Name: binder-client
Owner: Binder
Description: Command line tool for interfacing with the Binder API
Created: 2015-10-03 02:24:11.0
Updated: 2018-02-03 16:43:51.0
Pushed: 2017-08-18 02:46:55.0
Size: 27
Language: JavaScript
GitHub Committers
User | Most Recent Commit | # Commits |
---|
Other Committers
User | Most Recent Commit | # Commits |
---|
:books: Same functionality. Better performance for you. :books:
Over the past few months, we've been improving Binder's architecture and infrastructure. We're retiring this repo as it will no longer be actively developed. Future development will occur under the JupyterHub organization.
Thanks for updating your bookmarked links.
Command line tool for managing Binder services: build, register, launch. Implements both a programmatic interface and a CLI interface for the protocol specification in binder-protocol
.
Make it easy to mix and match use cases and environments: local or cloud deployment, building images or launching images directly, etc.
install binder-client -g
This module can be imported as a module, or used directly from the CLI (and the usage is pretty much the same in both cases). Every method that's exposed by the client takes at least a host and a port as parameters, and authorized endpoints also require an API key.
The client can communicate with any endpoint in the Binder API using the following format (see examples):
er (build|registry|deploy) (command) [options]
er.(build|registry|deploy).(command)(options, function (err, result) {
.
Unauthorized commands don't require an API token
er deploy deploy binder-project-example-requirements --host='deploy.mybinder.org' --port=8084'
Endpoints are specified as camel-case in binder-protocol
, but they should be translated to kebab-case before use with the client:
er registry fetch-all --host='build.mybinder.org' --port=8082 --api-token
If an endpoint has GET query parameters (i.e. binder registry fetch
) they are specified as additional CLI arguments
er registry fetch binder-project-example-requirements <deploy-id> --host='local' -api-token='blah'
Translating CLI commands into programmatic API commands is straightforward (CLI arguments just need to be inserted into an options object):
To start building an image/template for a repository:
buildOpts = {
st: '<build server host>',
rt: 8082,
iKey: <api key>,
pository: '<repo name>'
er.build.start(buildOpts, function (err, status) {
.
To query the status of a single build:
buildOpts = {
st: '<build server host>',
rt: 8082,
iKey: <api key>,
pository: '<repo name>'
er.build.status(buildOpts, function (err, status) {
.