Juttle Engine is an integrated environment for developing and executing
juttle programs and visualizations. It lets
you run Juttle programs stored on the file system and present the results in a
browser for experimentation, development, deployment, and debugging of juttle
To make it easy to interconnect analytics with visualization,
Juttle promotes a tight coupling between the language and the client-side
visualization library. Rather than
having to tie together by hand a data processing layer with a separate
visualization layer, Juttle enables you to integrate these two layers
so you don't have to worry about the details
of wiring Juttle dataflow computation to your browser-based views.
Juttle Engine is the development and execution environment in which you can
explore your data using Juttle and deploy rich visualizations into production.
Under the covers, Juttle Engine integrates
juttle-service, a node.js API server
that enables execution of Juttle programs using a REST API with the ability to
serve web application bundles. It comes with the
juttle-viewer application that
provides a simple in browser experience for loading and executing juttle
programs in the juttle-engine and rendering the results using the
juttle-viz visualization library. In the
future, Juttle Engine may be extended to support other application bundles.
Use npm to install Juttle Engine
$ npm install juttle-engine
This will install the Juttle Engine daemon and client binaries, the juttle interpreter, and the current set of supported juttle adapters.
We believe it to work with node.js 4.2.3 and 5.0, npm 2.14.17 and 3.5, and likely other combinations we have not tested against.
Running your first juttle program
Start the daemon by running juttle-engine:
$ juttle-engine -d -o juttle-engine.log
This will daemonize juttle-engine and output logs to juttle-engine.log. You can now run juttle programs against the juttle-engine daemon via the juttle-engine-client:
$ juttle-engine-client browser --path examples/core-juttle/hello_world.juttle
This will open a browser window and display the output of the program. You can make edits to your juttle, save the file, and reload the browser window to get the updated output.
We also offer a tutorial for developers new to Juttle.
Here's how the juttle-engine module fits into the overall Juttle Ecosystem:
Options and Configuration
Here are the full command line options supported by the daemon and client programs:
usage: [--port <port>] [--root <path>] [--config <juttle-config-path>] [--daemonize] [--output <logfile>] [--log-config <log4js-config-path>] [--log-level <level] [--help] -p, --port <port>: Run juttle-engine on specified port -r, --root <path>: Use <path> as the root directory for juttle programs -c, --config <juttle-config-path>: Read juttle config from <juttle-config-path> -d, --daemonize: Daemonize juttle-engine and log to configured log file -o, --output <logfile>: Log to specififed file when daemonized -L, --log-config <log4js-config-path>: Configure logging from <log4js-config-path>. Overrides any value of -o -l, --log-level <level>: Use a default log level of <level>. Overridden by any log level specified in -L -h, --help: Print this help and exit');
juttle-engine uses log4js for logging and by default logs to
usage: [--juttle-engine <hostname:port>] [--help] [COMMAND] [OPTIONS] [COMMAND]: one of the following, with the following options: subscribe (--job <job-id> | --observer <observer-id>) list_jobs [--job <job-id>] list_observers run --path <path-to-juttle-file> [--wait] [--observer <observer-id>] delete --job <job-id> get_inputs --path <path-to-juttle-file> --input name=val [--input name=val ...] push --path <path-to-juttle-file> [--topic <rendezvous-topic>] watch --path <path-to-juttle-file> [--topic <rendezvous-topic>] browser --path <path-to-juttle-file> [OPTIONS]: one of the following: --path <path-to-juttle-file> Path to file relative to configured root directory. used by: run,get_inputs,push,watch,browser --wait If true, wait for program to finish, otherwise run in background used by: run --input name=val One or more input values. used by: get_inputs --job <job-id> Job id. used by: subscribe,list_jobs,delete --observer <observer-id> Observer id. used by: subscribe,run --juttle-engine <hostname:port> Hostname/port of server --help Print this help and exit --topic <rendezvous-topic> Rendezvous topic used by: push,watch
Juttle config file
The Juttle compiler and runtime within juttle-engine are also configured via the juttle configuration file, typically at
$(HOME)/.juttle/config.json. For more information on the juttle configuration file, see the juttle configuration documentation.
For information on how modules are resolved, see the "Importing Modules" section in the general juttle documentation on programming constructs.
To run unit tests:
To check code style and perform lint checks:
Both are run automatically by Travis.