This library aims for headless rendering of three.js scenes based on models by common CAD formats.
Currently supports IFC models.
This library uses a custom build of web-ifc-three.
$ yarn install
$ npm_config_build_from_source=true yarn add --force https://github.com/bldrs-ai/web-ifc-three.git
$ yarn build
# Grab the coords from a model on Bldrs, e.g. the default homepage:
# https://bldrs.ai/share/v/p/index.ifc#c:-150.147,-85.796,167.057,-32.603,17.373,-1.347
#
# Then add them as an arg after the model filename:
headless-three> node src/headless.js models/index.ifc -150.147,-85.796,167.057,-32.603,17.373,-1.347 && open screenshot.png
headless-three> node src/headless.js models/Bunny.obj 0,.3,.25,0,0,-0.1 && open screenshot.png
Example render of index.ifc:
To run the server locally, you will need a working Docker installation.
$ docker compose build # Build the Docker image
$ docker compose up -d # Start the server, accessible at http://localhost:8001
$ curl -d '{"url": "http://server.com/path/to/a/model.ifc"}' \
-H 'content-type: application/json' \
-o rendered.png \
http://localhost:8001/render
for f in `ls models/*/*.{bld,fbx,ifc,obj,stl,pdb,xyz}` ; do curl -f -d "{\"url\": \"http://localhost:8090/$f\"}" -H 'content-type: application/json' -o "$f"-fit.png --fail --silent --show-error -D- http://localhost:8001/render > log ; grep -q '200 OK' log || /bin/mv -f log $f.err ; done