Taskero

Task automation tool with special focus on: asynchronizity, parametrization and automatic watchers functionality.

1. Installation

~$ npm install --save taskero

2. Usage

2.1. Example

This is a simple example. It simulates the compilation of JSX files to JS, and then it writes the concatenation of them into a file.

We can specify both, the termination of the files that will contain the compilation, and the file that will contain the concatenation of the compiled files.

The cool thing is that we can reuse the same function, because we parametrize some of the variables of the task, and when we run them, we redefine them.

const Taskero = require("taskero").Taskero;
const taskero = new Taskero({debug:false});
const compileJSX = function(i) {
 // Here, you can modify the contents of the file
 return i
};
const fs = require("fs");

taskero.register({
 name: "jsx:compile",
 onEach: [
   function(done, file, args) {
     done(compileJSX(file.contents));
   },
   function(done, file, args) {
     fs.writeFileSync(file.path + args.filesAppendix, file.contents, "utf8")
     done();
   }
 ],
 onDone: [
   function(done, files, args) {
     console.log("[taskero:jsx:compile] Compilation finished for: " + JSON.stringify(files, null, 3));
     done(files.reduce(function(prev, curr) {
       return prev + "\n" + curr;
     }, ""));
   }
 ],
 files: "**/**.jsx"
});

taskero.register({
 name: "say:ok",
 onDone: function() {console.log("OK!");}
});

taskero.run([
 {
   name: "jsx:compile",
   filesAppendix: ".compiled.js",
   onDoneFile: "dist/compilation.js"
 },{
   name: "say:ok"
}]).then(function() {
 console.log("Task finished.");
}).catch(function() {
 console.log("There were errors", error);
});

2.2. Step-by-step explanation

The following is a step-by-step explanation of the previous example.

1. Import Taskero class and create an instance:

const Taskero = require("taskero").Taskero;
const taskero = new Taskero({debug:false});

The object passed into the constructor sets the options for the instance created. The only option, for now, is debug. If set to true, it will log by console all the actions that taskero is carrying on under the hood. By default, this option is set to false, so this option is redundant, and the object itself is optional too.

2. Register the tasks we want.

taskero.register({
 name: "name of the task", // Must be unique per {Taskero} instance
 onEach: [],
 onDone: [],
 onDoneFile: "",
 files: [],
 filesEncoding: "utf8",
 watch: false,
 watchOptions: {persist:true}
});

The object passed to the Taskero#register(~) method here defines the default options known by the API. Moreover, you can provide your own properties. When you provide your own properties here, they will be set as default for the current task. They, though, will be extendable by the Taskero#run(~) method.

3. Run the tasks we want, overriding or adding the parameters we want.

taskero.run([{
 name: "task:1",
}, {
 name: "task:2",
}, {
 name: "task:3",
}, {
 name: "name of the already registered task",
 files: [/*Glob patterns*/],
 watch: true, // this will run the task as watcher
 onDoneFile: "dist/build.js",
 customArgument1: [],
 customArgument2: [],
 customArgument3: [],
 customArgument4: [],
 customArgument5: [],
 customArgument6: []
}]);

The Taskero#run(~) method accepts different types of arguments. The shown below is the most flexible, as it accepts multiple tasks (that will be run one after the other) and, at the same time, their parametrization. This way, we can reuse the same task against different files.

Remember that the watch option will let you run the task as a watcher, instead of as a simple execution, just providing a true flag, which can result very handy in order to reuse a task in watcher mode.

Note: the watcher mode means that the files matched will be listened for changes, and so, you can forget about running the task: it will be run automatically for you everytime a file is changed, or created, or removed.

3. API Reference

require("taskero").Taskero

Type: {Class}.

Description: Returns the main class of the Taskero's API.

new Taskero(Object:options)

Type: {Class}.

Property: {Array<Object>} tasksMap.

Property: {Array<Object>} watchers.

Property: {Object} options. By default: {debug:false}.

Parameter: {Object} options. Optional. Configurations of the new instance. By default:

{debug: false}

Return: {Object}.

Description: Creates and returns a new Taskero instance.

Taskero#tasksMap

Type: {Object}.

Description: Information of all the tasks (identified by their name) that have been registered by this Taskero instance.

Taskero#options

Type: {Object}

Description: Options passed to this Taskero instance. This is built by the default options, and the object provided as first argument to the Taskero constructor. Default options:

{
 debug: true
}

Taskero#watchers

Type: {Array<Object>}

Description: File watchers added by this Taskero instance.

Taskero#register(Object:taskInfo)

Type: {Function}

Description: Registers a new task for this instance.

Parameter: {Object} taskInfo. Required.

The Taskero#register(~) method accepts an object as parameter. This object can have any of the following properties:

• name: {String}. Required. The name of the task. It can not be duplicated. Otherwise, an error will be thrown.

• onEach: {Array<Function> | Function}. Optional. Functions that compose this task, that will be applied to each file (matched by the files argument). This implies that each of the functions provided by this argument will be applied once per each matched file. By default: [].

These functions (or function) will receive 3 parameters:

· argument 1: done: {Function}. This must be called to finish the callback, and to start the next call. When called with 1 parameter, the {String} passed to it will change the content of the file (but not written). When called with 2 parameters, the second parameter will throw an error, catched by the Taskero#run(...).then(...).catch(~) callback.

· argument 2: file: {Object}. This object has a .path and a .contents property. This object will be passed through all the onEach callbacks. The .contents property can be modified directly, or by the done(~) call. Take into account that any matched file will appear here.

· argument 3: args: {Object}. This object has all the parameters received by this task. So: name, onEach, onDone, onDoneFile, files... and any other parameter added by the method Taskero#register(~) or the method Taskero#run(~) will also appear here.

• onDone: {Array<Function> | Function}. Optional. Functions that compose this task, that will be applied after all the onEach callbacks are solved. The arguments that it receives are the same as the onEach callbacks, but as this set of callbacks are going to be called only once (and not once per file, as the onEach callbacks), the second arguments is not file, but files. If their contents were modified by onEach callbacks, they will remain modified here too. This callback is useful to concatenate the contents of all the files somewhere, or for tasks that are executed once, independently from the files. Also, note that when the done call is provided with 1 parameter, it will dump the passed {String} into the file specified at the parameter onDoneFile, if any. When it is called with 2, it will throw an error, catched by the Taskero#run(...).then(...).catch(~) callback. By default: [].

• onDoneFile: {String}. Optional. File that will be written with the {String} passed to the done (the first argument) of the onDone parameter. The file and folders, if not already created, will be created. If the said done call is not provided with 1 parameter, it will not dump anything to any file. By default: undefined.

• files: {Array<String> | String}. Optional. Glob pattern(s) that will be matched before proceeding to the onEach and onDone callbacks execution. It can be empty, in which case, no onEach callback will be called (but the onDone will be called anyway). By default: [].

• filesEncoding: {String}. Optional. By default: "utf8".

• watch: {Boolean}. Optional. When set to true, the task will be understood as a watcher, not as an executor. This means that the glob patterns will be used to be watched (by chokidar library), instead of matched (by globby library. When set to true, the watchers generated will be added into Taskero#watchers, which starts as an empty array. And they can be all closed and removed by calling to Taskero#closeWatchers(), directly. By default: false.

• watchOptions: {Object}. Optional. This object is passed to the chokidar library as the watch options. By default: {persist: true}.

Taskero#run(String:taskName)

Taskero#run(Object:taskParameters)

Taskero#run(Array<Object:taskParameters>:tasks)

Type: {void}

Parameter: {Array<Object>} tasks. Set of tasks to be run. They will be executed one after the other. The object will override the one passed to the task registration.

Returns: {Promise}. To chain the then(~) and catch(~) calls.

Description: Runs the provided tasks, in series (one after the other, in a synchronous way).

Taskero#closeWatchers()

Type: {Function}.

Parameters: {none}.

Returns: {void}.

Description: Removes all the Taskero#watchers added to the instance.

Taskero#debug(Any:message,...)

Type: {Function}.

Parameters: {none}.

Returns: {void}.

Description: Logs the provided messages, only if the Taskero#options.debug is set to true.

4. Tests, coverage and documentation

The package.json file is provided with 3 commands:

• npm run docs: it generates the documentation from the Javadoc comments of the src/taskero.js file, which contains the whole library.

• npm run test: it runs the tests and generates the coverage reports into coverage/ folder.

• npm run test-nocov: it runs the tests, but it does not generate coverage reports. Use this command if you want to run tests fastly.

5. Why?

I felt that Grunt is cool, and Gulp is cool, and Webpack is cool.

Yes, they are cool.

But:

• Webpack, which was the coolest, complicated things too much.

• Grunt, which was my favourite, lacks of parametrization.

• Gulp, which was very promising, has a strange API.

6. Problems

I am not an engineer. I just make tools that fit my needs. But I do not have a deep understanding, or know the best algorythms for doing what I want to do. I can imagine better ways to do them. But I still do not care too much about this point. It may result dangerous, when I do not have the deep knowledge I should have.

In other words: this tool seems to work great for some tasks. But I do not know what can happen if I put a hundred tasks to run, a thousand files to watch, etc.

But... I supose I can split my tasks into different files, and run them selectively, depending on which part of the project I am working on. I do not know. I want to think that I am not losing my time working on tools like this one. In the end, I am already using my own tools for my developments.

7. Conclusions

In conclusion, Taskero is a project that aims to ease the tasks automation, which is a very fundamental part of any software development project, and which takes a lot of time in development.

Also, I wanted to give more open source projects, I can not compite against companies, or engineers, but... maybe I can provide something interesting to them, despite not having a job yet. I do not know.