gulp-cause

Declarative gulp tasks composition and watching causality. Better with beverage.

Use

Work in Progress

Some of the functionality described below isn't implemented yet. See the trailing comments in the example about which parts that is.

Here is an example of how this would work:

var gulp = require('gulp');
// create task, two, three, more
require('gulp-cause')(gulp, [
  'alias', 'task',
  'both', {parallel: ['two', 'three']}, // not implemented yet
  'sequence', {series: ['both', 'more']}, // not implemented yet
  'task', ['src/*'], // shorthand, works fine in most cases (where watch is needed)
  'two', [['src/*'], fn], // this array is gulp-watch args
  'cool', {tasks: ['more'], watch: ['place/*']} // not implemented yet
]);

The code above instructs gulp-cause to:

1. make alias of an existing task
2. create task both to run tasks two and three in parallel
3. create task sequence to first run task both (declared above), then task more
4. task:watch will be created - any src/* changes will cause task to be run
5. two:watch uses gulp-watch rather than gulp.watch - documented further down
6. task cool watches place/* files and can have any name, not just more:watch

Rationale

There must be an even number of causality pairs in the array. Having two items per line is good for readability. The left one is always a task name to either be created or invoked - depending on the right-hand-side. The data is processed one pair of items at a time, thus with each step one can depend on tasks from the previous steps. If no tasks existed to begin with, then gulp-cause couldn't do a thing.

Watch

Using gulp.watch by default, unless the watch value is more complex than an array of globs to watch. In such a case, gulp-cause will interpret the array as a gulp-watch arguments list to be applied. See gulp-npm-test for example. Which of the two is preferable depends on the use case.

Test

npm test

Develop

License

MIT