snipeslow/Avatar-Toolkit
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 1 Wiki Activity

Added Comments

Browse Source
This commit is contained in:
Yusarina
2024-06-14 23:44:52 +01:00
parent 03c1fee622
commit 13f6e57f09
1 changed files with 103 additions and 101 deletions
Download Patch File Download Diff File Expand all files Collapse all files
core/pmx/import_pmx.py
+103 -101
View File
@@ -3,172 +3,174 @@ import struct
def read_pmx_header(file):
# Read PMX header information
magic = file.read(4)
magic = file.read(4) # Read magic bytes (should be "PMX ")
if magic != b'PMX ':
raise ValueError("Invalid PMX file")
version = struct.unpack('<f', file.read(4))[0]
version = struct.unpack('<f', file.read(4))[0] # Read version number (float, 4 bytes)
# Read additional header fields
data_size = struct.unpack('<i', file.read(4))[0]
encoding = struct.unpack('<b', file.read(1))[0]
additional_uvs = struct.unpack('<b', file.read(1))[0]
vertex_index_size = struct.unpack('<b', file.read(1))[0]
texture_index_size = struct.unpack('<b', file.read(1))[0]
material_index_size = struct.unpack('<b', file.read(1))[0]
bone_index_size = struct.unpack('<b', file.read(1))[0]
morph_index_size = struct.unpack('<b', file.read(1))[0]
rigid_body_index_size = struct.unpack('<b', file.read(1))[0]
data_size = struct.unpack('<i', file.read(4))[0] # Read size of remaining header data (int, 4 bytes)
encoding = struct.unpack('<b', file.read(1))[0] # Read encoding type (byte, 1 byte)
additional_uvs = struct.unpack('<b', file.read(1))[0] # Read number of additional UV layers (byte, 1 byte)
vertex_index_size = struct.unpack('<b', file.read(1))[0] # Read vertex index size (byte, 1 byte)
texture_index_size = struct.unpack('<b', file.read(1))[0] # Read texture index size (byte, 1 byte)
material_index_size = struct.unpack('<b', file.read(1))[0] # Read material index size (byte, 1 byte)
bone_index_size = struct.unpack('<b', file.read(1))[0] # Read bone index size (byte, 1 byte)
morph_index_size = struct.unpack('<b', file.read(1))[0] # Read morph index size (byte, 1 byte)
rigid_body_index_size = struct.unpack('<b', file.read(1))[0] # Read rigid body index size (byte, 1 byte)
# Read model name and comments
model_name = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace')
model_english_name = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace')
model_comment = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace')
model_english_comment = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace')
model_name = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace') # Read model name (string, variable length)
model_english_name = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace') # Read model English name (string, variable length)
model_comment = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace') # Read model comment (string, variable length)
model_english_comment = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace') # Read model English comment (string, variable length)
return version, encoding, additional_uvs, vertex_index_size, texture_index_size, material_index_size, bone_index_size, morph_index_size, rigid_body_index_size, model_name, model_english_name, model_comment, model_english_comment
def read_vertex(file, vertex_index_size):
position = struct.unpack('<3f', file.read(12))
normal = struct.unpack('<3f', file.read(12))
uv = struct.unpack('<2f', file.read(8))
position = struct.unpack('<3f', file.read(12)) # Read vertex position (float, 3 * 4 bytes)
normal = struct.unpack('<3f', file.read(12)) # Read vertex normal (float, 3 * 4 bytes)
uv = struct.unpack('<2f', file.read(8)) # Read vertex UV coordinates (float, 2 * 4 bytes)
if vertex_index_size == 1:
bone_indices = list(struct.unpack('<4B', file.read(4)))
bone_indices = list(struct.unpack('<4B', file.read(4))) # Read bone indices (byte, 4 * 1 byte)
elif vertex_index_size == 2:
bone_indices = list(struct.unpack('<4H', file.read(8)))
bone_indices = list(struct.unpack('<4H', file.read(8))) # Read bone indices (short, 4 * 2 bytes)
else:
bone_indices = list(struct.unpack('<4I', file.read(16)))
bone_indices = list(struct.unpack('<4I', file.read(16))) # Read bone indices (int, 4 * 4 bytes)
bone_weights = list(struct.unpack('<4f', file.read(16)))
edge_scale = struct.unpack('<f', file.read(4))[0]
bone_weights = list(struct.unpack('<4f', file.read(16))) # Read bone weights (float, 4 * 4 bytes)
edge_scale = struct.unpack('<f', file.read(4))[0] # Read edge scale (float, 4 bytes)
return position, normal, uv, bone_indices, bone_weights, edge_scale
def read_material(file, texture_index_size):
material_name = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace')
material_english_name = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace')
material_name = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace') # Read material name (string, variable length)
material_english_name = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace') # Read material English name (string, variable length)
diffuse_color = struct.unpack('<4f', file.read(16))
specular_color = struct.unpack('<3f', file.read(12))
specular_strength = struct.unpack('<f', file.read(4))[0]
ambient_color = struct.unpack('<3f', file.read(12))
diffuse_color = struct.unpack('<4f', file.read(16)) # Read diffuse color (float, 4 * 4 bytes)
specular_color = struct.unpack('<3f', file.read(12)) # Read specular color (float, 3 * 4 bytes)
specular_strength = struct.unpack('<f', file.read(4))[0] # Read specular strength (float, 4 bytes)
ambient_color = struct.unpack('<3f', file.read(12)) # Read ambient color (float, 3 * 4 bytes)
flag = struct.unpack('<b', file.read(1))[0]
edge_color = struct.unpack('<4f', file.read(16))
edge_size = struct.unpack('<f', file.read(4))[0]
texture_index = struct.unpack(f'<{texture_index_size}B', file.read(texture_index_size))[0]
sphere_texture_index = struct.unpack(f'<{texture_index_size}B', file.read(texture_index_size))[0]
sphere_mode = struct.unpack('<b', file.read(1))[0]
toon_sharing_flag = struct.unpack('<b', file.read(1))[0]
flag = struct.unpack('<b', file.read(1))[0] # Read flag (byte, 1 byte)
edge_color = struct.unpack('<4f', file.read(16)) # Read edge color (float, 4 * 4 bytes)
edge_size = struct.unpack('<f', file.read(4))[0] # Read edge size (float, 4 bytes)
texture_index = struct.unpack(f'<{texture_index_size}B', file.read(texture_index_size))[0] # Read texture index (byte, texture_index_size bytes)
sphere_texture_index = struct.unpack(f'<{texture_index_size}B', file.read(texture_index_size))[0] # Read sphere texture index (byte, texture_index_size bytes)
sphere_mode = struct.unpack('<b', file.read(1))[0] # Read sphere mode (byte, 1 byte)
toon_sharing_flag = struct.unpack('<b', file.read(1))[0] # Read toon sharing flag (byte, 1 byte)
if toon_sharing_flag == 0:
toon_texture_index = struct.unpack(f'<{texture_index_size}B', file.read(texture_index_size))[0]
toon_texture_index = struct.unpack(f'<{texture_index_size}B', file.read(texture_index_size))[0] # Read toon texture index (byte, texture_index_size bytes)
else:
toon_texture_index = struct.unpack('<b', file.read(1))[0]
toon_texture_index = struct.unpack('<b', file.read(1))[0] # Read shared toon texture index (byte, 1 byte)
comment = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace')
comment = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace') # Read material comment (string, variable length)
return material_name, material_english_name, diffuse_color, specular_color, specular_strength, ambient_color, flag, edge_color, edge_size, texture_index, sphere_texture_index, sphere_mode, toon_sharing_flag, toon_texture_index, comment
def read_bone(file, bone_index_size):
bone_name = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace')
bone_english_name = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace')
bone_name = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace') # Read bone name (string, variable length)
bone_english_name = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace') # Read bone English name (string, variable length)
position = struct.unpack('<3f', file.read(12))
parent_bone_index = struct.unpack(f'<{bone_index_size}B', file.read(bone_index_size))[0]
layer = struct.unpack('<i', file.read(4))[0]
flag = struct.unpack('<H', file.read(2))[0]
position = struct.unpack('<3f', file.read(12)) # Read bone position (float, 3 * 4 bytes)
parent_bone_index = struct.unpack(f'<{bone_index_size}B', file.read(bone_index_size))[0] # Read parent bone index (byte, bone_index_size bytes)
layer = struct.unpack('<i', file.read(4))[0] # Read bone layer (int, 4 bytes)
flag = struct.unpack('<H', file.read(2))[0] # Read bone flag (short, 2 bytes)
if flag & 0x0001:
tail_position = struct.unpack('<3f', file.read(12))
tail_position = struct.unpack('<3f', file.read(12)) # Read bone tail position (float, 3 * 4 bytes)
else:
tail_index = struct.unpack(f'<{bone_index_size}B', file.read(bone_index_size))[0]
tail_index = struct.unpack(f'<{bone_index_size}B', file.read(bone_index_size))[0] # Read bone tail index (byte, bone_index_size bytes)
if flag & 0x0100 or flag & 0x0200:
inherit_bone_parent_index = struct.unpack(f'<{bone_index_size}B', file.read(bone_index_size))[0]
inherit_bone_parent_influence = struct.unpack('<f', file.read(4))[0]
inherit_bone_parent_index = struct.unpack(f'<{bone_index_size}B', file.read(bone_index_size))[0] # Read inherit bone parent index (byte, bone_index_size bytes)
inherit_bone_parent_influence = struct.unpack('<f', file.read(4))[0] # Read inherit bone parent influence (float, 4 bytes)
if flag & 0x0400:
fixed_axis = struct.unpack('<3f', file.read(12))
fixed_axis = struct.unpack('<3f', file.read(12)) # Read fixed axis (float, 3 * 4 bytes)
if flag & 0x0800:
local_x_vector = struct.unpack('<3f', file.read(12))
local_z_vector = struct.unpack('<3f', file.read(12))
local_x_vector = struct.unpack('<3f', file.read(12)) # Read local X-axis vector (float, 3 * 4 bytes)
local_z_vector = struct.unpack('<3f', file.read(12)) # Read local Z-axis vector (float, 3 * 4 bytes)
if flag & 0x2000:
external_key = struct.unpack('<i', file.read(4))[0]
external_key = struct.unpack('<i', file.read(4))[0] # Read external key (int, 4 bytes)
if flag & 0x0020:
ik_target_bone_index = struct.unpack(f'<{bone_index_size}B', file.read(bone_index_size))[0]
ik_loop_count = struct.unpack('<i', file.read(4))[0]
ik_limit_radian = struct.unpack('<f', file.read(4))[0]
ik_link_count = struct.unpack('<i', file.read(4))[0]
ik_target_bone_index = struct.unpack(f'<{bone_index_size}B', file.read(bone_index_size))[0] # Read IK target bone index (byte, bone_index_size bytes)
ik_loop_count = struct.unpack('<i', file.read(4))[0] # Read IK loop count (int, 4 bytes)
ik_limit_radian = struct.unpack('<f', file.read(4))[0] # Read IK limit angle (float, 4 bytes)
ik_link_count = struct.unpack('<i', file.read(4))[0] # Read IK link count (int, 4 bytes)
ik_links = []
for _ in range(ik_link_count):
ik_link_bone_index = struct.unpack(f'<{bone_index_size}B', file.read(bone_index_size))[0]
ik_link_limit_min = struct.unpack('<3f', file.read(12))
ik_link_limit_max = struct.unpack('<3f', file.read(12))
ik_link_bone_index = struct.unpack(f'<{bone_index_size}B', file.read(bone_index_size))[0] # Read IK link bone index (byte, bone_index_size bytes)
ik_link_limit_min = struct.unpack('<3f', file.read(12)) # Read IK link limit minimum (float, 3 * 4 bytes)
ik_link_limit_max = struct.unpack('<3f', file.read(12)) # Read IK link limit maximum (float, 3 * 4 bytes)
ik_links.append((ik_link_bone_index, ik_link_limit_min, ik_link_limit_max))
return bone_name, bone_english_name, position, parent_bone_index, layer, flag, tail_position, inherit_bone_parent_index, inherit_bone_parent_influence, fixed_axis, local_x_vector, local_z_vector, external_key, ik_target_bone_index, ik_loop_count, ik_limit_radian, ik_links
def read_morph(file, morph_index_size):
morph_name = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace')
morph_english_name = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace')
morph_name = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace') # Read morph name (string, variable length)
morph_english_name = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace') # Read morph English name (string, variable length)
panel = struct.unpack('<b', file.read(1))[0]
morph_type = struct.unpack('<b', file.read(1))[0]
offset_size = struct.unpack('<i', file.read(4))[0]
panel = struct.unpack('<b', file.read(1))[0] # Read panel type (byte, 1 byte)
morph_type = struct.unpack('<b', file.read(1))[0] # Read morph type (byte, 1 byte)
offset_size = struct.unpack('<i', file.read(4))[0] # Read offset size (int, 4 bytes)
morph_data = []
for _ in range(offset_size):
if morph_type == 0: # Group
morph_index = struct.unpack(f'<{morph_index_size}B', file.read(morph_index_size))[0]
morph_value = struct.unpack('<f', file.read(4))[0]
morph_index = struct.unpack(f'<{morph_index_size}B', file.read(morph_index_size))[0] # Read morph index (byte, morph_index_size bytes)
morph_value = struct.unpack('<f', file.read(4))[0] # Read morph value (float, 4 bytes)
morph_data.append((morph_index, morph_value))
elif morph_type == 1: # Vertex
vertex_index = struct.unpack('<i', file.read(4))[0]
position_offset = struct.unpack('<3f', file.read(12))
vertex_index = struct.unpack('<i', file.read(4))[0] # Read vertex index (int, 4 bytes)
position_offset = struct.unpack('<3f', file.read(12)) # Read position offset (float, 3 * 4 bytes)
morph_data.append((vertex_index, position_offset))
elif morph_type == 2: # Bone
bone_index = struct.unpack(f'<{morph_index_size}B', file.read(morph_index_size))[0]
position_offset = struct.unpack('<3f', file.read(12))
rotation_offset = struct.unpack('<4f', file.read(16))
bone_index = struct.unpack(f'<{morph_index_size}B', file.read(morph_index_size))[0] # Read bone index (byte, morph_index_size bytes)
position_offset = struct.unpack('<3f', file.read(12)) # Read position offset (float, 3 * 4 bytes)
rotation_offset = struct.unpack('<4f', file.read(16)) # Read rotation offset (float, 4 * 4 bytes)
morph_data.append((bone_index, position_offset, rotation_offset))
elif morph_type == 3: # UV
vertex_index = struct.unpack('<i', file.read(4))[0]
uv_offset = struct.unpack('<4f', file.read(16))
vertex_index = struct.unpack('<i', file.read(4))[0] # Read vertex index (int, 4 bytes)
uv_offset = struct.unpack('<4f', file.read(16)) # Read UV offset (float, 4 * 4 bytes)
morph_data.append((vertex_index, uv_offset))
elif morph_type == 4: # UV extended1
vertex_index = struct.unpack('<i', file.read(4))[0]
uv_offset = struct.unpack('<4f', file.read(16))
vertex_index = struct.unpack('<i', file.read(4))[0] # Read vertex index (int, 4 bytes)
uv_offset = struct.unpack('<4f', file.read(16)) # Read UV offset (float, 4 * 4 bytes)
morph_data.append((vertex_index, uv_offset))
elif morph_type == 5: # UV extended2
vertex_index = struct.unpack('<i', file.read(4))[0]
uv_offset = struct.unpack('<4f', file.read(16))
vertex_index = struct.unpack('<i', file.read(4))[0] # Read vertex index (int, 4 bytes)
uv_offset = struct.unpack('<4f', file.read(16)) # Read UV offset (float, 4 * 4 bytes)
morph_data.append((vertex_index, uv_offset))
elif morph_type == 6: # UV extended3
vertex_index = struct.unpack('<i', file.read(4))[0]
uv_offset = struct.unpack('<4f', file.read(16))
vertex_index = struct.unpack('<i', file.read(4))[0] # Read vertex index (int, 4 bytes)
uv_offset = struct.unpack('<4f', file.read(16)) # Read UV offset (float, 4 * 4 bytes)
morph_data.append((vertex_index, uv_offset))
elif morph_type == 7: # UV extended4
vertex_index = struct.unpack('<i', file.read(4))[0]
uv_offset = struct.unpack('<4f', file.read(16))
vertex_index = struct.unpack('<i', file.read(4))[0] # Read vertex index (int, 4 bytes)
uv_offset = struct.unpack('<4f', file.read(16)) # Read UV offset (float, 4 * 4 bytes)
morph_data.append((vertex_index, uv_offset))
elif morph_type == 8: # Material
material_index = struct.unpack('<i', file.read(4))[0]
offset_type = struct.unpack('<b', file.read(1))[0]
diffuse_offset = struct.unpack('<4f', file.read(16))
specular_offset = struct.unpack('<3f', file.read(12))
specular_factor_offset = struct.unpack('<f', file.read(4))[0]
ambient_offset = struct.unpack('<3f', file.read(12))
edge_color_offset = struct.unpack('<4f', file.read(16))
edge_size_offset = struct.unpack('<f', file.read(4))[0]
texture_factor_offset = struct.unpack('<4f', file.read(16))
sphere_texture_factor_offset = struct.unpack('<4f', file.read(16))
toon_texture_factor_offset = struct.unpack('<4f', file.read(16))
material_index = struct.unpack('<i', file.read(4))[0] # Read material index (int, 4 bytes)
offset_type = struct.unpack('<b', file.read(1))[0] # Read offset type (byte, 1 byte)
diffuse_offset = struct.unpack('<4f', file.read(16)) # Read diffuse color offset (float, 4 * 4 bytes)
specular_offset = struct.unpack('<3f', file.read(12)) # Read specular color offset (float, 3 * 4 bytes)
specular_factor_offset = struct.unpack('<f', file.read(4))[0] # Read specular factor offset (float, 4 bytes)
ambient_offset = struct.unpack('<3f', file.read(12)) # Read ambient color offset (float, 3 * 4 bytes)
edge_color_offset = struct.unpack('<4f', file.read(16)) # Read edge color offset (float, 4 * 4 bytes)
edge_size_offset = struct.unpack('<f', file.read(4))[0] # Read edge size offset (float, 4 bytes)
texture_factor_offset = struct.unpack('<4f', file.read(16)) # Read texture factor offset (float, 4 * 4 bytes)
sphere_texture_factor_offset = struct.unpack('<4f', file.read(16)) # Read sphere texture factor offset (float, 4 * 4 bytes)
toon_texture_factor_offset = struct.unpack('<4f', file.read(16)) # Read toon texture factor offset (float, 4 * 4 bytes)
morph_data.append((material_index, offset_type, diffuse_offset, specular_offset, specular_factor_offset, ambient_offset, edge_color_offset, edge_size_offset, texture_factor_offset, sphere_texture_factor_offset, toon_texture_factor_offset))
return morph_name, morph_english_name, panel, morph_type, morph_data
@@ -179,42 +181,42 @@ def import_pmx(filepath):
version, encoding, additional_uvs, vertex_index_size, texture_index_size, material_index_size, bone_index_size, morph_index_size, rigid_body_index_size, model_name, model_english_name, model_comment, model_english_comment = read_pmx_header(file)
# Read vertices
vertex_count = struct.unpack('<i', file.read(4))[0]
vertex_count = struct.unpack('<i', file.read(4))[0] # Read vertex count (int, 4 bytes)
vertices = []
for _ in range(vertex_count):
position, normal, uv, bone_indices, bone_weights, edge_scale = read_vertex(file, vertex_index_size)
vertices.append((position, normal, uv, bone_indices, bone_weights, edge_scale))
# Read faces
face_count = struct.unpack('<i', file.read(4))[0]
face_count = struct.unpack('<i', file.read(4))[0] # Read face count (int, 4 bytes)
faces = []
for _ in range(face_count // 3):
face_indices = struct.unpack('<3i', file.read(12))
face_indices = struct.unpack('<3i', file.read(12)) # Read face indices (int, 3 * 4 bytes)
faces.append(face_indices)
# Read textures
texture_count = struct.unpack('<i', file.read(4))[0]
texture_count = struct.unpack('<i', file.read(4))[0] # Read texture count (int, 4 bytes)
textures = []
for _ in range(texture_count):
texture_path = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace')
texture_path = str(file.read(struct.unpack('<i', file.read(4))[0]), 'utf-16-le', errors='replace') # Read texture path (string, variable length)
textures.append(texture_path)
# Read materials
material_count = struct.unpack('<i', file.read(4))[0]
material_count = struct.unpack('<i', file.read(4))[0] # Read material count (int, 4 bytes)
materials = []
for _ in range(material_count):
material_name, material_english_name, diffuse_color, specular_color, specular_strength, ambient_color, flag, edge_color, edge_size, texture_index, sphere_texture_index, sphere_mode, toon_sharing_flag, toon_texture_index, comment = read_material(file, texture_index_size)
materials.append((material_name, material_english_name, diffuse_color, specular_color, specular_strength, ambient_color, flag, edge_color, edge_size, texture_index, sphere_texture_index, sphere_mode, toon_sharing_flag, toon_texture_index, comment))
# Read bones
bone_count = struct.unpack('<i', file.read(4))[0]
bone_count = struct.unpack('<i', file.read(4))[0] # Read bone count (int, 4 bytes)
bones = []
for _ in range(bone_count):
bone_name, bone_english_name, position, parent_bone_index, layer, flag, tail_position, inherit_bone_parent_index, inherit_bone_parent_influence, fixed_axis, local_x_vector, local_z_vector, external_key, ik_target_bone_index, ik_loop_count, ik_limit_radian, ik_links = read_bone(file, bone_index_size)
bones.append((bone_name, bone_english_name, position, parent_bone_index, layer, flag, tail_position, inherit_bone_parent_index, inherit_bone_parent_influence, fixed_axis, local_x_vector, local_z_vector, external_key, ik_target_bone_index, ik_loop_count, ik_limit_radian, ik_links))
# Read morphs
morph_count = struct.unpack('<i', file.read(4))[0]
morph_count = struct.unpack('<i', file.read(4))[0] # Read morph count (int, 4 bytes)
morphs = []
for _ in range(morph_count):
morph_name, morph_english_name, panel, morph_type, morph_data = read_morph(file, morph_index_size)