# Copyright 2025 The HuggingFace Inc. team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

# Benchmark image for VLABench integration tests.
# Extends the nightly GPU image with the PR's source code and VLABench setup.
#
# Build:  docker build -f docker/Dockerfile.benchmark.vlabench -t lerobot-benchmark-vlabench .
# Run:    docker run --gpus all --rm lerobot-benchmark-vlabench lerobot-eval ...

FROM huggingface/lerobot-gpu:latest

# Install VLABench from GitHub (not on PyPI) and pin MuJoCo/dm-control.
# Shallow-clone without submodule recursion (nested SSH-only submodules fail in CI).
# Editable install (-e) because VLABench/utils/ has no __init__.py, so
# find_packages() omits it from wheels; editable mode uses the source tree directly.
# rrt-algorithms has the same packaging issue (rrt/ dir missing __init__.py).
# Patch: constant.py calls os.listdir on ~100 asset/obj/meshes/* dirs at import
# time. Guard the call so missing dirs return [] instead of crashing (in case
# the asset download is partial).
RUN git clone --depth 1 https://github.com/OpenMOSS/VLABench.git ~/VLABench && \
    git clone --depth 1 https://github.com/motion-planning/rrt-algorithms.git ~/rrt-algorithms && \
    python3 -c "\
import pathlib; \
p = pathlib.Path.home() / 'VLABench/VLABench/configs/constant.py'; \
t = p.read_text(); \
p.write_text(t.replace( \
    'subdirs = os.listdir(xml_dir)', \
    'if not os.path.isdir(xml_dir): return []\n    subdirs = os.listdir(xml_dir)'))" && \
    uv pip install --no-cache -e ~/VLABench -e ~/rrt-algorithms \
      mujoco==3.2.2 dm-control==1.0.22 \
      open3d colorlog scikit-learn openai gdown

# Download VLABench mesh assets. Task configs reference object meshes
# (obj/meshes/fruit/, containers/basket/, etc.); without them the task builder
# picks from an empty mesh list and crashes at task-build time.
#
# Preferred source: an HF Hub mirror. Set VLABENCH_ASSETS_REPO at build time
# (e.g. --build-arg VLABENCH_ASSETS_REPO=lerobot/vlabench-assets) and we'll
# snapshot_download the repo into VLABench's assets dir. This is the reliable
# path for CI — Google Drive frequently returns HTTP 429 ("Too many users have
# viewed or downloaded this file recently") on shared academic files.
#
# Fallback: VLABench's own gdown-based script. Best-effort only; the build
# will NOT fail if gdown hits a Drive quota, letting us ship an image where
# task-build may still IndexError but the rest of the pipeline is exercised.
ARG VLABENCH_ASSETS_REPO=""
RUN ASSETS_DIR="$HOME/VLABench/VLABench/assets" && \
    if [ -n "${VLABENCH_ASSETS_REPO}" ]; then \
        echo "Downloading VLABench assets from HF Hub: ${VLABENCH_ASSETS_REPO}" && \
        python -c "from huggingface_hub import snapshot_download; \
snapshot_download(repo_id='${VLABENCH_ASSETS_REPO}', repo_type='dataset', local_dir='${ASSETS_DIR}')"; \
    else \
        echo "No VLABENCH_ASSETS_REPO set — falling back to gdown (best-effort)" && \
        python ~/VLABench/scripts/download_assets.py --choice all || \
          echo "WARN: VLABench asset download failed (likely Google Drive quota). Env task-build may crash."; \
    fi

# Overlay the PR's source code on top of the nightly image.
COPY --chown=user_lerobot:user_lerobot . .

# Re-install lerobot editably so the new source (with VLABenchEnv registration
# and updated obs handling) replaces the stale package baked into the nightly image.
RUN uv pip install --no-cache --no-deps -e .

CMD ["/bin/bash"]