"New"

- Simple, configurable and shareable... Blueprints.

October 13, 2016

Introduction

We all create boilerplate code. There are tons of tools that help us to do this. Most of them are either designed for a specific project or framework or large scaffolding apps for generating entire projects. "New" or "Blueprints" is something of a hybrid, it allows you to create boilerplate code for a wide array of projects. The boilerplate can be a single file, or it can be a chain of files. One of the great aspects about new is that it is meant to be both simple and flexible.

If you wanted to create a new blueprint called "component" then from where you were when you decided to create the blueprint, simply run:

  new new component  

This creates an empty "new" boilerplate folder, template, and configuration file. The folder (saved in the global "new" templates folder) is called component because that's what we named it using the previous command new new component. The configuration file is named .new-conf, "new" configuration files are always named .new-conf. The template file is named template and this can be changed to any name you want (it can even point to boilerplate files in other locations).

So we now have nestled away somewhere on our computer a folder like:

   ..
     component
       .new-conf
       template

Let's open up the .new-conf file. If you just installed "new" then it should be located at boilerplates/component/.new-conf from the directory you installed in. You could also run new -f to see "folder locations" or to simply edit a config file directly from your favorite set EDITOR, for any "new" boilerplate just running new -e component and that will open up the configuration file for the "component" boilerplate. To edit the current template run new -e component template

template=template
newpath=
after=

If the configuration looks like a bash script, ding ding ding! We told you that "new" is simple. The variable for template points "new" to the file to use as the boilerplate. The variable newpath is useful for working with project. Imagine my components go into the folder named 'app/components', then I wouldn't want to cd into that directory from the root of all my projects, so we could simply set the newpath=app/components.

The after= variable is special, this says that after regenerating this boilerplate to then run another boilerplate. We'll get to that and why it is so powerful later. For now, let's destroy this boilerplate.

# Use new list to see all your installed boilerplates
$ new list
  Found 1 templates:
      component

# Now let's destroy that boilerplate
$ new destroy component

# Run new list again to see that it is gone
$ new list
  Found 0 templates

Alright. So now we know how to make an empty boilerplate and how to destroy it. Let's actually do something useful with "new", we'll create a blueprint which when triggered will generate multiple boilerplate files, in seperate directory trees. More than likely when you create certain types of fi

Will-Change

• $NEW_TEMPLATES_DIR will be a "default" value which shall easily
  be overwritten by any number of configuration files located in
  the users $HOME_DIR or project folders.
• Somehow variables will be available for use inside templates. This feature requires some thought since "new"/blueprints must work for all file types, so it must be careful about how it chooses to identify variables intented to be templated and boilerplate with similar syntax that is just meant to be part of the template. (perhaps the variables will be explict mappings between search and replace values)
• There will be a restore. The backup is already done.