.

Edited July 15, 2018 by verbrannt

[eressea]

1 minute ago, verbrannt said:

BTW tree splitting produces too long paths in Windows, so it can't handle it properly =/
Max path length in Windows is 260 or something.

 

It works for me on Windows 10 with PHP 7.2.7, I just had to enable Win32 long paths - see https://www.ryadel.com/en/enable-ntfs-win32-long-paths-policy-remove-255-260-characters-limit-windows-10/

 

Also I've seen you haven't applied UTF-16le decoding part from the patch and still drop all non-ASCII characters, what's the reason for it?

[verbrannt]

1 minute ago, eressea said:

Also I've seen you haven't applied UTF-16le decoding part from the patch and still drop all non-ASCII characters, what's the reason for it?

 

I'm refactoring main.php right now thats the reason :)

[zconll]

There is 3 functions missing from eressea's compiler.

Quote

"235078277": {
        "name": "SetLocalMap",
        "arguments": [
            "int",
            "int",
            "int"        ],
        "type": "void"
    }

"235012742": {
        "name": "GetLocalMap",
        "arguments": [
            "int",
            "int"        ],
        "type": "int"
    }

"268828804": {
        "name": "DebugLog",
        "arguments": [
            "string",
            "string",
            "string",            "int",
            "int",
            "int"        ],
        "type": "void"
    }

 

 

[eressea]

4 minutes ago, zconll said:

There is 3 functions missing from eressea's compiler.

 

Because they're custom and unique for MyExt64... Maybe add data/gf-myext? :)

[zconll]

Problem with empty string in "MakeFString" function.

After descompiler:

Quote

Say(MakeFString(1000452, , __blankblank, _blank, _blank, _blank));

 

Should be:

 

Quote

Say(MakeFString(1000452, "", "", "", "", ""));

 

 

class  warrior_parameter affected and some more.

[verbrannt]

2 minutes ago, zconll said:

Problem with empty string in "MakeFString" function.

After descompiler:

 

Should be:

 

 

class  warrior_parameter affected and some more.

 

_blank is a property defined in the default_npc class.

string  _blank = "";

 

[zconll]

ok, I never has seen this parameter in GF/freya ai.obj, i need add this parameter to  "default_npc" class, but there a problem after descompiled.

Quote

Say(MakeFString(1000452, , __blankblank, _blank, _blank, _blank));

 

How you can to see second and third parameter show both together, when should be:

 

Quote

Say(MakeFString(1000452, , __blank, _blank, _blank, _blank, _blank));

 

[verbrannt]

5 minutes ago, zconll said:

ok, I never has seen this parameter in GF/freya ai.obj, i need add this parameter to  "default_npc" class, but there a problem after descompiled.

This parameter is from another AI (not default GF/Freya). AdvExt uses it, for example.

 

5 minutes ago, zconll said:

How you can to see second and third parameter show both together, when should be:

 

Please provide part of the ai.obj where this happens.

Edited July 15, 2018 by verbrannt

[zconll]

Sorry mistaken mine i updated last descompiler source and now show correctly all parameters.

 

des

[verbrannt]

New release: https://github.com/madyanov/nasc-decompiler/releases/tag/v1.4

 

• HF Support by Eressea
• GD Support by @ChaosPaladin
• Many NASC generator fixes
• Split result classes into tree
• Non-ASCII symbols support by Eressea

 

Edited July 15, 2018 by verbrannt

[eressea]

--output option disappeared? :)

 

EDIT: Ah, it never was there :D

Edited July 15, 2018 by eressea

[eressea]

Are you sure that GM_... aren't global map IDs?

 

EDIT: For example antharas uses global map ID of 10 which is same as gm_antaras_max... On the other side, it's usually defined as a parameter so it can't get resolved nevertheless...

Edited July 16, 2018 by eressea

[verbrannt]

42 minutes ago, eressea said:

Are you sure that GM_... aren't global map IDs?

It's definitely global map ids :) 

[eressea]

2 minutes ago, verbrannt said:

It's definitely global map ids :) 

 

I'm asking because of commit https://github.com/madyanov/nasc-decompiler/commit/87c177446322d3c2688ec3e53c75b216f50d2501