datagrok-tools

Datagrok-tools

Utility to upload and publish packages to Datagrok.

Installation

npm install datagrok-tools -g

Usage

1. Configure your environment with the following command:

   grok config
   

   Enter developer keys and set the default server. The developer key can be retrieved from your user profile (for example, see https://public.datagrok.ai/u).

2. Create a new package by running this command:

   grok create <package-name>
   

   A new folder <package-name> will be created automatically as well as its contents.

3. Run npm install in your package directory to get the required dependencies (the command is called automatically after package creation, if it ran successfully, skip this step).

4. Start working on the functionality of your package. Use grok add to create function templates.

5. Once you have completed the work on your package, upload it by running:

   grok publish
   

Run grok for instructions and grok <command> --help to get help on a particular command.

Read more about package development in Datagrok's documentation.

Commands

• config creates or updates a configuration file. The command shows the location of the config file with the developer keys and offers to interactively change them. It is also possible to reset the current configuration.

• create adds a package template to the current working directory when used without the name argument. The directory must be empty:

  grok create
  

  When called with an argument, the command creates a package in a folder with the specified name:

  grok create <package-name>
  

  Package name may only include letters, numbers, underscores, or hyphens. Read more about naming conventions here. Options:

  • --eslint adds a basic configuration for eslint
  • --ide adds an IDE-specific configuration for debugging (vscode)
  • --js creates a JavaScript package template
  • --ts creates a TypeScript package template (default)
  • --test adds tests support to package

• add puts an object template to your package:

  cd <package-name>
  grok add app <name>
  grok add connection <name>
  grok add detector <semantic-type-name>
  grok add function [tag] <name>
  grok add query <name>
  grok add script [tag] <language> <name>
  grok add view <name>
  grok add viewer <name>
  grok add tests
  

  In general, entity names follow naming rules for functions. Views and viewers should have class names, we recommend that you postfix them with 'View' and 'Viewer' respectively. Supported languages for scripts are javascript, julia , node, octave, python, r. Available function tags: panel, init.

• api creates wrapper functions for package scripts and queries. The output is stored in files /src/scripts-api.ts and /src/queries-api.ts respectively.

• publish uploads a package to the specified server (pass either a URL or a server alias from the config.yaml file). Additionally, you can use placeholders in JSON files under the /connections folder to substitute environment variables. For more information on configuring connections, refer to the Connections article.

  cd <package-name>
  grok publish [host]
  

  Options:

  • --build or --rebuild: boolean flags that indicate whether a local webpack bundle should be used or it should be generated on the server side
  • --debug or --release: boolean flags that determine whether to publish a debug version of the package visible only to the developer or a release version accessible by all eligible users and user groups
  • --key: a string containing a developer key that is not listed in the config file, e.g., the key for a new server
  • --suffix: a string containing package version hash

  Running grok publish is the same as running grok publish defaultHost --build --debug.

• check checks package content (function signatures, import statements of external modules, etc.). The check is also run during package publication.

• init modifies a package template by adding config files for linters, IDE, and so on (applies the same options as create).

• test runs package tests. First, it builds a package and publishes it (these steps can be skipped with flags --skip-build and --skip-publish correspondingly).

  cd <package-name>
  grok test
  

  The default host from the config.yaml file is used. You can specify another server alias via the --host option (e.g., grok test --host=dev). The results are printed to the terminal. To see tests execution, pass the --gui flag that disables the headless browser mode. If you want to save a test run result, add the --csv flag (the report will be saved in a CSV file in the package folder). You can find more details in local package testing instructions.

  Table with all available flags:

  Flag	Description
  catchUnhandled	Whether or not to catch unhandled exceptions during tests
  category	Runs tests of a specific category
  csv	Saves tests run results as a CSV file in the package folder
  gui	Disables the headless browser mode
  host	Specify server alias
  record	Records the test execution process in mp4 format
  report	Sends a report on test results
  skip-build	Skips package build step
  skip-publish	Skips package publish step
  verbose	Prints detailed information about passed and skipped tests in the console
  platform	Runs only platform tests (applicable for ApiTests package only)
  core	Runs package & core tests (applicable for DevTools package only)

• link command is used for public plugins development to link datagrok-api and libraries.