Gente duvida
34 views
Skip to first unread message
philipe rodrigues
Oct 13, 2021, 8:31:17 AM10/13/21
to Python Brasil
Como posso extrair um .dict pra formato txt? ou como posso editar um texto que está armazenado em um .dict pra que o .data dele tenha as informações de texto mudadas, por exemplo eu preciso traduzir um jogo que quando extraie as pastas estavam todos casados em .data e .dict, mas como nunca programei em python eu fiquei perdido, se alguém entender ou puder ajudar, agradeceria do fundo do coração.
Marcelo Valle
Oct 13, 2021, 9:55:28 AM10/13/21
to Python Brasil
Você pode usar o modulo json.
Exemplo:
>>> d = {"nome1": 35, "nome2": "abc"}
>>> d
{'nome1': 35, 'nome2': 'abc'}
>>> import json
>>> json.dumps(d)
'{"nome1": 35, "nome2": "abc"}'
>>>
On Wed, 13 Oct 2021 at 13:31, philipe rodrigues <phili...@gmail.com> wrote:
Como posso extrair um .dict pra formato txt? ou como posso editar um texto que está armazenado em um .dict pra que o .data dele tenha as informações de texto mudadas, por exemplo eu preciso traduzir um jogo que quando extraie as pastas estavam todos casados em .data e .dict, mas como nunca programei em python eu fiquei perdido, se alguém entender ou puder ajudar, agradeceria do fundo do coração.
--
--
------------------------------------
Grupo Python-Brasil
https://wiki.python.org.br/AntesDePerguntar
<*> Para visitar o site do grupo na web, acesse:
http://groups.google.com/group/python-brasil
<*> Para sair deste grupo, envie um e-mail para:
python-brasi...@googlegroups.com
---
Você recebeu essa mensagem porque está inscrito no grupo "Python Brasil" dos Grupos do Google.
Para cancelar inscrição nesse grupo e parar de receber e-mails dele, envie um e-mail para python-brasi...@googlegroups.com.
Para ver essa discussão na Web, acesse https://groups.google.com/d/msgid/python-brasil/58549ead-bc66-4926-be97-ac10f4c9d630n%40googlegroups.com.
Antonio Ennio de Jesus
Oct 13, 2021, 9:59:53 AM10/13/21
to python...@googlegroups.com
Olá tente json.dumps(dict) ,
Em qua, 13 13e out 13e 2021 às 9:31, philipe rodrigues<phili...@gmail.com> escreveu:
Como posso extrair um .dict pra formato txt? ou como posso editar um texto que está armazenado em um .dict pra que o .data dele tenha as informações de texto mudadas, por exemplo eu preciso traduzir um jogo que quando extraie as pastas estavam todos casados em .data e .dict, mas como nunca programei em python eu fiquei perdido, se alguém entender ou puder ajudar, agradeceria do fundo do coração.
--
philipe rodrigues
Oct 18, 2021, 1:20:11 PM10/18/21
to Python Brasil
Nas minhas pesquisas eu me deparei com esse JSON, mas a duvida em qual ide eu uso ou é no próprio console do python que eu digito as linhas de código?
Realmente tenho zero conhecimento em python, preciso extrair um dict de 4kb que tá em par com um .data de 500mb. Dentro desse data tem arquivos gráficos, modelos 3d esses eu consigo acessar com uma ferramenta especifica, mas eu não consigo extrair o Strings que é o que importa onde os textos estão armazenados e preciso extrai-los para traduzir e modificar ele pra que o jogo fique em português. Vou por um script que eu usei nesse mesmo software, mas que não extraiu a pasta que eu precisava apenas as outras.
vou colocar o script que extraiu tudo que tem nessa foto, menos os Strings.
Realmente tenho zero conhecimento em python, preciso extrair um dict de 4kb que tá em par com um .data de 500mb. Dentro desse data tem arquivos gráficos, modelos 3d esses eu consigo acessar com uma ferramenta especifica, mas eu não consigo extrair o Strings que é o que importa onde os textos estão armazenados e preciso extrai-los para traduzir e modificar ele pra que o jogo fique em português. Vou por um script que eu usei nesse mesmo software, mas que não extraiu a pasta que eu precisava apenas as outras.
Esse script foi capaz de extrair as pastas, mas não a Strings onde contém os textos. Em outro software eu consigo até visualizar as pastas mas não extrair.
vou colocar o script que extraiu tudo que tem nessa foto, menos os Strings.
from inc_noesis import *
import os
import pickle
import zlib
import math
#Version 1.1
# =================================================================
# Plugin options
# =================================================================
globalPath = None
bLoadMaterials = True
bLoadAnimations = True
#You can use these fields if the skeleton is misplaced for some reason.
skelPosCorrection = NoeVec3([0,0,0])
skelRotCorrection = NoeAngles([0,0,0])
# =================================================================
# Misc
# =================================================================
def registerNoesisTypes():
handle = noesis.register("Luigi's Mansion 3 dictionnary",".dict")
noesis.setHandlerTypeCheck(handle, CheckDictType)
noesis.setHandlerLoadModel(handle, ExtractDict)
handle = noesis.register("Luigi's Mansion 3 texture",".lm3t")
noesis.setHandlerTypeCheck(handle, CheckTextureType)
noesis.setHandlerLoadRGBA(handle, LoadRGBA)
handle = noesis.register("Luigi's Mansion 3 model",".lm3m")
noesis.setHandlerTypeCheck(handle, CheckModelType)
noesis.setHandlerLoadModel(handle, LoadModel)
return 1
def getFileNum(fileNum,isGlobal=False):
if isGlobal:
result = globalPath + os.sep + "File_Data" + os.sep + "global_" + str(fileNum) + '.lm3'
else:
result = filePath + os.sep + os.path.basename(os.path.dirname(filePath)) + '_' + str(fileNum) + '.lm3'
return result
def InitializeFromDict():
global dataFileName
dataFileName = rapi.getInputName()[:-5] + ".data"
if not rapi.checkFileExists(dataFileName):
print("Associated data file not found")
#file paths
global destPath
global textPath
global modelPath
global animPath
global animBundlePath
global filePath
destPath = os.path.dirname(rapi.getInputName()) + os.sep + rapi.getLocalFileName(dataFileName)[:-5]
modelPath = destPath + os.sep + "Models"
textPath = destPath + os.sep + "Textures"
animPath = destPath + os.sep + "Animations"
animBundlePath = destPath + os.sep + "AnimationPacks"
filePath = destPath + os.sep + "File_Data"
if not os.path.exists(modelPath):
os.makedirs(modelPath)
if not os.path.exists(textPath):
os.makedirs(textPath)
if not os.path.exists(animPath):
os.makedirs(animPath)
if not os.path.exists(animBundlePath):
os.makedirs(animBundlePath)
if not os.path.exists(filePath):
os.makedirs(filePath)
def InitializeFromAsset():
global textureList
textureList = []
global modelList
modelList = []
global animationList
animationList = []
global textureHashToIndex
textureHashToIndex = {}
global rootPath
rootPath = os.path.dirname(rapi.getInputName())
rootPath = os.path.dirname(rootPath)
#file paths
global textPath
global modelPath
global animPath
global filePath
modelPath = rootPath + os.sep + "Models"
textPath = rootPath + os.sep + "Textures"
animPath = rootPath + os.sep + "Animations"
animBundlePath = rootPath + os.sep + "AnimationPacks"
filePath = rootPath + os.sep + "File_Data"
if not os.path.exists(modelPath):
os.makedirs(modelPath)
if not os.path.exists(textPath):
os.makedirs(textPath)
if not os.path.exists(animPath):
os.makedirs(animPath)
if not os.path.exists(animBundlePath):
os.makedirs(animBundlePath)
if not os.path.exists(filePath):
os.makedirs(filePath)
def InitializeFileStream(num):
#file streams
if num == 0:
global bs0
if rapi.checkFileExists(getFileNum(0)):
bs0 = NoeBitStream(rapi.loadIntoByteArray(getFileNum(0)))
bs0.setEndian(NOE_LITTLEENDIAN)
else:
print("file 0 not found")
elif num == 52:
global bs52
if rapi.checkFileExists(getFileNum(52)):
bs52 = NoeBitStream(rapi.loadIntoByteArray(getFileNum(52)))
bs52.setEndian(NOE_LITTLEENDIAN)
else:
print("file 52 not found")
elif num == 53:
global bs53
if rapi.checkFileExists(getFileNum(53)):
bs53 = NoeBitStream(rapi.loadIntoByteArray(getFileNum(53)))
bs53.setEndian(NOE_LITTLEENDIAN)
else:
print("file 53 not found")
elif num == 54:
global bs54
if rapi.checkFileExists(getFileNum(54)):
bs54 = NoeBitStream(rapi.loadIntoByteArray(getFileNum(54)))
bs54.setEndian(NOE_LITTLEENDIAN)
else:
print("file 54 not found")
elif num == 63:
global bs63
if rapi.checkFileExists(getFileNum(63)):
bs63 = NoeBitStream(rapi.loadIntoByteArray(getFileNum(63)))
bs63.setEndian(NOE_LITTLEENDIAN)
else:
print("file 63 not found")
elif num == 65:
global bs65
if rapi.checkFileExists(getFileNum(65)):
bs65 = NoeBitStream(rapi.loadIntoByteArray(getFileNum(65)))
bs65.setEndian(NOE_LITTLEENDIAN)
else:
print("file 65 not found")
# =================================================================
# Noesis check type
# =================================================================
def CheckDictType(data):
bs = NoeBitStream(data)
if len(data) < 16:
print("Invalid dict file, too small")
return 0
return 1
def CheckModelType(data):
bs = NoeBitStream(data)
if len(data) < 16:
print("Invalid model file, too small")
return 0
return 1
def CheckTextureType(data):
return 1
# =================================================================
# Classes
# =================================================================
class ChunkType1:
def __init__(self):
self.id = None
self.unk1 = None
self.headerSize = None
self.headerOffset = None
self.dataType = None
self.flags = None
self.chunkSize = None
self.chunkOffset = None
def __str__(self):
return '(ID: '+str(self.id)+', chunkFlag: '+str(self.chunkFlag)+\
', chunkSize: '+str(self.chunkSize)+', chunkOffset: '+str(self.chunkOffset)+\
', dataType: '+str(self.dataType)+ ')'
def parse(self, bs):
self.id = bs.readUShort()
self.unk1 = bs.readUShort()
self.headerSize = bs.readUInt()
self.headerOffset = bs.readUInt()
self.dataType = bs.readUShort()
self.flags = bs.readUShort()
self.chunkSize = bs.readUInt()
self.chunkOffset = bs.readUInt()
class ChunkType2:
def __init__(self):
self.dataType = None
self.chunkFlag = None
self.chunkSize = None
self.chunkOffset = None
def __str__(self):
return '(chunkFlag: '+str(self.chunkFlag)+\
', chunkSize: '+str(self.chunkSize)+', chunkOffset: '+str(self.chunkOffset)+\
', dataType: '+str(self.dataType)+ ')'
def parse(self, bs):
self.dataType = bs.readUShort()
self.chunkFlag = bs.readUShort()
self.chunkSize = bs.readUInt()
self.chunkOffset = bs.readUInt()
class LM3TextureAsset:
def __init__(self):
self.hashName = None
self.headerOffset = None
self.headerSize = None
self.dataOffset = None
self.dataSize = None
class LM3ModelAsset:
def __init__(self):
self.hashName = None
self.meshAssetList = []
self.buffersOffset = -1
self.materialOffsets = []
self.materialMap = None
self.pairedTextureFileIndices = []
self.pairedGlobalTextureFileIndices = []
self.jointList = []
self.parentList = []
self.animationIndices = []
self.boneIDB1ToHash = {}
self.hashToBoneIDB1 = {}
self.boneID71ToHash = {}
self.hashToBoneID71 = {}
self.b001Offset = -1
self.b001Size = -1
self.b003Offset = -1
self.b003Size = -1
self.b004Offset = -1
self.b004Size = -1
self.b005Offset = -1
self.b005Size = -1
self.b006Offset = -1
self.b006Size = -1
self.b007Offset = -1
self.b007Size = -1
self.b102Offset = -1
self.b102Size = -1
self.b103Offset = -1
self.b103Size = -1
self.s7103Offset = -1
self.s7103Size = -1
self.s7105Offset = -1
self.s7105Size = -1
self.s7106Offset = -1
self.s7106Size = -1
class LM3SkeletonAsset:
def __init__(self):
self.pairedModelHashName = None
self.s7103Offset = -1
self.s7103Size = -1
self.s7105Offset = -1
self.s7105Size = -1
self.s7106Offset = -1
self.s7106Size = -1
class LM3AnimationAsset:
def __init__(self):
self.hashName = None
self.dataOffset = None
self.dataSize = None
class LM3AnimationBundleAsset:
def __init__(self):
self.animationIndices =[]
class LM3BoneHeader:
def __init__(self):
self.hash = None
self.index = -1
self.magic = None
self.type = -1
self.opcode = -1
self.offset = -1
class LM3MeshAsset:
def __init__(self):
self.vertexBufferOffset = -1
self.skinningBufferOffset = -1
self.indexBufferOffset = -1
self.indexFormat = None
self.indexCount = -1
self.vertexCount = -1
self.isSkinned = None
# =================================================================
# Load texture
# =================================================================
def LoadRGBA(data, texList):
InitializeFromAsset()
InitializeFileStream(63)
InitializeFileStream(65)
textureAsset = pickle.load(open( rapi.getInputName(), "rb" ))
processTexture([textureAsset])
for tex in textureList:
texList.append(tex)
global bs63, bs65
del bs63,bs65
return 1
# =================================================================
# Load model
# =================================================================
def LoadModel(data, mdlList):
ctx = rapi.rpgCreateContext()
InitializeFromAsset()
modelAsset = pickle.load(open( rapi.getInputName(), "rb" ))
processModel([modelAsset])
for mod in modelList:
mdlList.append(mod)
return 1
# =================================================================
# Data extraction
# =================================================================
def ExtractAssets():
chunkType1List = []
chunkType2List = []
#Texture
textureAssetList = []
textureHashesToFileIndex = {}
textureIndex = -1
#Model
modelAssetList = []
modelHashesToFileIndex = {}
modelIndex = -1
#Skeleton
skeletonAssetList = []
skeletonIndex = -1
#Animation
animationAssetList = []
animationHashesToFileIndex = {}
#Animation bundle
animationBundleAssetList = []
#Material
materialFlag = b'\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\x00\x00\x00\x00\xff\xff\xff\xff\xff\xff\xff\xff\x00\x00\x00\x00'
#file streams
global bs0,bs52,bs53,bs54,bs63,bs65
if rapi.checkFileExists(getFileNum(0)):
bs0 = NoeBitStream(rapi.loadIntoByteArray(getFileNum(0)))
else:
print("file 0 not found")
if rapi.checkFileExists(getFileNum(52)):
bs52 = NoeBitStream(rapi.loadIntoByteArray(getFileNum(52)))
else:
print("file 52 not found")
if rapi.checkFileExists(getFileNum(53)):
bs53 = NoeBitStream(rapi.loadIntoByteArray(getFileNum(53)))
else:
print("file 53 not found")
if rapi.checkFileExists(getFileNum(63)):
bs63 = NoeBitStream(rapi.loadIntoByteArray(getFileNum(63)))
else:
print("file 63 not found")
bs0.setEndian(NOE_LITTLEENDIAN)
bs52.setEndian(NOE_LITTLEENDIAN)
bs53.setEndian(NOE_LITTLEENDIAN)
bs63.setEndian(NOE_LITTLEENDIAN)
#Read chunk entries from file 0
while(True):
flag = bs0.readUShort()
if flag==0x1301:
bs0.seek(-2,1)
chunkType1 = ChunkType1()
chunkType1.parse(bs0)
chunkType1List.append(chunkType1)
else:
bs0.seek(-2,1)
break
while(bs0.tell()+12 <= bs0.getSize()):
chunkType2 = ChunkType2()
chunkType2.parse(bs0)
chunkType2List.append(chunkType2)
del bs0
#Parsing chunkType1
for chunk in chunkType1List:
if chunk.dataType == 0xB000:
bs52.seek(chunk.headerOffset+4)
modelAsset = LM3ModelAsset()
modelAsset.hashName = hex(bs52.readUInt())
modelHashesToFileIndex[modelAsset.hashName] = len(modelAssetList)
modelAssetList.append(modelAsset)
elif chunk.dataType == 0x7000:
bs52.seek(chunk.headerOffset+4)
animationAsset = LM3AnimationAsset()
animationAsset.hashName = hex(bs52.readUInt())
animationAsset.dataOffset = chunk.chunkOffset
animationAsset.dataSize = chunk.chunkSize
animationHashesToFileIndex[animationAsset.hashName] = len(animationAssetList)
animationAssetList.append(animationAsset)
elif chunk.dataType == 0x7100:
bs52.seek(chunk.headerOffset+4)
skeletonAsset = LM3SkeletonAsset()
skeletonAsset.pairedModelHashName = hex(bs52.readUInt())
skeletonAssetList.append(skeletonAsset)
elif chunk.dataType == 0x1302:
bs53.seek(chunk.chunkOffset)
animationBundleAsset = LM3AnimationBundleAsset()
count = bs53.readUInt()
bs53.readUInt()
for j in range(count):
bs53.readUInt()
hash = hex(bs53.readUInt())
if hash in animationHashesToFileIndex:
animationBundleAsset.animationIndices.append(animationHashesToFileIndex[hash])
else:
continue #anim file in global
animationBundleAsset.animationIndices = list(sorted(set(animationBundleAsset.animationIndices)))
animationBundleAssetList.append(animationBundleAsset)
#Processing chunkType2
for chunk in chunkType2List:
#Texture
#Texture Info
if chunk.dataType == 0xB501:
textureIndex += 1
textureAsset = LM3TextureAsset()
bs63.seek(chunk.chunkOffset)
textureAsset.hashName = hex(bs63.readUInt())
textureAsset.headerOffset = chunk.chunkOffset
textureAsset.headerSize = chunk.chunkSize
textureAssetList.append(textureAsset)
#necessary because for some reason some textures may have broken duplicates (see maid and Hellen in Shiny)
if textureAsset.hashName not in textureHashesToFileIndex:
textureHashesToFileIndex[textureAsset.hashName] = textureIndex
# Texture Data
elif chunk.dataType == 0xB502:
textureAssetList[textureIndex].dataOffset = chunk.chunkOffset
textureAssetList[textureIndex].dataSize = chunk.chunkSize
#Model
#Model matrix, used by some models to align to the skeleton
elif chunk.dataType == 0xB001:
modelAssetList[modelIndex].b001Size = chunk.chunkSize
modelAssetList[modelIndex].b001Offset = chunk.chunkOffset
#Submesh info
elif chunk.dataType == 0xB003:
modelAssetList[modelIndex].b003Size = chunk.chunkSize
modelAssetList[modelIndex].b003Offset = chunk.chunkOffset
#Vertex data
elif chunk.dataType == 0xB004:
modelAssetList[modelIndex].b004Size = chunk.chunkSize
modelAssetList[modelIndex].b004Offset = chunk.chunkOffset
#Index buffers
elif chunk.dataType == 0xB005:
modelAssetList[modelIndex].b005Size = chunk.chunkSize
modelAssetList[modelIndex].b005Offset = chunk.chunkOffset
#Material data
elif chunk.dataType == 0xB006:
modelIndex+=1
modelAssetList[modelIndex].b006Size = chunk.chunkSize
modelAssetList[modelIndex].b006Offset = chunk.chunkOffset
#Material additionnal data
elif chunk.dataType == 0xB007:
modelAssetList[modelIndex].b007Size = chunk.chunkSize
modelAssetList[modelIndex].b007Offset = chunk.chunkOffset
#Incomplete boneset coords
elif chunk.dataType == 0xB102:
modelAssetList[modelIndex].b102Size = chunk.chunkSize
modelAssetList[modelIndex].b102Offset = chunk.chunkOffset
#Incomplete boneset hashes
elif chunk.dataType == 0xB103:
modelAssetList[modelIndex].b103Size = chunk.chunkSize
modelAssetList[modelIndex].b103Offset = chunk.chunkOffset
#Skeleton
#Skeleton header
elif chunk.dataType == 0x7101:
skeletonIndex += 1
#Complete boneset coords
elif chunk.dataType == 0x7103:
skeletonAssetList[skeletonIndex].s7103Size = chunk.chunkSize
skeletonAssetList[skeletonIndex].s7103Offset = chunk.chunkOffset
#Hashes to complete boneset ID
elif chunk.dataType == 0x7105:
skeletonAssetList[skeletonIndex].s7105Size = chunk.chunkSize
skeletonAssetList[skeletonIndex].s7105Offset = chunk.chunkOffset
#Parenting info
elif chunk.dataType == 0x7106:
skeletonAssetList[skeletonIndex].s7106Size = chunk.chunkSize
skeletonAssetList[skeletonIndex].s7106Offset = chunk.chunkOffset
for i,textAsset in enumerate(textureAssetList):
pickle.dump( textAsset, open( textPath + os.sep + "texture_" + str(i) +".lm3t", "wb" ))
pickle.dump( textureHashesToFileIndex, open( textPath + os.sep + "textureMap.lm3tMap", "wb" ))
for i,modelAsset in enumerate(modelAssetList):
#B005
modelAsset.buffersOffset = modelAsset.b005Offset
#B003 and B004
meshCount = modelAsset.b003Size//0x40
checkPoint = modelAsset.b004Offset
b004Size = modelAsset.b004Offset + modelAsset.b004Size
for j in range(meshCount):
bs52.seek(modelAsset.b003Offset+j*0x40)
meshAsset = LM3MeshAsset()
#parse section
bs52.readUInt() #hashName
meshAsset.indexBufferOffset = bs52.readUInt()
indexFlags = bs52.readUInt()
meshAsset.indexCount = (indexFlags & 0xFFFFFF)
type = (indexFlags >> 24)
if (type == 0x80):
meshAsset.indexFormat = 1
else:
meshAsset.indexFormat = 0
meshAsset.vertexCount = bs52.readUInt()
bs52.readUInt()
bs52.readUShort()
bs52.readUShort()
bs52.readUInt64()
bs52.readUInt()
meshAsset.isSkinned = True if bs52.readUInt() != 0xFFFFFFFF else False
bs52.seek(checkPoint)
if(meshAsset.isSkinned and bs52.tell()+4<=b004Size):
meshAsset.skinningBufferOffset = bs52.readUInt()
if(bs52.tell()+4<=b004Size):
meshAsset.vertexBufferOffset = bs52.readUInt()
if(bs52.tell()+4<=b004Size):
bs52.readUInt()
if(bs52.tell()+4<=b004Size):
bs52.readUInt()
checkPoint = bs52.tell()
modelAsset.meshAssetList.append(meshAsset)
#B006
#load global textureMap if relevant
if globalPath is not None:
globalTextureHashesToFileIndex = pickle.load(open( globalPath + os.sep + "Textures" + os.sep + "textureMap.lm3tMap", "rb" ))
#process section
bs52.seek(modelAsset.b006Offset)
while(bs52.tell() < modelAsset.b006Offset+modelAsset.b006Size):
temp = hex(bs52.readUInt())
if temp in textureHashesToFileIndex:
modelAsset.pairedTextureFileIndices.append(textureHashesToFileIndex[temp])
elif globalPath is not None:
if temp in globalTextureHashesToFileIndex:
modelAsset.pairedGlobalTextureFileIndices.append(globalTextureHashesToFileIndex[temp])
modelAsset.pairedTextureFileIndices = list(sorted(set(modelAsset.pairedTextureFileIndices)))
modelAsset.pairedGlobalTextureFileIndices = list(sorted(set(modelAsset.pairedGlobalTextureFileIndices)))
#B007
bs52.seek(modelAsset.b007Offset)
while(len(modelAsset.materialOffsets)<len(modelAsset.meshAssetList)):
checkPoint = bs52.tell()
temp = bs52.readBytes(0x1C)
if temp == materialFlag:
bs52.seek(checkPoint-4)
modelAsset.materialOffsets.append(bs52.readUInt())
bs52.seek(checkPoint+4)
modelAsset.materialOffsets.append(modelAsset.b006Size+4)
#Skeleton scan + match
for i,skeletonAsset in enumerate(skeletonAssetList):
if skeletonAsset.pairedModelHashName in modelHashesToFileIndex:
modelAssetList[modelHashesToFileIndex[skeletonAsset.pairedModelHashName]].s7103Size = skeletonAsset.s7103Size
modelAssetList[modelHashesToFileIndex[skeletonAsset.pairedModelHashName]].s7103Offset = skeletonAsset.s7103Offset
modelAssetList[modelHashesToFileIndex[skeletonAsset.pairedModelHashName]].s7105Size = skeletonAsset.s7105Size
modelAssetList[modelHashesToFileIndex[skeletonAsset.pairedModelHashName]].s7105Offset = skeletonAsset.s7105Offset
modelAssetList[modelHashesToFileIndex[skeletonAsset.pairedModelHashName]].s7106Size = skeletonAsset.s7106Size
modelAssetList[modelHashesToFileIndex[skeletonAsset.pairedModelHashName]].s7106Offset = skeletonAsset.s7106Offset
if i < len(animationBundleAssetList):
modelAssetList[modelHashesToFileIndex[skeletonAsset.pairedModelHashName]].animationIndices = animationBundleAssetList[i].animationIndices
for i,modelAsset in enumerate(modelAssetList):
if modelAsset.s7105Offset > 0:
#B103
bs52.seek(modelAsset.b103Offset)
for j in range(modelAsset.b103Size//0x4):
hash = hex(bs52.readUInt())
modelAsset.boneIDB1ToHash[j] = hash
modelAsset.hashToBoneIDB1[hash] = j
#7105
bs53.seek(modelAsset.s7105Offset)
for j in range(modelAsset.s7105Size//0x8):
hash = hex(bs53.readUInt())
id = bs53.readUInt()
modelAsset.boneID71ToHash[id] = hash
modelAsset.hashToBoneID71[hash] = id
for i,modelAsset in enumerate(modelAssetList):
pickle.dump( modelAsset, open( modelPath + os.sep + "model_" + str(i) +".lm3m", "wb" ))
pickle.dump( modelHashesToFileIndex, open( modelPath + os.sep + "modelMap.lm3mMap", "wb" ))
for i,animationAsset in enumerate(animationAssetList):
pickle.dump( animationAsset, open( animPath + os.sep + "anim_" + str(i) +".lm3a", "wb" ))
for i,animationBundleAsset in enumerate(animationBundleAssetList):
pickle.dump( animationBundleAsset, open( animBundlePath + os.sep + "animPack_" + str(i) +".lm3ap", "wb" ))
print("Files and assets extracted")
def ExtractDict(data, mdlList):
ctx = rapi.rpgCreateContext()
bs = NoeBitStream(data)
InitializeFromDict()
#Dict parsing, simplified since we don't need everything
bs.readBytes(0xc)
fileCount = bs.readUByte()
chunkCount = bs.readUByte()
bs.seek(0x2 + chunkCount*0x18,1)
offsets = {}
sizes = {}
#verify if global dir is a valid one before proceeding
if globalPath is not None:
if os.path.isdir(globalPath) is not True:
print("invalid directory")
return 0
for i in range(fileCount):
offset = bs.readUInt()
size = bs.readUInt()
zSize = bs.readUInt()
bs.readBytes(0x2)
source = bs.readUByte()
bs.readUByte()
#Only keep the .data related files, we don't have .debug and .nxpc
if source == 0:
offsets[i] = offset
sizes[i] = size
indices = list(offsets.keys())
#Decompress the zlib compressed files and put them in the File_Data folder
for i, key in enumerate(indices):
if sizes[key] > 0:
if i +1 < len(indices):
fileData = rapi.loadIntoByteArray(dataFileName)[offsets[indices[i]]:offsets[indices[i+1]]]
else:
continue
decompressedData = rapi.decompInflate(fileData,sizes[key])
destinationFileName = filePath + os.sep + rapi.getLocalFileName(dataFileName)[:-5] + '_' + str(key) + '.lm3'
with open(destinationFileName,"wb") as destFile:
destFile.write(decompressedData)
#Extraction completed, next we'll parse the chunk table and prepare the assets
ExtractAssets()
try:
mdl = rapi.rpgConstructModel()
except:
mdl = NoeModel()
mdlList.append(mdl)
return 1
# =================================================================
# Data processing
# =================================================================
def processTexture(textureAssets):
for i,textureAsset in enumerate(textureAssets):
bs63.seek(textureAsset.headerOffset+4) #skip the hash, not needed
width = bs63.readUShort()
height = bs63.readUShort()
bs63.readBytes(4)
format = bs63.readUByte()
bs63.readBytes(3)
maxBlockHeight = 16
# print(i)
# print(hex(format))
if format == 0x0 or format == 0x1:
format = "r8g8b8a8"
elif format == 0x11:
format = noesis.NOESISTEX_DXT1
elif format == 0x15:
format = noesis.FOURCC_ATI2
elif format == 0x16:
format = noesis.FOURCC_BC5
elif format == 0x19:
format = "ASTC_4_4"
elif format == 0x1A:
format = "ASTC_5_4"
elif format == 0x1B:
format = "ASTC_5_5"
elif format == 0x1C:
format = "ASTC_6_5"
elif format == 0x1D:
format = "ASTC_6_6"
elif format == 0x1E:
format = "ASTC_8_5"
elif format == 0x1F:
format = "ASTC_8_6"
elif format == 0x20:
format = "ASTC_8_8"
else:
print("UNKNOWN TEXTURE FORMAT !" + str(hex(format)))
format = noesis.NOESISTEX_UNKNOWN
textureName = str(i) + '.dds'
bs65.seek(textureAsset.dataOffset)
textureData = bs65.readBytes(textureAsset.dataSize)
bRaw = type(format) == str
if bRaw and format.startswith("ASTC"):
blockWidth, blockHeight = list(map(lambda x: int(x), format.split('_')[1:]))
widthInBlocks = (width + (blockWidth - 1)) // blockWidth
heightInBlocks = (height + (blockHeight - 1)) // blockHeight
blockSize = 16
maxBlockHeight = 8 if width <= 256 or height <= 256 else 16
maxBlockHeight = 4 if width <= 128 or height <= 128 else maxBlockHeight
maxBlockHeight = 2 if width <= 64 or height <= 64 else maxBlockHeight
if format == "ASTC_6_6":
maxBlockHeight = 4 if width <= 64 or height <= 64 else maxBlockHeight
maxBlockHeight = 2 if width <= 32 or height <= 32 else maxBlockHeight
maxBlockHeight = 1 if width <= 32 and height <= 32 else maxBlockHeight
#check kboo text9
textureData = rapi.callExtensionMethod("untile_blocklineargob", textureData, widthInBlocks, heightInBlocks, blockSize, maxBlockHeight)
textureData = rapi.callExtensionMethod("astc_decoderaw32", textureData, blockWidth, blockHeight, 1, width, height, 1)
format = noesis.NOESISTEX_RGBA32
elif bRaw:
blockWidth = blockHeight = 1
widthInBlocks = (width + (blockWidth - 1)) // blockWidth
heightInBlocks = (height + (blockHeight - 1)) // blockHeight
maxBlockHeight = 8 if width <= 256 or height <= 256 else 16
maxBlockHeight = 4 if width <= 128 or height <= 128 else maxBlockHeight
maxBlockHeight = 2 if width <= 32 or height <= 32 else maxBlockHeight
textureData = rapi.callExtensionMethod("untile_blocklineargob", textureData, widthInBlocks, heightInBlocks, 16, maxBlockHeight)
textureData = rapi.imageDecodeRaw(textureData, width, height, format,2)
format = noesis.NOESISTEX_RGBA32
else:
blockWidth = blockHeight = 4
blockSize = 8 if format == noesis.NOESISTEX_DXT1 else 16
widthInBlocks = (width + (blockWidth - 1)) // blockWidth
heightInBlocks = (height + (blockHeight - 1)) // blockHeight
maxBlockHeight = 8 if width <= 256 or height <= 256 else 16
maxBlockHeight = 4 if width <= 128 or height <= 128 else maxBlockHeight
maxBlockHeight = 2 if width <= 32 or height <= 32 else maxBlockHeight
textureData = rapi.callExtensionMethod("untile_blocklineargob", textureData, widthInBlocks, heightInBlocks, blockSize, maxBlockHeight)
textureData = rapi.imageDecodeDXT(textureData, width, height, format,0.0,2)
format = noesis.NOESISTEX_RGBA32
tex = NoeTexture(textureAsset.hashName, width, height, textureData, format)
textureHashToIndex[textureAsset.hashName]=len(textureList)
textureList.append(tex)
def processModel(modelAssets):
for a,modelAsset in enumerate(modelAssets):
if bLoadMaterials:
#Local textures
InitializeFileStream(63)
InitializeFileStream(65)
tempTextureList = []
for tex in modelAsset.pairedTextureFileIndices:
tempTextureList.append(pickle.load(open( textPath+os.sep+"texture_" + str(tex)+".lm3t", "rb" )))
processTexture(tempTextureList)
global bs63,bs65
del bs63,bs65
#Global textures, if relevant
if globalPath is not None:
tempTextureList = []
global bs63,bs65
bs63 = NoeBitStream(rapi.loadIntoByteArray(getFileNum(63,True)))
bs65 = NoeBitStream(rapi.loadIntoByteArray(getFileNum(65,True)))
bs63.setEndian(NOE_LITTLEENDIAN)
bs65.setEndian(NOE_LITTLEENDIAN)
for tex in modelAsset.pairedGlobalTextureFileIndices:
tempTextureList.append(pickle.load(open( globalPath + os.sep + "Textures"+os.sep+"texture_" + str(tex)+".lm3t", "rb" )))
processTexture(tempTextureList)
del bs63,bs65
#Materials (kinda...)
InitializeFileStream(52)
materialList = []
for index in range(len(modelAsset.materialOffsets)):
modelAsset.materialOffsets[index]+=modelAsset.b006Offset-8
for i,offset in enumerate(modelAsset.materialOffsets[:-1]):
material = NoeMaterial('submesh_' + str(i), "")
# "Good" material
if offset + 448 in modelAsset.materialOffsets:
bs52.seek(offset)
bs52.readBytes(0x14)
hash = hex(bs52.readUInt())
#added for Hellen in Shiny.data, fixed some building stuff
if hash == hex(0x81800000):
bs52.seek(-8,1)
hash = hex(bs52.readUInt())
if hash in textureHashToIndex:
material.setTexture(textureList[textureHashToIndex[hash]].name)
bs52.readBytes(4)
hash = hex(bs52.readUInt())
if hash in textureHashToIndex:
material.setNormalTexture(textureList[textureHashToIndex[hash]].name)
material.setNormalTexture(textureList[2].name)
materialList.append(material)
else:
if hash in textureHashToIndex:
material.setNormalTexture(textureList[textureHashToIndex[hash]].name)
bs52.readBytes(8)
bs52.readBytes(0xC)
hash = hex(bs52.readUInt())
if hash in textureHashToIndex:
material.setTexture(textureList[textureHashToIndex[hash]].name)
# bs52.readBytes(0x38)
# hash = bs52.readUInt()
# print(bs52.tell())
# if hash in model.textureHashesToIndex:
# print("spec found !")
# material.setSpecularTexture(textureList[model.textureHashesToIndex[hash]].name)
materialList.append(material)
#just for egadd glasses...
elif offset == 330752 and offset+196 == modelAsset.materialOffsets[-1]:
bs52.seek(offset)
bs52.readBytes(0x10)
hash = hex(bs52.readUInt())
if hash in textureHashToIndex:
material.setTexture(textureList[textureHashToIndex[hash]].name)
bs52.readBytes(8)
bs52.readBytes(0xC)
hash = hex(bs52.readUInt())
if hash in textureHashToIndex:
material.setNormalTexture(textureList[textureHashToIndex[hash]].name)
# bs52.readBytes(0x38)
# hash = bs52.readUInt()
# print(bs52.tell())
# if hash in model.textureHashesToIndex:
# print("spec found !")
# material.setSpecularTexture(textureList[model.textureHashesToIndex[hash]].name)
materialList.append(material)
elif offset + 192 in modelAsset.materialOffsets:
bs52.seek(offset)
bs52.readBytes(0x10)
hash = hex(bs52.readUInt())
a = bs52.readUInt()
if a != 0x81800000:
if hash in textureHashToIndex:
material.setNormalTexture(textureList[textureHashToIndex[hash]].name)
bs52.readBytes(0x4)
bs52.readBytes(0xC)
hash = hex(bs52.readUInt())
if hash in textureHashToIndex:
material.setTexture(textureList[textureHashToIndex[hash]].name)
materialList.append(material)
else:
if hash in textureHashToIndex:
material.setTexture(textureList[textureHashToIndex[hash]].name)
bs52.readBytes(4)
hash = hex(bs52.readUInt())
if hash in textureHashToIndex:
material.setNormalTexture(textureList[textureHashToIndex[hash]].name)
material.setNormalTexture(textureList[2].name)
materialList.append(material)
else:
bs52.seek(offset)
a = 0
while a != 0x81800000 and a!= 0x01800000 and bs52.tell() < modelAsset.materialOffsets[-1]:
a= bs52.readUInt()
if a == 0x81800000:
bs52.seek(-8,1)
hash = hex(bs52.readUInt())
# print(hash)
if hash in textureHashToIndex:
material.setTexture(textureList[textureHashToIndex[hash]].name)
bs52.readBytes(8)
hash = hex(bs52.readUInt())
if hash in textureHashToIndex:
material.setNormalTexture(textureList[textureHashToIndex[hash]].name)
# bs52.readBytes(8)
# hash = bs52.readUInt()
# if hash in model.textureHashesToIndex:
# print("spec found !")
# material.setSpecularTexture(textureList[model.textureHashesToIndex[hash]].name)
materialList.append(material)
elif a == 0x01800000: #added because of Hellen concert Hall sm 0
bs52.seek(-8,1)
bs52.readBytes(0x18)
hash = hex(bs52.readUInt())
# print(hash)
if hash in textureHashToIndex:
material.setTexture(textureList[textureHashToIndex[hash]].name)
bs52.readBytes(8)
hash = hex(bs52.readUInt())
if hash in textureHashToIndex:
material.setNormalTexture(textureList[textureHashToIndex[hash]].name)
# bs52.readBytes(8)
# hash = bs52.readUInt()
# if hash in model.textureHashesToIndex:
# print("spec found !")
# material.setSpecularTexture(textureList[model.textureHashesToIndex[hash]].name)
materialList.append(material)
else:
InitializeFileStream(52)
materialList = []
#Skeleton
#Parenting info
InitializeFileStream(53)
global bs53
if modelAsset.s7106Offset > 0:
bs53.seek(modelAsset.s7106Offset)
for i in range(modelAsset.s7106Size//0x2):
modelAsset.parentList.append(bs53.readUShort())
#Transform info
if modelAsset.s7103Offset > 0:
bs53.seek(modelAsset.s7103Offset)
for i in range(modelAsset.s7103Size//0x1C):
quaternion = [bs53.readFloat() for j in range(4)]
position = [bs53.readFloat() for j in range(3)]
boneMatrixTransform = NoeQuat(quaternion).toMat43().inverse()
boneMatrixTransform[3] = NoeVec3(position)
# bone = NoeBone(i, 'bone_' + str(i), boneMatrixTransform, None, modelAsset.parentList[i])
bone = NoeBone(i, 'bone_' + str(i), boneMatrixTransform, None, modelAsset.parentList[i])
modelAsset.jointList.append(bone)
for bone in modelAsset.jointList:
parentId = bone.parentIndex
if parentId != 65535:
bone.setMatrix(bone.getMatrix() * modelAsset.jointList[parentId].getMatrix())
else:
bone.setMatrix(bone.getMatrix()*NoeAngles([90,0,-90]).toMat43())
#Animation
if bLoadAnimations:
animBPath = rapi.loadPairedFileGetPath("animPack file", ".lm3ap")
if animBPath is not None:
animationBundleAsset = pickle.load(open( animBPath[1], "rb" ))
# animList = modelAsset.animationIndices if bLoadAnimations else []
animList = animationBundleAsset.animationIndices
for n,animationindex in enumerate(animList):
animationAsset = pickle.load(open( animPath+os.sep+"anim_" + str(animationindex)+".lm3a", "rb" ))
keyframedJointList = []
# print(animationAsset.dataOffset)
# print(animationAsset.dataSize)
bs53.seek(animationAsset.dataOffset)
bs53.readUInt() # 0, always ?
boneHeaderCount = bs53.readUShort()
frameCount = bs53.readUShort()
duration = bs53.readFloat()
bs53.readUInt() # 0, always ?
bs53.seek(0x18,1) # Not sure what the two first chunks mean
rotNoeKeyFramedValues = {}
posNoeKeyFramedValues = {}
scaleNoeKeyFramedValues = {}
unknownRotOpcode = {}
unknownPosOpcode = {}
unknownScaleOpcode = {}
for i in range(boneHeaderCount-2): #Sometimes additionnal chunks starting with 0x00000505, don't really care since we have offsets
boneHeader = LM3BoneHeader()
boneHeader.hash = hex(bs53.readUInt())
boneHeader.index = bs53.readUByte()
boneHeader.magic = bs53.readUByte()
boneHeader.type = bs53.readUByte()
boneHeader.opcode = bs53.readUByte()
boneHeader.offset = bs53.readUInt()
checkpoint = bs53.tell()
if boneHeader.hash in modelAsset.hashToBoneID71:
# if boneHeader.magic != 0xC0 and boneHeader.magic != 0xC1 and boneHeader.magic != 0xC2:
# print("New magic found")
# print(boneHeader.magic)
#rotation
if boneHeader.type == 1:
rotNoeKeyFramedValues[boneHeader.hash] = []
if boneHeader.opcode == 0x0F: #confirmed
bs53.seek(animationAsset.dataOffset + boneHeader.offset)
for j in range(frameCount):
quaternion = NoeQuat([bs53.readFloat() for a in range(4)]).transpose()
rotationKeyframedValue = NoeKeyFramedValue(duration/frameCount*j,quaternion)
rotNoeKeyFramedValues[boneHeader.hash].append(rotationKeyframedValue)
elif boneHeader.opcode == 0x13: #not sure but makes sense
bs53.seek(animationAsset.dataOffset + boneHeader.offset)
count = bs53.readUInt()
for j in range(count):
quaternion = NoeQuat([bs53.readFloat() for a in range(4)]).transpose()
rotationKeyframedValue = NoeKeyFramedValue(duration/count*j,quaternion)
rotNoeKeyFramedValues[boneHeader.hash].append(rotationKeyframedValue)
elif boneHeader.opcode == 0x15: #confirmed
bs53.seek(animationAsset.dataOffset + boneHeader.offset)
quaternion = NoeQuat([bs53.readFloat() for a in range(4)]).transpose()
rotationKeyframedValue = NoeKeyFramedValue(0,quaternion)
rotNoeKeyFramedValues[boneHeader.hash].append(rotationKeyframedValue)
# elif boneHeader.opcode == 0x16: #not sure at all
# flag to say nothing do be done ?
# bs53.seek(animationAsset.dataOffset + boneHeader.offset)
# quaternion = NoeQuat([bs53.readShort()/0x7FFF for a in range(4)]).transpose()
# rotationKeyframedValue = NoeKeyFramedValue(0,quaternion)
# rotNoeKeyFramedValues[boneHeader.hash].append(rotationKeyframedValue)
elif boneHeader.opcode == 0x17: #not sure at all
# Some weird thing is going on. Normalization or something needed ?
bs53.seek(animationAsset.dataOffset + boneHeader.offset)
quaternion = NoeAngles([bs53.readShort()/180,0,0]).toQuat()
rotationKeyframedValue = NoeKeyFramedValue(0,quaternion)
rotNoeKeyFramedValues[boneHeader.hash].append(rotationKeyframedValue)
elif boneHeader.opcode == 0x18: #somewhat confirmed (priestess cloth)
bs53.seek(animationAsset.dataOffset + boneHeader.offset)
for j in range(frameCount):
quaternion = NoeAngles([0,0,bs53.readShort()/180]).toQuat()
rotationKeyframedValue = NoeKeyFramedValue(duration/frameCount*j,quaternion)
rotNoeKeyFramedValues[boneHeader.hash].append(rotationKeyframedValue)
elif boneHeader.opcode == 0x19: #somewhat confirmed (priestess arms)
bs53.seek(animationAsset.dataOffset + boneHeader.offset)
for j in range(frameCount):
quaternion = NoeAngles([0,bs53.readShort()/180,0]).toQuat()
rotationKeyframedValue = NoeKeyFramedValue(duration/frameCount*j,quaternion)
rotNoeKeyFramedValues[boneHeader.hash].append(rotationKeyframedValue)
# elif boneHeader.opcode == 0x1A: #need confirm, king boo
# bs53.seek(animationAsset.dataOffset + boneHeader.offset)
# bs53.readShort()
# quaternion = NoeAngles([bs53.readShort()/180,0,0]).toQuat()
# rotationKeyframedValue = NoeKeyFramedValue(0,quaternion)
# rotNoeKeyFramedValues[boneHeader.hash].append(rotationKeyframedValue)
# elif boneHeader.opcode == 0x1B: #seen on maid
# bs53.seek(animationAsset.dataOffset + boneHeader.offset)
# bs53.readShort()
# quaternion = NoeAngles([bs53.readShort()/180,0,0]).toQuat()
# rotationKeyframedValue = NoeKeyFramedValue(0,quaternion)
# rotNoeKeyFramedValues[boneHeader.hash].append(rotationKeyframedValue)
# elif boneHeader.opcode == 0x1C: #need confirm (pianist tail)
# bs53.seek(animationAsset.dataOffset + boneHeader.offset)
# bs53.readShort()
# quaternion = NoeAngles([0,bs53.readShort()/180,0]).toQuat()
# rotationKeyframedValue = NoeKeyFramedValue(0,quaternion)
# rotNoeKeyFramedValues[boneHeader.hash].append(rotationKeyframedValue)
else:
if hex(boneHeader.opcode) in unknownRotOpcode:
unknownRotOpcode[hex(boneHeader.opcode)] += 1
else:
unknownRotOpcode[hex(boneHeader.opcode)] = 1
#location
elif boneHeader.type == 3:
posNoeKeyFramedValues[boneHeader.hash] = []
if boneHeader.opcode == 0x6: #confirmed, somewhat (priestess movement)
bs53.seek(animationAsset.dataOffset + boneHeader.offset)
for j in range(frameCount):
position = NoeVec3([bs53.readFloat() for a in range(3)])
positionKeyFramedValue = NoeKeyFramedValue(duration/frameCount*j, position)
posNoeKeyFramedValues[boneHeader.hash].append(positionKeyFramedValue)
elif boneHeader.opcode == 0x8: #guess
bs53.seek(animationAsset.dataOffset + boneHeader.offset)
for j in range(frameCount):
position = NoeVec3([bs53.readHalfFloat() for a in range(3)])
positionKeyFramedValue = NoeKeyFramedValue(duration/frameCount*j, position)
posNoeKeyFramedValues[boneHeader.hash].append(positionKeyFramedValue)
elif boneHeader.opcode == 0x9: #to be confirmed but makes sense
bs53.seek(animationAsset.dataOffset + boneHeader.offset)
count = bs53.readUInt()
for j in range(count):
position = NoeVec3([bs53.readFloat() for a in range(3)])
positionKeyFramedValue = NoeKeyFramedValue(duration/count*j, position)
posNoeKeyFramedValues[boneHeader.hash].append(positionKeyFramedValue)
elif boneHeader.opcode == 0xA: #to be confirmed but makes sense
bs53.seek(animationAsset.dataOffset + boneHeader.offset)
count = bs53.readUShort()
for j in range(count):
position = NoeVec3([bs53.readHalfFloat() for a in range(3)])
positionKeyFramedValue = NoeKeyFramedValue(duration/count*j, position)
posNoeKeyFramedValues[boneHeader.hash].append(positionKeyFramedValue)
elif boneHeader.opcode == 0xB: #to be confirmed but makes sense
bs53.seek(animationAsset.dataOffset + boneHeader.offset)
position = NoeVec3([bs53.readFloat() for a in range(3)])
positionKeyFramedValue = NoeKeyFramedValue(0, position)
posNoeKeyFramedValues[boneHeader.hash].append(positionKeyFramedValue)
elif boneHeader.opcode == 0xC: #confirmed (maid stuff)
bs53.seek(animationAsset.dataOffset + boneHeader.offset)
position = NoeVec3([bs53.readHalfFloat() for a in range(3)])
positionKeyFramedValue = NoeKeyFramedValue(0, position)
posNoeKeyFramedValues[boneHeader.hash].append(positionKeyFramedValue)
elif boneHeader.opcode == 0xD: #confirmed (pianist movement)
bs53.seek(animationAsset.dataOffset + boneHeader.offset)
axis = bs53.readUInt()
bs53.readBytes(0x8)
count = bs53.readUInt()
for j in range(count):
data = bs53.readFloat()
if axis == 0:
position = NoeVec3([data,0,0])
elif axis == 1:
position = NoeVec3([0,data,0])
elif axis == 2:
position = NoeVec3([0,0,data])
positionKeyFramedValue = NoeKeyFramedValue(duration/count*j, position)
posNoeKeyFramedValues[boneHeader.hash].append(positionKeyFramedValue)
elif boneHeader.opcode == 0xE: #confirmed somewhat (Hellen outfit)
bs53.seek(animationAsset.dataOffset+ boneHeader.offset)
a = bs53.readUInt()
bs53.readUShort()
count = bs53.readUShort()
for j in range(count):
position = NoeVec3([bs53.readHalfFloat(),0,0])
positionKeyFramedValue = NoeKeyFramedValue(duration/count*j, position)
posNoeKeyFramedValues[boneHeader.hash].append(positionKeyFramedValue)
else:
if hex(boneHeader.opcode) in unknownPosOpcode:
unknownPosOpcode[hex(boneHeader.opcode)] += 1
else:
unknownPosOpcode[hex(boneHeader.opcode)] = 1
#Not confirmed at all, no clue yet. Probably something other than scale
# elif boneHeader.type == 2:
# scaleNoeKeyFramedValues[boneHeader.hash] = []
# posNoeKeyFramedValues[boneHeader.hash] = []
# if boneHeader.opcode == 0x2A: #always of length 12 so either 6 HF or 3 F, don't have any clear example
# bs53.seek(animationAsset.dataOffset + boneHeader.offset)
# scale = NoeVec3([bs53.readHalfFloat() for a in range(3)])
# scaleKeyFramedValue = NoeKeyFramedValue(0, scale)
# scaleNoeKeyFramedValues[boneHeader.hash].append(scaleKeyFramedValue)
# bs53.seek(animationAsset.dataOffset + boneHeader.offset)
# position = NoeVec3([bs53.readFloat() for a in range(3)])
# positionKeyFramedValue = NoeKeyFramedValue(0, position)
# posNoeKeyFramedValues[boneHeader.hash].append(positionKeyFramedValue)
# else:
# if hex(boneHeader.opcode) in unknownScaleOpcode:
# unknownScaleOpcode[hex(boneHeader.opcode)] += 1
# else:
# unknownScaleOpcode[hex(boneHeader.opcode)] = 1
bs53.seek(checkpoint)
# print("anim " + str(animationindex))
# print(unknownRotOpcode)
# print(unknownPosOpcode)
for hash in modelAsset.hashToBoneID71:
if hash in posNoeKeyFramedValues or hash in rotNoeKeyFramedValues or hash in scaleNoeKeyFramedValues:
actionBone = NoeKeyFramedBone(modelAsset.hashToBoneID71[hash])
#root bone rotation ignored as it seem to screw up some stuff
if hash in rotNoeKeyFramedValues and hash != "0x2e51a3":
actionBone.setRotation(rotNoeKeyFramedValues[hash], noesis.NOEKF_ROTATION_QUATERNION_4)
if hash in posNoeKeyFramedValues:
actionBone.setTranslation(posNoeKeyFramedValues[hash], noesis.NOEKF_TRANSLATION_VECTOR_3)
if hash in scaleNoeKeyFramedValues:
actionBone.setScale(scaleNoeKeyFramedValues[hash], noesis.NOEKF_SCALE_VECTOR_3)
keyframedJointList.append(actionBone)
anim = NoeKeyFramedAnim("anim_"+str(animationindex), modelAsset.jointList, keyframedJointList, 30)
animationList.append(anim)
del bs53
#Geometry
InitializeFileStream(54)
bs52.seek(modelAsset.b001Offset)
modelMatrix = NoeMat44.fromBytes(bs52.readBytes(0x40))
for i,mesh in enumerate(modelAsset.meshAssetList):
#vertices
bs54.seek(modelAsset.buffersOffset + mesh.vertexBufferOffset)
finalVertexBuffer = bs54.readBytes(0x30*mesh.vertexCount)
#since UVs are split between normal coords we need to construct a new buffer
uv1 = noesis.deinterleaveBytes(finalVertexBuffer,0xC,0x4,0x30)
uv2 = noesis.deinterleaveBytes(finalVertexBuffer,0x1C,0x4,0x30)
finalUVBuffer = noesis.interleaveUniformBytes(uv1+uv2,4,mesh.vertexCount)
#indices
bs54.seek(modelAsset.buffersOffset + mesh.indexBufferOffset)
multiplier = 1 if mesh.indexFormat else 2
finalIndicesBuffer = bs54.readBytes(mesh.indexCount * multiplier)
#skinning, don't care about non animated stuff for now so we ignore objects with only B1 info
if mesh.isSkinned and modelAsset.s7103Offset > 0:
bs54.seek(modelAsset.buffersOffset)
bs54.seek(mesh.skinningBufferOffset,1)
checkpoint = bs54.tell()
#Indices, need to be taken from B1 to 71
finalBlendIndicesBuffer = bytes()
for j in range(mesh.vertexCount):
bs54.seek(checkpoint+j*0x14)
temp = [bs54.readUByte() for j in range(4)]
a = []
for t in temp:
h = modelAsset.boneIDB1ToHash[t]
f = modelAsset.hashToBoneID71[h]
finalBlendIndicesBuffer+=struct.pack('<B', f)
bs54.seek(checkpoint)
finalBlendWeightsBuffer = noesis.deinterleaveBytes(bs54.readBytes(0x14*mesh.vertexCount),0x4,0x10,0x14)
meshName = 'submesh_' + str(i)
rapi.rpgSetMaterial(meshName)
rapi.rpgSetName(meshName)
rapi.rpgClearBufferBinds()
rapi.rpgBindPositionBufferOfs(finalVertexBuffer, noesis.RPGEODATA_FLOAT, 0x30,0x0)
rapi.rpgBindNormalBufferOfs(finalVertexBuffer, noesis.RPGEODATA_FLOAT, 0x30,0x10)
rapi.rpgBindUV1Buffer(finalUVBuffer, noesis.RPGEODATA_FLOAT, 0x8)
if mesh.isSkinned and modelAsset.s7103Offset > 0:
rapi.rpgBindBoneIndexBuffer(finalBlendIndicesBuffer, noesis.RPGEODATA_UBYTE, 0x4, 0x4)
rapi.rpgBindBoneWeightBuffer(finalBlendWeightsBuffer, noesis.RPGEODATA_FLOAT,0x10, 0x4)
correctionMatrix = skelRotCorrection.toMat43()
correctionMatrix[3] = skelPosCorrection
transfMatrix = modelMatrix.toMat43() * NoeAngles([90,0,0]).toMat43() * correctionMatrix
rapi.rpgSetTransform(transfMatrix)
rapi.rpgCommitTriangles(finalIndicesBuffer,noesis.RPGEODATA_UBYTE if mesh.indexFormat else noesis.RPGEODATA_USHORT, mesh.indexCount,noesis.RPGEO_TRIANGLE, 1)
try:
mdl = rapi.rpgConstructModel()
except:
mdl = NoeModel()
mdl.setModelMaterials(NoeModelMaterials(textureList, []))
if len(modelAsset.jointList) > 0:
mdl.setBones(modelAsset.jointList)
if len(animationList) > 0:
mdl.setAnims(animationList)
mdl.setModelMaterials(NoeModelMaterials(textureList, materialList))
modelList.append(mdl)
philipe rodrigues
Oct 18, 2021, 1:20:36 PM10/18/21
to Python Brasil
Olha eu consegui extrair os textos, mas não consegui reinserir, como eu reinsiro essa entrada 96 editada de volta no ini.dict? Existe alguma ferramenta? Eu quero editar essa entrada 69 que está no init.dict, eu consigo extrair ela, mas não consigo reimportar pro dict. 
Em quarta-feira, 13 de outubro de 2021 às 10:59:53 UTC-3, eia...@yahoo.com escreveu:
Reply all
Reply to author
Forward
0 new messages