Add TMPFakeKanjiAdder

scripts/CharacterInfoExtraction/TMPFakeKanjiAdder.py

@@ -0,0 +1,203 @@
+import sys
+import os
+import struct
+
+class DataScanner:
+	def __init__(self, data):
+		self.offset = 0
+		self.data = data
+
+	def advance(self, length):
+		self.offset += length
+
+	def read(self, length):
+		output = self.peek(length)
+		self.advance(length)
+		return output
+
+	def peek(self, length):
+		return self.data[self.offset:(self.offset + length)]
+
+	def readString(self):
+		length = int.from_bytes(self.peek(4), byteorder="little")
+		length += (4 - length) % 4
+		return self.read(length + 4)
+
+	def rest(self):
+		return self.data[self.offset:]
+
+def readString(str):
+	length = int.from_bytes(str[0:4], byteorder="little")
+	return str[4:(4+length)].decode("utf-8")
+
+class FontFile:
+	def __init__(self, data, filename, emptyAtlasPoint=None):
+		stringLength = int.from_bytes(data[(4 * 15):(4 * 16)], byteorder="little")
+		if stringLength in range(4, 26): # Assumes filenames are between 4 and 25 chars
+			print(f"Detected {filename} as coming from TextMesh Pro")
+			self.type = "TMP"
+			self._fromAssetCreator(data, filename, emptyAtlasPoint)
+		else:
+			stringLength = int.from_bytes(data[(4*12):(4*13)], byteorder="little")
+			if stringLength == 1:
+				print(f"Detected {filename} as coming from a newer Higurashi game")
+				self.type = "Hima"
+			else:
+				print(f"Detected {filename} as coming from an older Higurashi game")
+				self.type = "Oni"
+			self._fromGame(data, filename, emptyAtlasPoint)
+		self.filename = readString(self.Filename)
+
+	def _readArray(self, data, itemLength, emptyAtlasPoint):
+		self.Array = data.read(4)
+		length = int.from_bytes(self.Array, byteorder="little")
+		atlasWidth = struct.unpack("<f", self.AtlasWidth)[0]
+		atlasHeight = struct.unpack("<f", self.AtlasHeight)[0]
+		for _ in range(length):
+			info = bytearray(data.read(4 * 8))
+			if itemLength > 8:
+				data.advance((itemLength - 8) * 4)
+			x, y, width, height = struct.unpack("<ffff", info[4:20])
+			if emptyAtlasPoint is not None:
+				if width == 0 and height == 0:
+					charid = int.from_bytes(info[:4], byteorder="little")
+					print(f"Relocating 0x0 character U+{charid:x} to ({emptyAtlasPoint[0]}, {emptyAtlasPoint[1]})")
+					info[4:12] = struct.pack("<ff", emptyAtlasPoint[0], emptyAtlasPoint[1])
+			if x > atlasWidth:
+				info[4:8] = self.AtlasWidth
+			if y > atlasHeight:
+				info[8:12] = self.AtlasHeight
+			self.Array += info
+
+	def _fromGame(self, data, filename, emptyAtlasPoint):
+		data = DataScanner(data)
+		self.Header = data.read(4 * 7)
+		self.Filename = data.readString()
+		if self.type == "Hima":
+			self.BeforeFontName = data.read(4)
+		else:
+			self.BeforeFontName = bytes()
+		self.FontName = data.readString()
+		# Font face data
+		self.PointSize = data.read(4)
+		self.Padding = data.read(4)
+		self.LineHeight = data.read(4)
+		self.Baseline = data.read(4)
+		self.Ascender = data.read(4)
+		self.Descender = data.read(4)
+		self.CenterLine = data.read(4)
+		self.SuperscriptOffset = data.read(4)
+		self.SubscriptOffest = data.read(4)
+		self.SubSize = data.read(4)
+		self.Underline = data.read(4)
+		self.UnderlineThickness = data.read(4)
+		self.TabWidth = data.read(4)
+		self.CharacterCount = data.read(4)
+		self.AtlasWidth = data.read(4)
+		self.AtlasHeight = data.read(4)
+
+		self.AfterFontFace = data.read(4 * 8)
+		self._readArray(data, 8, emptyAtlasPoint)
+		self.Footer = data.rest()
+
+	def _fromAssetCreator(self, data, filename, emptyAtlasPoint):
+		data = DataScanner(data)
+		self.Header = data.read(4 * 15)
+		self.Filename = data.readString()
+		self.BeforeFontName = data.read(4 * 7)
+		self.FontName = data.readString()
+		# Font face data
+		self.PointSize = data.read(4)
+		self.Scale = data.read(4)
+		self.CharacterCount = data.read(4)
+		self.LineHeight = data.read(4)
+		self.Baseline = data.read(4)
+		self.Ascender = data.read(4)
+		self.CapHeight = data.read(4)
+		self.Descender = data.read(4)
+		self.CenterLine = data.read(4)
+		self.SuperscriptOffset = data.read(4)
+		self.SubscriptOffest = data.read(4)
+		self.SubSize = data.read(4)
+		self.Underline = data.read(4)
+		self.UnderlineThickness = data.read(4)
+		self.strikethrough = data.read(4)
+		self.strikethroughThickness = data.read(4)
+		self.TabWidth = data.read(4)
+		self.Padding = data.read(4)
+		self.AtlasWidth = data.read(4)
+		self.AtlasHeight = data.read(4)
+
+		self.AfterFontFace = data.read(4 * 3)
+		self._readArray(data, 9, emptyAtlasPoint)
+		self.Footer = data.rest()
+
+	def toBytes(self):
+		out = bytes()
+		out += self.Header
+		out += self.Filename
+		out += self.BeforeFontName
+		out += self.FontName
+
+		out += self.PointSize
+		out += self.Padding
+		out += self.LineHeight
+		out += self.Baseline
+		out += self.Ascender
+		out += self.Descender
+		out += self.CenterLine
+		out += self.SuperscriptOffset
+		out += self.SubscriptOffest
+		out += self.SubSize
+		out += self.Underline
+		out += self.UnderlineThickness
+		out += self.TabWidth
+		out += self.CharacterCount
+		out += self.AtlasWidth
+		out += self.AtlasHeight
+
+		out += self.AfterFontFace
+		out += self.Array
+		out += self.Footer
+
+		return out
+
+	def fakeChars(self, charlist):
+		zeroSizeCharInfo = None
+		oldSize = int.from_bytes(self.Array[0:4], byteorder="little")
+		existingChars = set()
+		for i in range(oldSize):
+			index = 4 + (i * 4 * 8)
+			unicodescalar = int.from_bytes(self.Array[index:index+4], byteorder="little")
+			existingChars.add(unicodescalar)
+			if unicodescalar == 0x3000:
+				zeroSizeCharInfo = bytearray(self.Array[index:index+32])
+		if zeroSizeCharInfo is None:
+			print("Starting character set should contain a U+3000 ideographic space to base fake kanji info off of.")
+			return;
+		newSize = oldSize
+		newArray = bytearray(self.Array)
+		for char in charlist:
+			if char in existingChars:
+				continue
+			zeroSizeCharInfo[0:4] = char.to_bytes(4, byteorder="little")
+			newArray.extend(zeroSizeCharInfo)
+			existingChars.add(char)
+			newSize += 1
+		finalSizeInfo = newSize.to_bytes(4, byteorder="little")
+		newArray[0:4] = finalSizeInfo
+		self.CharacterCount = finalSizeInfo
+		self.Array = bytes(newArray)
+
+if len(sys.argv) <= 3:
+	print("Usage: " + sys.argv[0] + " monobehaviour.dat charset.txt outputMonoBehaviour.dat")
+
+with open(sys.argv[1], "rb") as tmp:
+	tmpAsset = FontFile(tmp.read(), sys.argv[1])
+
+with open(sys.argv[2], "r", encoding="utf-8") as charsetFile:
+	charsetText = charsetFile.read()
+	chars = [ord(x) for x in charsetText]
+tmpAsset.fakeChars(chars)
+with open(sys.argv[3], "wb") as output:
+	output.write(tmpAsset.toBytes())