Python-UNO bridge to libreoffice inside docker container.

docker/lo-ubuntu2204/Dockerfile

@@ -1,4 +1,3 @@
-# Stage 1: Build LibreOffice
 FROM ubuntu:22.04
 SHELL ["/bin/bash", "-c"]
 
@@ -18,6 +17,9 @@ RUN DEBIAN_FRONTEND="noninteractive" apt-get install -y \
         curl git vim sudo wget tzdata \
         x11-xserver-utils
 RUN apt-get install -y libreoffice
+RUN apt-get update && \
+    apt-get install -y python3 python3-pip python3-uno libreoffice-dev
+RUN pip install Flask
 
 # clean up the apt cache to reduce the image size
 RUN rm -rf /var/lib/apt/lists/*
@@ -31,11 +33,15 @@ RUN useradd --create-home -s /bin/bash $user \
         && adduser $user sudo \
         && echo "$user ALL=(ALL) NOPASSWD:ALL" >> /etc/sudoers
 
+COPY --chown=$user:$user container/* $home
+RUN chmod +x $home/start.sh
+RUN chmod +x $home/update-libreoffice-config.sh
+
 WORKDIR $home
 
 ENV USER=$user
 ENV SHELL=/bin/bash
 USER $user
 
-CMD ["/bin/bash"]
+CMD ["./start.sh"]
 

docker/lo-ubuntu2204/README.md

@@ -0,0 +1,39 @@
+# Run LibreOffice from Ubuntu 22.04 docker container
+
+The scripts in this folder allows you to run LibreOffice version 7.5.9 inside a docker
+container. Currently, there are two scripts that can be used to run LibreOffice:
+
+- `docker-soffic.sh` : This script runs LibreOffice the same way that you would use
+   outside the container. So running `./docker-soffice.sh main.fodt` will open the
+main document.
+- `start-container.sh` : This script will run LibreOffice as a daemon (a standalone background process that
+   listens to messages over a socket) inside the docker container. The host can then communicate
+   with this daemon using a Python UNO bridge script that is running inside the docker container.
+   The communication between the host machine and docker container is done using a rest API against a
+   flask web server running inside the docker container. The API currently only has a single
+   end point "open-document" which will open a given FODT document and also update its indexes.
+
+   To simplify the use of this end point, you can use the Python script `lodocker-open-file <file.fodt>`.
+   See information about installing the Python script below.
+
+## Installation of the python scripts
+- Requires python3 >= 3.10
+
+### Using poetry
+For development it is recommended to use poetry:
+
+- Install [poetry](https://python-poetry.org/docs/)
+- Then run:
+```
+$ poetry install
+$ poetry shell
+```
+
+### Installation into virtual environment
+If you do not plan to change the code, you can do a regular installation into a VENV:
+
+```
+$ python -m venv .venv
+$ source .venv/bin/activate
+$ pip install .
+```

docker/lo-ubuntu2204/build-image.sh

@@ -1,3 +1,3 @@
 #! /bin/bash
 
-docker build --network=host -t lo-ubuntu2204 .
+docker build -t lo-ubuntu2204 .

docker/lo-ubuntu2204/container/docker-server.py

@@ -0,0 +1,81 @@
+import logging
+import os
+#import subprocess
+from pathlib import Path
+
+from flask import Flask, request
+
+import uno
+
+from unohelper import systemPathToFileUrl, absolutize
+from com.sun.star.beans import PropertyValue
+
+app = Flask(__name__)
+
+def update_indexes(doc):
+    """
+    Update all indexes in the given document.
+    """
+    try:
+        indexes = doc.getDocumentIndexes()
+        for i in range(0, indexes.getCount()):
+            index = indexes.getByIndex(i)
+            index.update()
+        logging.info("Indexes updated successfully.")
+    except Exception as e:
+        logging.error("Error updating indexes and tables: " + str(e))
+
+def open_document_with_libreoffice(doc_path: str):
+    # Connect to the running instance of LibreOffice
+    local_context = uno.getComponentContext()
+    resolver = local_context.ServiceManager.createInstanceWithContext(
+        "com.sun.star.bridge.UnoUrlResolver", local_context
+    )
+    port = int(os.getenv('LIBREOFFICE_PORT', 2002))
+    ctx = resolver.resolve(f"uno:socket,host=localhost,port={port};urp;StarOffice.ComponentContext")
+    smgr = ctx.ServiceManager
+    desktop = smgr.createInstanceWithContext("com.sun.star.frame.Desktop", ctx)
+    # Load the document
+    cwd = systemPathToFileUrl( os.getcwd() )
+    path_ = "parts" / Path(doc_path)
+    file_url = absolutize( cwd, systemPathToFileUrl(str(path_)) )
+
+    load_props = []
+    #load_props.append(PropertyValue(Name="Hidden", Value=True))  # Open document hidden
+    #load_props.append(PropertyValue(Name="UpdateDocMode",
+    #            Value=uno.getConstantByName("com.sun.star.document.UpdateDocMode.QUIET_UPDATE")))
+    load_props = tuple(load_props)
+
+    logging.info("Loading {}".format(file_url))
+    doc = desktop.loadComponentFromURL(file_url, "_blank", 0, load_props)
+
+    update_indexes(doc)
+
+    # Save the document
+    # The user can save the document from the menu for now.
+    # If we want to automate the saving process we can do that by running libreoffice
+    # in headless mode using doc.store() as shown below
+    #doc.store()
+    #doc.dispose()
+    logging.info("Done")
+
+
+@app.route('/open-document', methods=['POST'])
+def open_document():
+    # Extract the document path from the request
+    doc_path = request.json.get('path')
+
+    # Replace this with the command to open the document using LibreOffice
+    open_document_with_libreoffice(doc_path)
+    return "Document opened", 200
+
+if __name__ == '__main__':
+    logging.basicConfig(level=logging.INFO)
+    # ASSUME: libreoffice deamon process has already been started and listening
+    #   on port 2002 at localhost :
+    #
+    #    libreoffice --accept="socket,host=localhost,port=2002;urp;" --pidfile=lo_pid.txt
+    #
+    port = int(os.getenv('FLASK_PORT', 8080))
+    app.run(debug=True, host='0.0.0.0', port=port)
+

docker/lo-ubuntu2204/container/open-blank-document.py

@@ -0,0 +1,20 @@
+
+import logging
+import os
+import uno
+
+try:
+    # Connect to the running instance of LibreOffice
+    local_context = uno.getComponentContext()
+    resolver = local_context.ServiceManager.createInstanceWithContext(
+        "com.sun.star.bridge.UnoUrlResolver", local_context
+    )
+    port = int(os.getenv('LIBREOFFICE_PORT', 2002))
+    ctx = resolver.resolve(f"uno:socket,host=localhost,port={port};urp;StarOffice.ComponentContext")
+    smgr = ctx.ServiceManager
+    desktop = smgr.createInstanceWithContext("com.sun.star.frame.Desktop", ctx)
+
+    # Open a blank document
+    document = desktop.loadComponentFromURL("private:factory/swriter", "_blank", 0, ())
+except Exception as e:
+    logging.error(f"Error in opening blank document: {e}")

docker/lo-ubuntu2204/container/start.sh

@@ -0,0 +1,56 @@
+#!/bin/bash
+
+# This file is run when the Docker container is started "docker run" without additional arguments.
+# We have wrapped the "docker run" in a script "start-container.sh" that takes care of
+# mounting the shared directory and the fonts directory, and also sets up the X11 connection.
+#
+# This script "start.sh" starts the LibreOffice daemon and the Python Flask server in the
+# Docker container that enables the host to communicate with the Docker container using a REST API.
+
+# Further, we would like to get rid of the popup dialog when opening a file with links to other files.
+#  The popup dialog has the message:
+#
+#  "The document main.fodt contains one or more links to external data.
+#   Would you like to change the document, and update all links to get
+#   the most recent data?", see:
+#
+# https://ask.libreoffice.org/t/avoid-popup-dialog-at-startup-would-you-like-to-update-all-links-to-get-the-most-recent-data/99189
+#
+# I first tried to pass "com.sun.star.document.UpdateDocMode.QUIET_UPDATE" to the loadComponentFromURL()
+# method in the python script "docker-server.py", but that did not work.
+# I have submitted a PR to fix this: https://gerrit.libreoffice.org/c/core/+/161628
+#
+# In the mean time, I found a workaround by modifying the user profile file
+#  "registrymodifications.xcu". This file is located in the directory
+#  "/home/docker-user/.config/libreoffice/4/user/" and needs to be modified
+#  as shown in the script "update-libreoffice-config.sh".
+#
+# However, we cannot modify this file before LibreOffice has been started once
+#  to initialize the user profile. Therefore we start LibreOffice, wait 5 seconds,
+#  kill it, modify the file, and start it again. See below:
+
+libreoffice --accept="socket,host=localhost,port=${LIBREOFFICE_PORT};urp;" --headless \
+  --norestore --nofirststartwizard --nologo --nodefault --pidfile=/home/docker-user/lo_pid.txt &
+
+
+echo "Waiting 5 seconds for LibreOffice to start and intialize..."
+sleep 5
+
+# Call a script that opens a blank document in LibreOffice Writer
+python3 open-blank-document.py
+echo "Waiting 5 seconds for LibreOffice to open a blank document..."
+sleep 5
+echo "Killing LibreOffice..."
+kill $(cat /home/docker-user/lo_pid.txt)
+
+# Modify the user profile to get rid of popup dialog when opening a file with
+# links to other files
+./update-libreoffice-config.sh
+
+echo "Restarting LibreOffice..."
+# Restart LibreOffice with the modified user profile
+libreoffice --accept="socket,host=localhost,port=${LIBREOFFICE_PORT};urp;" \
+  --norestore --nofirststartwizard --nologo --nodefault --pidfile=/home/docker-user/lo_pid.txt &
+
+# Start the Flask server
+exec python3 docker-server.py

docker/lo-ubuntu2204/container/update-libreoffice-config.sh

@@ -0,0 +1,38 @@
+#!/bin/bash
+
+# This file is needed to avoid the popup dialog with message:
+#  "The document main.fodt contains one or more links to external data.
+#   Would you like to change the document, and update all links to get
+#   the most recent data?", see:
+#
+# https://ask.libreoffice.org/t/avoid-popup-dialog-at-startup-would-you-like-to-update-all-links-to-get-the-most-recent-data/99189
+#
+
+CONFIG_FILE="/home/docker-user/.config/libreoffice/4/user/registrymodifications.xcu"
+TEMP_FILE="/home/docker-user/.config/libreoffice/4/user/temp_registrymodifications.xcu"
+
+# Append the SecureURL item
+append_secure_url() {
+    echo '<item oor:path="/org.openoffice.Office.Common/Security/Scripting"><prop oor:name="SecureURL" oor:op="fuse"><value><it>$(home)/parts</it></value></prop></item>' >> "$TEMP_FILE"
+}
+
+# Append the Update LinkMode item
+append_update_link_mode() {
+    echo '<item oor:path="/org.openoffice.Office.Writer/Content/Update"><prop oor:name="Link" oor:op="fuse"><value>2</value></prop></item>' >> "$TEMP_FILE"
+}
+
+# Copy original content excluding the last line (</oor:items>)
+head -n -1 "$CONFIG_FILE" > "$TEMP_FILE"
+
+# Append new configuration items
+append_secure_url
+append_update_link_mode
+
+# Append the final closing tag
+echo '</oor:items>' >> "$TEMP_FILE"
+
+# Replace the original file with the modified one
+mv "$TEMP_FILE" "$CONFIG_FILE"
+
+# Display a message
+echo "Modified registrymodifications.xcu successfully"