jupyterlab_gcsfilebrowser
- Provides a mechanism browse and interact with
Google Cloud Storage through a file browser.
- Python 3.5+
- JupyterLab
- Virtualenv (Recommended for local development)
This should work on Google Cloud Deep Learning VM M19+.
This extension requires credentials to exist on the JupyterLab server.
If credentials do not exist on the server, follow the Google Cloud Getting Started with Authentication instructions.
Use the deploy-latest.sh script to upload and install from the latest publicly released tarball on a DLVM over SSH using the instance name. Requires gcloud from the Google Cloud SDK to be installed.
./deploy-latest.sh ${INSTANCE_NAME?}
Use the deploy.sh script to build the extension locally, upload, and install on a DLVM over SSH using the instance name. Requires gcloud from the Google Cloud SDK to be installed.
./deploy.sh ${INSTANCE_NAME?}
# Build the Python source distribution package
local$ python setup.py sdist
# Copy the dist/jupyterlab_gcsfilebrowser-x.x.x.tar.gz archive to the JupyterLab
# server
# Install the Python package
server$ sudo pip3 install jupyterlab_gcsfilebrowser-x.x.x.tar.gz
# Force Jupyter to rebuild the front-end packages
server$ sudo jupyter lab build
server$ sudo service jupyter restart
For a development install (requires npm version 4 or later), do the following in the repository directory:
You will need to have Python3, virtualenv, and npm installed.
# Create a Python 3 virtualenv and install jupyterlab and the project in edit mode
virtualenv -p python3 venv
source venv/bin/activate
# Install the version of jupyterlab used by DLVM images
pip install jupyterlab
pip install .
# Install the npm package and the extension
npm install
jupyter labextension install . --no-build
# Now, run npm start which starts the Typescript compiler in watch mode on the
# extension directory as well as the JupyterLab server
npm start
See: go/jupyterlab-gcsfilebrowser-release-notes