[Dev] Map Structure

Quote:

---

Originally Posted by DragonLord

This is getting really sad at this point.
You guys have the source available and you're making all this enormous mess and pointless threads.

Every .dat file is essentially an .obj container.
By default, the first entry of a .dat file is the map info itself, and it consists of the following:

Code:

---

struct MAP_DATA
{
    float    fTileSize;
    int        nVertexNumber;
    int        nTileInfoNumber;
    int        nObjectNumber;
};

struct VERTEXINFO
{
    float[3]    pos;
    float[3]    nor;
};

struct TILEINFOCLIENT
{
    int                useTexNumber;
    unsigned int    dwWaterType;
    int                waterHeight;
    short            waterTexNumber;
    unsigned char    bMove;
    int                bEnableLand;
    unsigned char    bEventType;
}

struct OBJECTINFOCLIENT
{
    unsigned int    dwObjType;
    int                nObjCount;
    float[3]        vPos;
    float[3]        vVel;
    float[3]        vUp;
    unsigned int    dwObjectMonsterUniqueNumber;
    unsigned char    bObjectTexIndex;
    unsigned int    nObjectSrcIndex;
    unsigned char    bBodyCondition;
    int                bIsEvent;
    unsigned char    bEventType;
    short            sEventIndexFrom;
    short            sEventIndexTo;
    short            sEventIndex3;
    char            strEventParam[40];
}

---

MAP_DATA, followed by VERTEXINFO * nVertexNumber, followed by TILEINFOCLIENT * nTileInfoNumber, and finally OBJECTINFOCLIENT * nObjectNumber.

Was it really so hard to gather all this information from the source?

---

The problem here is that i see a lot of guys reading the structures value by value like this:

Code:

---

using (BinaryReader BR = new BinaryReader(Path, Encoding.ASCII))
{
    MyStruct str = new MyStruct();
    str.x = BR.ReadInt etc etc.;
    str.j = BR.ReadBoolean etc etc;
}

---

and after complaining because it dont work <_<

Are you able to post the .dat extension?

Quote:

---

Originally Posted by St34lth4ng3l

...
I started my own one in Unity, but got stuck on model positioning, rotation etc.
http://fs5.directupload.net/images/160103/yxzguxlm.jpg

---

Out of sheer curiosity, did you write your own DirectX X file parser or did you use the Microsoft one through .NET's DirectX API?

Because if you did and it also supports correctly reading/parsing animations, I could trade you the correct rotation/translation matrices you're missing for it.

Quote:

---

Originally Posted by ThatGuy

Out of sheer curiosity, did you write your own DirectX X file parser or did you use the Microsoft one through .NET's DirectX API?

Because if you did and it also supports correctly reading/parsing animations, I could trade you the correct rotation/translation matrices you're missing for it.

---

I'm using an external program during runtime to export them, because natively unity only supports a few meshtypes.

Quote:

---

Originally Posted by St34lth4ng3l

I'm using an external program during runtime to export them, because natively unity only supports a few meshtypes.

---

meshconvert from the Direct X SDK?

if i didn't make this simple wrong, you are would not to share your informations.
:)

Quote:

---

Originally Posted by Murat Yereseren

if i didn't make this simple wrong, you are would not to share your informations.
:)

---

Are you bleeping kidding me!?
It's as easy as reading the goddamn code!
Don't try to pin the blame on people like just because you're a blithering idiot!

just if anyone need to know how to read .dat files in c# (.net).

Attachment 155337

bergi, BOOL in C++ is an int, not a bool type.
So you're looking at 4 bytes versus one byte.
As well as char being 2 bytes due to unicode. I recommend using byte instead.

Quote:

---

Originally Posted by DragonLord

bergi, BOOL in C++ is an int, not a bool type.
So you're looking at 4 bytes versus one byte.
As well as char being 2 bytes due to unicode. I recommend using byte instead.

---

This is all about Marshaling in C# really. You can apply native / unmanaged type attributes like he did with:

Code:

---

[MarshalAs(UnmanagedType.ByValArray, SizeConst = 40)]

---

or optionally

Code:

---

[MarshalAs(UnmanagedType.Bool)]

---

which will make a C# bool always a 4-byte integer. This is the default btw. so it's not necessary to specify it explicitly!

The struct posted by @bergi9 works because:

• The default Marshaling for Boolean Types marshals every C# bool to a 32-bit integer
• He applied a ByValArray Type with a constant size marshal tag to the "strEventParam" Array