callisto-nbviewer 1.0.2

Jupyter notebook viewer

callisto

A jupyter notebook viewer that has the following features:

• Use ContentsManager as content provider. Therefore it is completely work with any of your jupyterhub settings.
• Showing Table of Contents for quickly reference to your notebook
• Directly import the notebook to your jupyterhub server.

Getting Started

1. Install the package:

pip install callisto-nbviewer

2. start the server:

callisto start --port 5000

3. Visit website localhost:5000

Configuration

You can use your own config to customize callisto. see example on tests/fixtures/config.py

ontents_manager_cls = "s3contents.S3ContentsManager"
# `contents_manager_cls` can be string or ContentsManger class
# You can use any existing ContentsManager to load data.
# If the value is not set, the default would be FileContentsManager
# and will serve contents on current working directory


contents_manager_kwargs = {
    "bucket": "test-bucket",
    "access_key_id": "testing",
    "secret_access_key": "testing",
    "endpoint_url": "http://localhost:3000"
}
# [Optional] `contents_manager_kwargs` is a dict of kwargs to configure the required settings
# for whatever ContentsManager you choose.


jupyterhub_base_url = "https://jupyterhub.example.com"
# [Optional] `jupyterhub_base_url` if set, you can import the notebook directly to your jupyterhub server
# This is the base url to your jupyterhub.

import_link_with_hubshare_preview = True
# [Optional] `import_link_with_hubshare_preview`: if you have installed jupyterlab-hubshare plugin,
# You can use this link to show preview on your notebook server


import_link_func = lambda path: "prefix/" + path
# [Optional] `import_link_func` is a function in that takes the current path as an input,
# You can do any magic to update the final path in case the path on the site is different
# from the path on jupyterhub.

and then start callisto with

callisto --config my_callisto_config.py

Config for different contents manager

Local File

contents_manager_cls = "managerjupyter_server.services.contents.filemanager.FileContentsManager"
contents_manager_kwargs = {"root_dir": "/absolute/path/to/your/notebook/folder"}

S3

Note that s3contents seems to be fetching the directory last_modified time, and is very slow. Therefore I make a work around file in callisto.contents_managers.s3.SimplifiedS3ContentsManager class please take a look on the settings.

contents_manager_cls = "callisto.contents_managers.s3.SimplifiedS3ContentsManager"
contents_manager_kwargs = {
    "bucket": "my-s3-bucket",
    "prefix": "prefix/to/your/notebook/folder",
    "aws_access_key_id": "my_access_key",
    "aws_secret_access_key": "my_secret_access_key",
}

if your file s3 doesn't contains a lot of folders, you can still use s3contents (please ensure you also install s3contents)

contents_manager_cls = "s3contents.S3ContentsManager"
contents_manager_kwargs = {
    "bucket": "my-s3-bucket",
    "prefix": "prefix/to/your/notebook/folder",
}

for more settings, please refer to: s3contents

Development

to start a dev version, download the git repo:

git clone https://github.com/lydian/callisto.git

and run

cd callisto
make start-dev

The dev server will running on http://localhost:5001