feat(DATAGO-123135): Add ability to preview binary artifacts (PDF. PPTX, DOCX)

.github/workflows/ci.yaml

@@ -189,6 +189,8 @@ jobs:
           cache-to: ${{ steps.image_name.outputs.cache_to }}
           provenance: false
           sbom: false
+          build-args: |
+            INSTALL_LIBREOFFICE=${{ secrets.INSTALL_LIBREOFFICE || 'false' }}
 
   # ----------------------------------------------------
   # PHASE 3: Merge Multi-Platform Manifest (Always)

Dockerfile

@@ -129,24 +129,48 @@
 ENV PYTHONUNBUFFERED=1
 ENV PATH="/opt/venv/bin:$PATH"
 
+# Build argument to control LibreOffice installation for binary artifact preview
+# LibreOffice is NOT installed by default to keep image size smaller
+# To enable binary artifact preview (DOCX/PPTX), set:
+#   docker build --build-arg INSTALL_LIBREOFFICE=true -t sam .
+# Or via environment variable:
+#   INSTALL_LIBREOFFICE=true docker build --build-arg INSTALL_LIBREOFFICE -t sam .
+#
+# IMPORTANT: LibreOffice is a separate open-source application licensed under MPL-2.0.
+# See THIRD_PARTY_LICENSES/LIBREOFFICE.md for full license and attribution details.
+# Source code: https://www.libreoffice.org/download/source-code/
+ARG INSTALL_LIBREOFFICE
+
 # Copy Node.js 25 from the official node image
 COPY --from=node-binaries /usr/local/bin/node /usr/local/bin/node
 COPY --from=node-binaries /usr/local/bin/npm /usr/local/bin/npm
 COPY --from=node-binaries /usr/local/bin/npx /usr/local/bin/npx
 COPY --from=node-binaries /usr/local/lib/node_modules /usr/local/lib/node_modules
 
-# Install minimal runtime dependencies (no uv for licensing compliance, no curl - due to vulnerabilities)
+# Install minimal runtime dependencies
 # Add unstable repo with APT pinning to only upgrade libtasn1-6 (CVE-2025-13151 fix)
+# LibreOffice is optionally installed for document conversion (DOCX/PPTX to PDF for preview)
 RUN echo "deb http://deb.debian.org/debian unstable main" > /etc/apt/sources.list.d/unstable.list && \
     printf "Package: *\nPin: release a=unstable\nPin-Priority: 50\n\nPackage: libtasn1-6\nPin: release a=unstable\nPin-Priority: 900\n" > /etc/apt/preferences.d/99pin-libtasn1 && \
     apt-get update && \
     apt-get install -y --no-install-recommends \
     ffmpeg=7:7.1.3-0+deb13u1 \
-    git \
-    libatomic1 \
-    libtasn1-6/unstable \
-    libssl3t64=3.5.4-1~deb13u2 \
-    openssl=3.5.4-1~deb13u2 && \
+    git && \
+    if [ "${INSTALL_LIBREOFFICE}" = "true" ]; then \
+        echo "============================================================" && \
+        echo "NOTICE: Installing LibreOffice - a separate open-source application" && \
+        echo "LibreOffice is licensed under Mozilla Public License 2.0 (MPL-2.0)" && \
+        echo "License: https://www.mozilla.org/en-US/MPL/2.0/" && \
+        echo "Source:  https://www.libreoffice.org/download/source-code/" && \
+        echo "See THIRD_PARTY_LICENSES/LIBREOFFICE.md for full attribution" && \
+        echo "============================================================" && \
+        apt-get install -y --no-install-recommends \
+        libreoffice-writer-nogui \
+        libreoffice-impress-nogui \
+        libreoffice-calc-nogui; \
+    else \
+        echo "Skipping LibreOffice installation (set INSTALL_LIBREOFFICE=true to enable binary artifact preview)"; \
+    fi && \
     apt-get clean && \
     rm -rf /var/lib/apt/lists/* /etc/apt/sources.list.d/unstable.list /etc/apt/preferences.d/99pin-libtasn1
 
@@ -189,9 +213,9 @@
 ENV SOLACE_DEV_MODE=True
 
 # Set the following environment variables to appropriate values before deploying
 ENV SESSION_SECRET_KEY="REPLACE_WITH_SESSION_SECRET_KEY"
 ENV LLM_SERVICE_ENDPOINT="REPLACE_WITH_LLM_SERVICE_ENDPOINT"
 ENV LLM_SERVICE_API_KEY="REPLACE_WITH_LLM_SERVICE_API_KEY"
 ENV LLM_SERVICE_PLANNING_MODEL_NAME="REPLACE_WITH_PLANNING_MODEL_NAME"
 ENV LLM_SERVICE_GENERAL_MODEL_NAME="REPLACE_WITH_GENERAL_MODEL_NAME"
 

THIRD_PARTY_LICENSES/LIBREOFFICE.md

@@ -0,0 +1,49 @@
+# LibreOffice Third-Party License Notice
+
+## About LibreOffice
+
+LibreOffice is a free and open-source office productivity software suite developed by The Document Foundation. When the SAM Docker image is built with `INSTALL_LIBREOFFICE=true`, the unmodified LibreOffice packages from Debian are installed as a **separate application** alongside SAM to provide document conversion capabilities (DOCX, PPTX, XLSX to PDF).
+
+**Important:** LibreOffice is NOT bundled by default. It is only installed when explicitly requested via the `INSTALL_LIBREOFFICE=true` build argument.
+
+## License
+
+LibreOffice is licensed under the **Mozilla Public License Version 2.0 (MPL-2.0)**.
+
+Some components may also be available under the GNU Lesser General Public License (LGPL).
+
+### Mozilla Public License Version 2.0
+
+The full text of the MPL-2.0 license can be found at:
+- https://www.mozilla.org/en-US/MPL/2.0/
+
+### LGPL License
+
+The full text of the LGPL can be found at:
+- https://www.gnu.org/licenses/lgpl-3.0.html
+
+## Source Code
+
+LibreOffice source code is available from:
+- **Official website:** https://www.libreoffice.org/download/source-code/
+- **Git repository:** https://git.libreoffice.org/core
+- **Debian source packages:** Available via `apt-get source libreoffice-writer-nogui` (when using Debian-based images)
+
+## How LibreOffice is Used
+
+SAM uses LibreOffice in **headless mode** via command-line interface to convert Office documents (DOCX, PPTX, XLSX) to PDF format for in-browser preview. The conversion is performed by calling the `soffice` binary with the `--headless --convert-to pdf` flags.
+
+No modifications have been made to LibreOffice's source code. The standard Debian packages are installed as-is:
+- `libreoffice-writer-nogui`
+- `libreoffice-impress-nogui`
+- `libreoffice-calc-nogui`
+
+## Additional Information
+
+- **LibreOffice Official Website:** https://www.libreoffice.org/
+- **The Document Foundation:** https://www.documentfoundation.org/
+- **LibreOffice License FAQ:** https://www.libreoffice.org/about-us/licenses/
+
+---
+
+*This notice is provided in compliance with LibreOffice's licensing requirements. LibreOffice is a trademark of The Document Foundation.*