ci: enable deployment on huggingface spaces

Author: vemonet

Dockerfile

@@ -20,6 +20,17 @@ ENV SPARQL_FILE='/app/fetch_rhea.rq'
 # SPARQL endpoint to fetch compounds from
 ENV SPARQL_ENDPOINT='https://sparql.rhea-db.org/sparql'
 
+# Precompute fingerprints from the bundled compounds.tsv so the server starts
+# fast (no recompute on cold start). The cache is keyed on file content + fp
+# types, so the runtime entrypoint below reuses it directly.
+RUN uv run python -c "from mol_search_sparql_service.rdkit_fingerprints import engine; engine.load_file('compounds.tsv', cache_dir='.mol-search-service')"
+
+# Run as a non-root user (Hugging Face Spaces convention). Give it ownership of
+# /app so the precomputed cache and venv remain readable/writable at runtime.
+RUN useradd -m -u 1000 user && chown -R user:user /app
+USER user
+ENV HOME=/home/user
+
 EXPOSE 8010
 ENTRYPOINT ["sh", "-c", "uv run mol-search-sparql-service -f compounds.tsv -p $SERVER_PORT -w $WORKERS"]
 # ENTRYPOINT ["sh", "-c", "uv run mol-search-sparql-service  -s fetch_rhea.rq -e https://sparql.rhea-db.org/sparql -p $SERVER_PORT -w $WORKERS"]

README.md

@@ -1,3 +1,13 @@
+---
+title: Chemistry Search SPARQL Service
+emoji: 🫆
+colorFrom: indigo
+colorTo: blue
+sdk: docker
+app_port: 8010
+pinned: false
+---
+
 # 🫆 Chemistry Search SPARQL Service
 
 A service for finding similar chemicals and substructures using molecular fingerprints. It provides both a SPARQL endpoint and an MCP (Model Context Protocol) server interface.