Punchcard CMS

by punchcard-cms

GitHub Readme.md

Punchcard Demo Build Status Coverage Status

Deploy

Quickly deploy Punchcard CMS to Heroku, or use this repository as an example to deploy Punchcard anywhere.

Requirements

Punchcard is a Node.js based headless Content Management System that uses PostgreSQL as its database. It requires the following at a minimum to run:

  • Node.js - 6.0 or greater
  • PostgreSQL - 9.5 or greater

Running Locally

Ensure that PostgreSQL is available on the command line and accessible via the postgres command. Create a user that has a password W@ts0n and ensure that there is a punchcard database created.

If working on a Mac, Homebrew is a convenient way to install PostgreSQL up and running. After installing Homebrew, do the following:

  1. Update Homebrew - brew update
  2. Install PostgreSQL - brew install postgres
  3. Start the PostgreSQL Server - postgres -D /user/local/var/postgres
  4. In a new terminal window or tab, create a new PostgreSQL user and set the password to W@ts0n when prompted - createuser punchcard -P
  5. Create a local database - createdb -0 punchcard punchcard

Local Development Commands

When working locally, there are two things that need to run: the database and the server. There are a number of helper scripts that ship with this repository to help with that:

  • npm run dev:database - Runs PostgreSQL, assuming it's been installed via Homebrew
  • npm run dev:start - Runs Gulp default task, as defined in the Punchcard Runner
  • npm run dev - Runs dev:database and dev:start
  • npm run lint - Runs the Gulp lint task, as defined in the Punchcard Runner
  • npm run build - Runs the Gulp build task, as defined in the Punchcard Runner
  • npm test - Runs lint, then tests, then build

Testing

Canonically, Punchcard uses AVA for testing (both as a runner and as an assertion library) and nyc for code coverage. Due to outstanding issues in later versions, nyc is pegged to v6. JavaScript is linted using ESLint Config Punchcard, using the relevant configurations (node, ava, or browser). Sass is linted using Sass Lint.

The basic test that ships with this repository checks to ensure that Punchcard can run with any modifications an end-user may want to make to the application coming from Punchcard. Custom input plugins are tested using Punchcard Shared Tests.

Punchcard Boilerplate

The following files and folders represent a fairly basic example of Punchcard Boilerplate to get Punchcard up and running. They are divided in to server, browser, test, and config. There are also generated files that can be mostly ignored.

Server Boilerplate

Server boilerplate are required (marked with *) and optional files and folders to get Punchcard running. All files and folders are relative to the running root of the application.

  • index.js* - The main file to actually run. The file presented is the bare minimum file recommended to make Punchcard both run and be testable. Punchcard is a Promise that returns a configured Express application that can be extended as if it were a local Express application.
  • config* - Punchcard config configuration files
    • config/default.js* - Default configuration to be shared regardless of environment. Extends the core Punchcard config to provide all default configuration. Without this configuration extension, Punchcard will not run.
    • config/test.js - Configuration to be used when the Node Environment is test. Recommended to use sqlite3 for testing as it's faster than a full PostgreSQL database. Options present here override the default config and fall back to the default if not present.
    • config/production.js - Configuration to be used the Node Environment is production. Recommend for production environments, with Environment Variables used to hold keys and other secrets. Options present here override the default config and fall back to the default if not present.
  • content-types* - The folder to hold Content Types for use with Punchcard. Punchcard extends the Content Type definition with a top-level attribute workflow that should be set to an ID of a provided workflow (below).
    • content-types/article.yml - A basic content type for an article. Includes file upload, repeatable, and reference attributes.
    • content-types/author.yml - A basic content type for a content author. Includes a file upload attribute, as well as an attribute from a custom input plugin.
  • workflows* -The folder to hold workflows for use with defined content types. Workflows can be set for a given content type by adding workflow: ID in a Content Type, with ID replaced by the ID of the approval workflow.
    • workflows/self-publish.yml - A workflow with one step that allows the individual who authored a piece of content to approve it
  • input-plugins - The folder to hold custom Input Plugins for use in Punchcard. A sample input plugin is provided.
    • input-plugins/utils.js - Helpful functions for creating custom input plugins
    • input-plugins/input-plugin-address - Custom Address input plugin
      • input-plugins/input-plugin-address/README.md - README for Address input plugin
      • input-plugins/input-plugin-address/index.js - Input plugin definition file
      • input-plugins/input-plugin-address/package.json - Basic Node configuration for custom input plugin. Custom input plugin dependencies should be saved to the main Node configuration
  • views - Nunjucks templates to be made available to Punchcard. Putting an identically named template at the identical relative path to a template in the core Punchcard module will overwrite the core template with the provided one
    • views/content/add.html - Template override for the Content Add template from Punchcard. Changes the URL of the provided browser JavaScript to the compiled one for the custom implementation.

Browser Boilerplate

Browser boilerplate is based on the default source and destination targets using Punchcard Runner. These can be modified, and methods for compiling can be changed, as long as settings are updated as appropriate for Punchcard Runner.

  • src - The folder to hold static files that will be served to the browser. Compiled versions will be made available in the public folder.
    • src/images - Images to be available in the browser. Will render to public/images and available in the /images folder from the browser
    • src/js - JavaScript files to be available in the browser. Will render to public/js and available in the /js folder from the browser
    • src/sass - Sass files to be rendered to CSS and available in the browser. Will render to public/css and available in the /css folder from the browser
      • src/sass/style.scss - Sass file that imports the Punchcard styling as an eyeglass module. Will render to public/css/style.css and available at /css/style.css from the browser
  • Gulpfile.js - Gulp task running file. The application, server, and tasks.nodemon configurations are for the server, whereas tasks.build is for the browser files. The two additional tasks, punchcard:js and punchcard:images, bring Punchcard provided JavaScript and Images in to this application's compiled public folder.

Test Boilerplate

Tests, while not required to run Punchcard, are recommended in order to ensure Punchcard runs and is extended as expected.

AVA Tests

  • tests - Folder to run all tests
    • tests/server.js - Basic test using SuperTest to ensure Punchcard is able to start
    • tests/plugin.js - Tests for the custom input plugin. Tests come from Punchcard Shared Tests

Linting

  • .eslintrc.yml - ESLint configuration for all Node code
  • .sass-lint.yml - Sass Lint configuration for all Sass code
  • src/js/.eslintrc.yml - ESLint configuration for all Browser code
  • tests/.eslintrc.yml - ESLint configuration for all Test code

Config Boilerplate

There is a handful of general configuration files. Files that are required are marked with *.

  • .editorconfig - EditorConfig to hint to text editors how to behave
  • .gitignore - Files and file patterns for Git to ignore
  • .nvmrc - Node version for Node Version Manager to use. Helps ensure that the correct version of Node is being used given multiple versions of Node available on a system
  • .slugignore - Files and file patterns to be ignored when Heroku (or Heroku-compatible system) builds the server deployment
  • .travis.yml - Configuration file for Travis Continuous Integration. Configures Travis to use Node 6, cache node_modules, run tests, and on success send code coverage to Coveralls, create a new semantic release, update labels using Reparo, and deploy to Heroku after successful tests are run on the master branch, all while ignoring running any CI on release branches
  • Procfile - Configuration file to tell Heroku (or Heroku-compatible system) what command to run for different contexts. Configures Heroku to run npm start (the start script in package.json) when in a web context
  • app.json - Heroku (or Heroku-compatible system) application manifest
  • package.json* - Node configuration. In a production environment where deploys can happen from a Continuous Integration server, gulp, gulp-concat, gulp-imagemin, gulp-uglify, and punchcard-runner should all be devDependencies and the prestart script should not exist. All dependencies that start with input-plugin-, as well as lodash, are dependencies for the sample content types and the custom input plugin