File size: 2,822 Bytes
6237959 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 | """Depth-buffered previews without a graphics server. Requires a C++ compiler."""
from pathlib import Path
import ctypes
import hashlib
import math
import subprocess
import tempfile
import numpy as np
from PIL import Image
def render_parts(parts, visible, moved, colours, path, state, mode, camera, size):
cpp=Path(__file__).with_name('rasterizer.cpp')
tag=hashlib.sha256(cpp.read_bytes()).hexdigest()[:12]
libpath=Path(tempfile.gettempdir())/f'orcasat_raster_{tag}.so'
if not libpath.is_file():
subprocess.run(['g++','-O3','-shared','-fPIC','-std=c++17',str(cpp),'-o',str(libpath)],check=True)
lib=ctypes.CDLL(str(libpath))
ptrd=np.ctypeslib.ndpointer(dtype=np.float64,flags='C_CONTIGUOUS')
ptrf=np.ctypeslib.ndpointer(dtype=np.float32,flags='C_CONTIGUOUS')
ptrb=np.ctypeslib.ndpointer(dtype=np.uint8,flags='C_CONTIGUOUS')
lib.rasterize.argtypes=[ctypes.c_int,ptrd,ptrf,ptrf,ctypes.c_int,ctypes.c_int,ptrd,ptrd,ptrb]
tri=[];normals=[];cc=[]
direction=np.array([camera[0],camera[1],camera[2]*.8],dtype=float)
direction/=np.linalg.norm(direction)
right=np.cross([0,0,1],direction);right/=np.linalg.norm(right)
up=np.cross(direction,right)
basis=np.stack([right,up,direction],axis=1)
from export_glb import tessellate_absolute
for p in parts:
if not visible(p,state,mode):continue
vs,ts=tessellate_absolute(moved(p,mode))
vv=np.array([[v.x,v.y,v.z] for v in vs],dtype=np.float64)
tt=np.array(ts,dtype=np.int32)
face=np.cross(vv[tt[:,1]]-vv[tt[:,0]],vv[tt[:,2]]-vv[tt[:,0]])
vn=np.zeros_like(vv)
for i in range(3):np.add.at(vn,tt[:,i],face)
vn/=np.maximum(np.linalg.norm(vn,axis=1)[:,None],1e-15)
keep=face@direction>1e-12
tt=tt[keep]
tri.append((vv@basis)[tt]);normals.append(vn[tt]);cc.append(np.tile(colours[p.colour],(len(tt),1)))
triangles=np.concatenate(tri)
nm=np.ascontiguousarray(np.concatenate(normals),dtype=np.float32)
color=np.ascontiguousarray(np.concatenate(cc),dtype=np.float32)
flat=triangles.reshape(-1,3);lo=flat.min(0);hi=flat.max(0)
ss=2;w,h=size[0]*ss,size[1]*ss
scale=min((w-100*ss)/(hi[0]-lo[0]),(h-100*ss)/(hi[1]-lo[1]))
centre=(lo+hi)/2
triangles[:,:,0]=(triangles[:,:,0]-centre[0])*scale+w/2
triangles[:,:,1]=h/2-(triangles[:,:,1]-centre[1])*scale
triangles=np.ascontiguousarray(triangles,dtype=np.float64)
light=up*.85-right*.65+direction*1.05;light/=np.linalg.norm(light)
image=np.empty((h,w,3),dtype=np.uint8)
lib.rasterize(len(triangles),triangles,nm,color,w,h,
np.ascontiguousarray(light,dtype=np.float64),
np.ascontiguousarray(direction,dtype=np.float64),image)
Image.fromarray(image).resize(size,Image.Resampling.LANCZOS).save(path)
|