Interactive cli tool that lists and run package.json scripts.
An iPipeTo workflow
- Install
- Usage
- Features
- Customize
- Repeat the last ran task
- Run multiple tasks
- Run in autocomplete or fuzzy search mode
- Tips
- Help
- License
$ npm install -g ntl
Navigate to any folder containing a package.json file (usually a Node.js project) that has configured scripts, then just use the ntl command:
ntlYou can also specify a path to a project folder containing a package.json file:
ntl ./my-node-project
- Interactive interface that lists all
package.jsonscripts - Select any item in the interactive interface to execute that task
- Add descriptions to each task that can be shown in the UI
- Multiple interactive interfaces (menu list, autocomplete fuzzy search)
- Many options to customize the UI (exclude scripts, amount of items, etc)
- Easy to repeat last ran script (
ntandrerunoptions) - Run multiple tasks
- Repeat previously set of ran tasks
- Preserve order of selection
- Customize rerun cache options
By default Node Task List tries to use npm to run the configured script tasks. In case you want to use a custom runner (e.g: yarn or pnpm) you can configure ntl to use whatever runner you prefer.
That can be configured system-wide by setting the environment variable in your profile file (.bashrc or .bash_profile for macos). In case you only want to ever use yarn as your node task runner, you should set:
# .bashrc (linux) OR .bash_profile (macos)
export NTL_RUNNER=yarnKeep in mind environment variables are flexible enough that they can also be set temporarily prior to running a command, so the following would also work:
NTL_RUNNER=yarn ntl❗ You can also define a runner in a per-project basis using the ntl configuration of your package.json, e.g:
{
"name": "<project>",
"version": "1.0.0",
"ntl": {
"runner": "yarn"
}
}📝 You can define descriptions for your tasks in your package.json file by defining a ntl section, e.g:
{
"name": "<project>",
"version": "1.0.0",
"scripts": {
"build": "make build",
"coverage": "jest --coverage",
"test": "jest"
},
"ntl": {
"descriptions": {
"build": "Builds the project",
"coverage": "Run test outputing code coverage",
"test": "Run project's tests"
}
}
}These descriptions will be shown anytime you run ntl.
Use the --info or simply -i option in order to display the contents of each script task, like:
$ ntl -i
⬢ Node Task List
? Select a task to run: (Use arrow keys)
❯ generate-manual › maked-man README.md > man/man1/ntl.1
pretest › eslint cli.js rerun.js test
test › cross-env NTL_NO_RERUN_CACHE=1 tapTask contents will also be shown when using the --descriptions option and no description is available for a given item.
You can define a list of scripts to be excluded from the interactive menu using the --exclude option:
$ ntl -e coverall tasks
⬢ Node Task List
? Select a task to run: (Use arrow keys)
❯ test
test:watch
You can also use a wildcard character to exclude multiple scripts:
$ ntl -e "test*"
⬢ Node Task List
? Select a task to run: (Use arrow keys)
❯ coveralls
tasks
You can also filter out items that doesn't have a description using the --descriptions-only or -o option.
ntl uses a cache system that stores the last ran command for each project in order to make it easier for users to repeat it. Given its importance, the following environment variables are available in order to customize its location and size:
NTL_RERUN_CACHE_DIR: Defines a directory to store the cache fileNTL_RERUN_CACHE_NAME: Filename to use for the cacheNTL_RERUN_CACHE_MAX: Number of items to store in the cache (defaults to10)NTL_NO_RERUN_CACHE: When defined, avoid the cache system completely
For example, if a given user wanted to store its cache in ~/.ntl/cache location and save up to 100 items in it, they could add the following to their .bashrc (linux) or .bash_profile (macos):
export NTL_RERUN_CACHE_DIR=$HOME
export NTL_RERUN_CACHE_NAME=cache
export NTL_RERUN_CACHE_MAX=100The cache can also be customized through command line options:
--rerun-cache-dirDefines a directory to store the cache file--rerun-cache-name: Filename to use for the cache--no-rerun-cache: Avoids the cache system completely
You can increase/reduce the size of the presented UI list using the --size or -s option. In this example we just increased the size of the list to show up to 12 items at once:
$ ntl -s 12
⬢ Node Task List
? Select a task to run:
❯ build
generate-manual
hello
bomdia
bonjour
test
test:dev
test:ci
test:integration
test:unit
test:e2e
start
The default size value is 7 items.
ntl provides many options to make it easier to rerun the last task, either through having it selected as default option the next time you run the ntl command, or by using one of the following:
- The ultra convenient way:
ntcommand shorthand 😎 (You should think ofntas: "ok, just run the last node task", in contrast tontlwhich should be interpreted as: "ok, give me the node task list again") in case no previous task is available, runningntwill behave exactly asntl - Using a
--rerunor-rflag, e.g:ntl -r - Prepending the
NTL_RERUNenv variable, e.g:NTL_RERUN=true ntl
Using the --multiple or -m option, the interface becomes a checkbox-based list that allows you to select multiple tasks and run them in serial.
$ ntl -m
⬢ Node Task List
? Select a task to run:
◯ start
◉ test
❯◉ test:ci
Better yet, combine that with the rerun feature and you can repeat multiple tasks using the nt command.
You can run multiple tasks in the order you pick them 😎 Use the --ordered or -O option:
$ ntl -O
⬢ Node Task List
? Select a task to run: (Press <space> to select, <r> to res
et)
◯ bar › Test task
◯ foo › Test task
❯2 hello › Hello world
◯ start › Start the development server
1 test › Run all tests for the projectUse --autocomplete or -A option in order to use an interface variation that allows you to type the name of the task instead of browsing through an arrow-based menu. This mode can be very helpful when managing a long list of tasks.
$ ntl -A
⬢ Node Task List
? Select a task to run: t
❯ pretest
test
start
You can define ntl as a dev dependency and one of the tasks of your project, specially start - so whenever someone runs npm start or yarn start they get the convenient ntl interface. Like in the following package.json example:
{
"name": "<project>",
"version": "1.0.0",
"scripts": {
"start": "ntl"
},
"devDependencies": {
"ntl": "^5.0.0"
}
}You can also define a task that invokes ntl while excluding other tasks, e.g:
{
"scripts": {
"test": "jest --coverage",
"test:watch": "jest --coverage --watchAll",
"coveralls": "jest --coverage --coverageReporters=text-lcov | coveralls",
"tasks": "ntl --exclude coverall tasks"
}
}ntlThe default commandnodetasklistLonghand version in case users have conflictingntlcommandsnpm-tasklistLegacy longhand versionntRerun last script shortcutnodetaskRerun last script longhand
Still feel like you could use some --help ?
Usage:
ntl [<path>] Build an interactive interface and run any script
nt [<path>] Rerun last executed script
Options:
-a, --all Includes pre and post scripts on the list [boolean]
-A, --autocomplete Starts in autocomplete mode [boolean]
-D, --debug Prints to stderr any internal error [boolean]
-d, --descriptions Displays the descriptions of each script [boolean]
-o, --descriptions-only Limits output to scripts with a description [boolean]
-h, --help Shows this help message [boolean]
-i, --info Displays the contents of each script [boolean]
-e, --exclude Excludes specific scripts [array]
-m, --multiple Allows the selection of multiple items [boolean]
-O, --ordered Selects multiple items in order [boolean]
-s, --size Amount of lines to display at once [number]
--rerun-cache-dir Defines the rerun task cache location [string]
--rerun-cache-name Defines the rerun task cache filename [string]
--no-rerun-cache Never write to or read from cache [boolean]
-v, --version Show version number [boolean]
-r, --rerun Rerun last executed script [boolean]
Visit https://github.com/ruyadorno/ntl for more infoMIT © 2020 Ruy Adorno
