Tools for the files used by Higurashi Mei
List files are encrypted using the algorithm shown in Crypto.swift with the key 0x215901a9b1a1c553. hash.dat contains the fnv hash of the decrypted data.
| Name | Type |
|---|---|
| NumGroups | UInt32 |
| Groups | Group[NumGroups] |
| NumItems | UInt32 |
| Items | Item[NumItems] |
| Name | Type |
|---|---|
| Size | UInt32 |
| Name | CString |
| Name | Type | Notes |
|---|---|---|
| Name | CString | |
| Group | UInt16 | Offset into main file's Groups array |
| ID | UInt32 | |
| Pos | UInt64 | Position in data.bin |
| Size | UInt32 | Length in data.bin |
| Hash | UInt64 | FNV hash of decrypted data, and also the encryption key used |
These files are stored with the extension .bin and have Sx1.05 as their file magic
| Name | Type | Notes |
|---|---|---|
| Magic | CString | Should be Sx1.05 |
| Type | UInt8 | Image encoding, 1 => ARGB32, 2 => PVRTC RGB4, 3 => ETC RGB4, 4 => DXT1, 5 => DXT5 |
| UInt64 | Purpose unknown | |
| Width | UInt16 | |
| Height | UInt16 | |
| ImageSize | UInt32 | Decompressed size of Image |
| Image | Compressed Data | See below, keep reading until the decompressed data is ImageSize bytes |
| ImageASize | UInt32 | Decompressed size of ImageA, only present for Types 2 and 3 |
| ImageA | Compressed Data | Same as Image, only present for types 2 and 3 |
| NumSections | UInt32 | |
| Sections | Section[NumSections] |
For files with an ImageA (types 2 and 3), Image contains the final image's red and green channels, while ImageA contains its blue and alpha, therefore to create the final image, use the following mapping:
- Final.R = Image.R
- Final.G = Image.G
- Final.B = ImageA.R
- Final.A = ImageA.B
To decompress data:
- Read an Int16
- If the value from ① is greater than or equal to zero, read that number of bytes into the output buffer and go to ①
- Read a UInt16
- Copy ③ bytes starting at ① bytes away from the end of the output buffer and go to ①
| Name | Type |
|---|---|
| Name | CString |
| Width | UInt16 |
| Height | UInt16 |
| x0 | Float32 |
| y0 | Float32 |
| x1 | Float32 |
| y1 | Float32 |
| u0 | Float32 |
| v0 | Float32 |
| u1 | Float32 |
| v1 | Float32 |
x and y seem to always be 0 and 1, they probably map how the image should be drawn onto an image of size Width x Height
u and v are the texture coordinates from the main file's image texture