Month: December 2018

Styling a Select Like It’s 2019

, , , ,

A Quick CSS Audit and General Notes About Design Systems

I’ve been auditing a ton of CSS lately and thought it would be neat to jot down how I’m going about doing that. I’m sure there are a million different ways to do this depending on the size and scale of your app and how your CSS works under the hood, so please take all this with a grain of salt.

First a few disclaimers: at Gusto, the company I work for today, our engineers and designers all write in Sass and use webpack to compile those files into CSS. Our production environment minifies all that code into a single CSS file. However, our CSS is made up of three separate domains. so I downloaded them all to my desktop because I wanted to test them individually.

Here’s are those files and what they do:

  • manifest.css: a file that’s generated from all our Sass functions, mixins and contains all of our default HTML styles and utility classes.
  • components.css: a file that consists of our React components such as Button.scss, Card.scss, etc. This and manifest.css both come from our Component Library repo and are imported into our main app.
  • app.css: a collection of styles that override our components and manifest. Today, it exists in our main application repo.

After I downloaded everything, I threw them into an S3 bucket and ran them through CSS Stats. (I couldn’t find a command line tool that I liked, so I decided stuck with this tool.) The coolest thing about CSS Stats is that it provides a ton of clarity about the health and quality of a site’s CSS, and in turn, a design system. It does this by showing the number of unique font-size and unique background-color CSS declarations there are, as well as a specificity graph for that particular CSS file.

I wanted to better understand our manifest.css file first. As I mentioned, this file contains all our utility classes (such as padding-top-10px and c-salt-500) as well as our normalize and reset CSS files, so it’s pretty foundational for everything else. I started digging through the results:

There are some obvious issues here, like the fact that there are 101 unique colors and 115 unique background colors. Why is this a big deal? Well, it’s a little striking to me because our team had already made a collection of Sass functions to output a very specific number of colors. In our Figma UI Kit and variables_color.scss (which gets compiled into our manifest file, we declare a total of 68 unique colors:

So, where are all these extra colors coming from? Well, I assume that they’re coming from Bootstrap. Back when we started building the application, we hastily built on top of Bootstrap’s styles without refactoring things as we went. There was a certain moment when this started to hurt as we found visual inconsistencies across our application and hundreds of lines of code being written that simply overrode Bootstrap. In a rather gallant CSS refactor, I removed Bootstrap’s CSS from our main application and archived it inside manifest.css, waiting for the day when we could return to it and refactor it all.

These extra colors are likely come from that old Bootstrap file, but it’s probably worth investigating some more. Anyway, the real issue with this for me is that my understanding of the design system is different from what’s in the front-end. That’s a big problem! If my understanding of the design system is different from how the CSS works, then there’s enormous potential for engineers and designers to pick up on the wrong patterns and for confusion to disseminate across our organization. Think about the extra bloat and lack of maintainability, not to mention other implications.

I was reading Who Are Design Systems For? by Matthew Ström and perked up when he quotes a talk by Julie Ann-Horvath where she’s noted as saying, “a design system doesn’t exist until it’s in production.” Following the logic, it’s clear the design system I thought we had didn’t actually exist.

Going back to manifest.css though: the specificity graph for this file should be perfectly gradual and yet there are some clear spikes that show there’s probably a bit more CSS that needs to be refactored in there:

Anyway, next up is our components.css. Remember that’s the file that our styles for our components come from so I thought beforehand that it’s bound to be a little messier than our manifest file. Throwing it into CSS Stats returns the following:

CSS-Stats shows some of the same problems — like too many font sizes (what the heck is going on with that giant font size anyway?) — but there are also way too many custom colors and background-colors. I already had a hunch about what the biggest issue with this CSS file was before I started and I don’t think the problem is not shown in this data here at all.

Let me explain.

A large number of our components used to be Bootstrap files of one kind or another. Take our Accordion.jsx React component, for instance. That imports an accordion.css file which is then compiled with all the other component’s CSS into a components.css file. The problem with this is that some Accordion styles affect a lot more than just that component. CSS from this this file bleeds into other patterns and classes that aren’t tied to just one component. It’s sort of like a poison in our system and that impacts our team because it makes it difficult to reliably make changes to a single component. It also leads to a very fragile codebase.

So I guess what I’m saying here is that tools like CSS Stats are wondrous things to help us check core vital signs for CSS health, but I don’t think they’ll ever really capture the full picture.

Anyway, next up is the app.css file:

This is the “monolith” — the codebase that our design systems team is currently trying to better understand and hopefully refactor into a series of flexible and maintainable React components that others can reuse again and again.

What worries me about this codebase is the specificity of it all what happens when something changes in the manifest.css or in our components.css? Will those styles be overridden in the monolith? What will happen to the nice and tidy component styles that we import into a new project?

Subsequently, I don’t know where I stole this, but I’ve been saying it an awful lot lately — you should always be able to predict what your CSS is going to do, whether that’s a single line of code or a giant codebase of intermingled styles. That’s what design systems are all about — designing and building predictable interfaces for the future. And if our compiled CSS has all these unpredictable and unknowable parts to it, then we need to gather everyone together to fix it.

Anywho, I threw some of the data into a Dropbox Paper doc after all this to make sure we start tackling these issues and see gradual improvements over time. That looks something like this today:

How have you gone about auditing your CSS? Does your team code review CSS? Are there any tricks and tips you’d recommend? Leave a comment below!

The post A Quick CSS Audit and General Notes About Design Systems appeared first on CSS-Tricks.


, , , , , ,

Gradient Borders in CSS

Let’s say you need a gradient border around an element. My mind goes like this:

  • There is no simple obvious CSS API for this.
  • I’ll just make a wrapper element with a linear-gradient background, then an inner element will block out most of that background, except a thin line of padding around it.

Perhaps like this:

See the Pen Gradient with Wrapper by Chris Coyier (@chriscoyier) on CodePen.

If you hate the idea of a wrapping element, you could use a pseudo-element, as long as a negative z-index value is OK (it wouldn’t be if there was much nesting going on with parent elements with their own backgrounds).

Here’s a Stephen Shaw example of that, tackling border-radius in the process:

See the Pen Gradient border + border-radius by Shaw (@shshaw) on CodePen.

You could even place individual sides as skinny pseudo-element rectangles if you didn’t need all four sides.

But don’t totally forget about border-image, perhaps the most obtuse CSS property of all time. You can use it to get gradient borders even on individual sides:

See the Pen Gradient Border on 2 sides with border-image by Chris Coyier (@chriscoyier) on CodePen.

Using both border-image and border-image-slice is probably the easiest possible syntax for a gradient border, it’s just incompatible with border-radius, unfortunately.

See the Pen CSS Gradient Borders by Chris Coyier (@chriscoyier) on CodePen.

The post Gradient Borders in CSS appeared first on CSS-Tricks.



Gulp for WordPress: Creating the Tasks

This is the second post in a two-part series about creating a Gulp workflow for WordPress theme development. Part one focused on the initial installation, setup, and organization of Gulp in a WordPress theme project. This post goes deep into the tasks Gulp will run by breaking down what each task does and how to tailor them to streamline theme development.

Now that we spent the first part of this series setting up a WordPress theme project with Grunt installed in it, it’s time to dive into the tasks we want it to do for us as we develop the theme. We’re going to get our hands extremely dirty in this post, get ready to write some code!

Article Series:

  1. Initial Setup
  2. Creating the Tasks (This Post)

Creating the style task

Let’s start by compiling src/bundle.scss from Sass to CSS, then minifying the CSS output for production mode and putting the completed bundle.css file into the dist directory.

We’re going to use a couple of Gulp plugins to do the heavy lifting. We’ll use gulp-sass to compile things and gulp-clean-css to minify. Then, gulp-if will allow us to conditionally run functions which, In our case, will check if we are in production or development modes before those tasks run and then execute accordingly.

We can install all three plugins in one fell swoop:

npm install --save-dev gulp-sass gulp-clean-css gulp-if

Let’s make sure we have something in our bundle.scss file so we can test the tasks:

$ colour: #f03;  body {   background-color: $ colour; }

Alright, back to the Gulpfile to import the plugins and define the task that runs them:

import { src, dest } from 'gulp'; import yargs from 'yargs'; import sass from 'gulp-sass'; import cleanCss from 'gulp-clean-css'; import gulpif from 'gulp-if'; const PRODUCTION =;  export const styles = () => {   return src('src/scss/bundle.scss')     .pipe(sass().on('error', sass.logError))     .pipe(gulpif(PRODUCTION, cleanCss({compatibility:'ie8'})))     .pipe(dest('dist/css')); }

Let’s walk through this code to explain what’s happening.

  • The src and dest functions are imported from Gulp. src will read the file that you pass as an argument and return a node stream.
  • We pull in yargs to create our flag that separates tasks between the development and production modes.
  • The three plugins are called into action.
  • The PRODUCTION flag is defined and held in the prod command.
  • We define styles as the task name we will use to run these tasks in the command line.
  • We tell the task what file we want processed (bundle.scss) and where it lives (src/scss/bundle.scss).
  • We create “pipes” that serve as the plungs that run when the styles command is executed. Those pipes run in the order they are written: convert Sass to CSS, minify the CSS (if we’re in production), and place the resulting CSS file into the dist/css directory.

Go ahead. Run gulp styles in the command line and see that a new CSS file has been added to your CSS directory dist/css.

Now do gulp styles --prod. The same thing happens, but now that CSS file has been minified for production use.

Now, assuming you have a functioning WordPress theme with header.php and footer.php, the CSS file (as well as JavaScript files when we get to those tasks) can be safely enqueued, likely in your functions.php file:

function _themename_assets() {   wp_enqueue_style( '_themename-stylesheet', get_template_directory_uri() . '/dist/css/bundle.css', array(), '1.0.0', 'all' ); } add_action('wp_enqueue_scripts', '_themename_assets');

That’s all good, but we can make our style command even better.

For example, try inspecting the body on the homepage with the WordPress theme active. The styles that we added should be there:

As you can see, it says that our style is coming from bundle.css, which is true. However, it would be much better if the name of the original SCSS file is displayed here instead for our development purposes — it makes it so much easier to locate code, particularly when we’re working with a ton of partials. This is where source maps come into play. That will detail the location of our styles in DevTools. To further illustrate this issue, let’s also add some SCSS inside src/scss/components/slider.scss and then import this file in bundle.scss.

//src/scss/components/slider.scss body {   background-color: aqua; }
//src/scss/bundle.scss @import './components/slider.scss'; $ colour: #f03; body {   background-color: $ colour; }

Run gulp styles again to recompile your files. Your inspector should then look like this:

The DevTools inspector will show that both styles are coming from bundle.css. But we would like it to show the original file instead (i.e bundle.scss and slider.scss). So let’s add that to our wish list of improvements before we get to the code.

The other thing we’ll want is vendor prefixing to be handled for us. There’s nothing worse than having to write and manage all of those on our own, and Autoprefixer is the tool that can do it for us.

And, in order for Autoprefixer to work its magic, we’ll need the PostCSS plugin.

OK, that adds up to three more plugins and tasks we need to run. Let’s install all three:

npm install --save-dev gulp-sourcemaps gulp-postcss autoprefixer

So gulp-sourcemaps will obviously be used for sourcemaps. gulp-postcss and autoprefixer will be used to add autoprefixing to our CSS. postcss is a famous plugin for transforming CSS files and autoprefixer is just a plugin for postcss. You can read more about the other things that you can do with postcss here.

Now at the very top let’s import our plugins into the Gulpfile:

import postcss from 'gulp-postcss'; import sourcemaps from 'gulp-sourcemaps'; import autoprefixer from 'autoprefixer';

And then let’s update the task to use these plugins:

export const styles = () => {   return src('src/scss/bundle.scss')     .pipe(gulpif(!PRODUCTION, sourcemaps.init()))     .pipe(sass().on('error', sass.logError))     .pipe(gulpif(PRODUCTION, postcss([ autoprefixer ])))     .pipe(gulpif(PRODUCTION, cleanCss({compatibility:'ie8'})))     .pipe(gulpif(!PRODUCTION, sourcemaps.write()))     .pipe(dest('dist/css')); }

To use the the sourcemaps plugin we have to follow some steps:

  1. First, we initialize the plugin using sourcemaps.init().
  2. Next, pipe all the plugins that you would like to map.
  3. Finally, Create the source map file by calling sourcemaps.write() just before writing the bundle to the destination.

Note that all the plugins piped between sourcemaps.init() and sourcemaps.write() should be compatible with gulp-sourcemaps. In our case, we are using sass(), postcss() and cleanCss() and all of them are compatible with sourcemaps.

Notice that we only run the Autoprefixer begind the production flag since there’s really no need for all those vendor prefixes during development.

Let’s run gulp styles now, without the production flag. Here’s the output in bundle.css:

body {   background-color: aqua; } body {   background-color: #f03; } /*#sourceMappingURL=data:application/json;charset=utf8;base64,eyJ2ZXJzaW9uIjozLCJmaWxlIjoiYnVuZGxlLmNzcyIsInNvdXJjZXMiOlsiYnVuZGxlLnNjc3MiLCJjb21wb25lbnRzL3NsaWRlci5zY3NzIl0sInNvdXJjZXNDb250ZW50IjpbIkBpbXBvcnQgJy4vY29tcG9uZW50cy9zbGlkZXIuc2Nzcyc7XG5cbiRjb2xvdXI6ICNmMDM7XG5ib2R5IHtcbiAgICBiYWNrZ3JvdW5kLWNvbG9yOiAkY29sb3VyO1xufVxuOjpwbGFjZWhvbGRlciB7XG4gICAgY29sb3I6IGdyYXk7XG59IiwiYm9keSB7XG4gICAgYmFja2dyb3VuZC1jb2xvcjogYXF1YTtcbn0iXSwibmFtZXMiOltdLCJtYXBwaW5ncyI6IkFDQUEsQUFBQSxJQUFJLENBQUM7RUFDRCxnQkFBZ0IsRUFBRSxJQUFJLEdBQ3pCOztBRENELEFBQUEsSUFBSSxDQUFDO0VBQ0QsZ0JBQWdCLEVBRlgsSUFBSSxHQUdaOztBQUNELEFBQUEsYUFBYSxDQUFDO0VBQ1YsS0FBSyxFQUFFLElBQUksR0FDZCJ9 */#

The extra text below is source maps. Now, when we inspect the site in DevTools, we see:

Nice! Now onto production mode:

gulp styles --prod

Check DevTools against style rules that require prefixing (e.g. display: grid;) and confirm those are all there. And make sure that your file is minified as well.

One final notice for this task. Let’s assume we want multiple CSS bundles: one for front-end styles and one for WordPress admin styles. We can create add a new admin.scss file in the src/scss directory and pass an array of paths in the Gulpfile:

export const styles = () => {   return src(['src/scss/bundle.scss', 'src/scss/admin.scss'])     .pipe(gulpif(!PRODUCTION, sourcemaps.init()))     .pipe(sass().on('error', sass.logError))     .pipe(gulpif(PRODUCTION, postcss([ autoprefixer ])))     .pipe(gulpif(PRODUCTION, cleanCss({compatibility:'ie8'})))     .pipe(gulpif(!PRODUCTION, sourcemaps.write()))     .pipe(dest('dist/css')); }

Now we have bundle.css and admin.css in the dist/css directory. Just make sure to properly enqueue any new bundles that are separated out like this.

Creating the watch task

Alright, next up is the watch task, which makes our life so much easier by looking for files with saved changes, then executing tasks on our behalf without have to call them ourselves in the command line. How great is that?

Like we did for the styles task:

import { src, dest, watch } from 'gulp';

We’ll call the new task watchForChanges:

export const watchForChanges = () => {   watch('src/scss/**/*.scss', styles); }

Note that watch is unavailable as a name since we already have a variable using it.

Now let’s run gulp watchForChanges the command line will be on a constant, ongoing watch for changes in any .scss files inside the src/scss directory. And, when those changes happen, the styles task will run right away with no further action on our part.

Note that src/scss/**/*.scss is a glob pattern. That basically means that this string will match any .scss file inside the src/scss directory or any sub-folder in it. Right now, we are only watching for .scss files and running the styles task. Later, we’ll expand its scope to watch for other files as well.

Creating the images task

As we covered earlier, the images task will compress images in src/images and then move them to dist/images. Let’s install a gulp plugin that will be responsible for compressing images:

npm install --save-dev gulp-imagemin

Now, import this plugin at the top of the Gulpfile:

import imagemin from 'gulp-imagemin';

And finally, let’s write our images task:

export const images = () => {   return src('src/images/**/*.{jpg,jpeg,png,svg,gif}')     .pipe(gulpif(PRODUCTION, imagemin()))     .pipe(dest('dist/images')); }

We give the src() function a glob that matches all .jpg, .jpeg, .png, .svg and .gif images in the src/images directory. Then, we run the imagemin plugin, but only for production. Compressing images can take some time and isn’t necessary during development, so we can leave it out of the development flow. Finally, we put the compressed versions of images in dist/images.

Now any images that we drop into src/images will be copied when we run gulp images. However, running gulp images --prod, will both compress and copy the image over.

Last thing we need to do is modify our watchForChanges task to include images in its watch:

export const watchForChanges = () => {   watch('src/scss/**/*.scss', styles);   watch('src/images/**/*.{jpg,jpeg,png,svg,gif}', images); }

Now, assuming the watchForChanges task is running, the images task will be run automatically whenever we add an image to the src/images folder. It does all the lifting for us!

Important: If the watchForChanges task is running and when the Gulpfile is modified, it will need to be stopped and restarted in order for the changes to take effect.

Creating the copy task

You probably have been in situations where you’ve created files, processed them, then needed to manually grab the production files and put them where they need to be. Well, as we saw in the images task, we can use the copy feature to do this for us and help prevent moving wrong files.

export const copy = () => {   return src(['src/**/*','!src/{images,js,scss}','!src/{images,js,scss}/**/*'])     .pipe(dest('dist')); }

Try to read the array of paths supplied to src() carefully. We are telling Gulp to match all files and folders inside src (src/**/*), except the images, js and scss folders (!src/{images,js,scss}) and any of the files or sub-folders inside them (!src/{images,js,scss}/**/*).

We want our watch task to look for these changes as well, so we’ll add it to the mix:

export const watchForChanges = () => {   watch('src/scss/**/*.scss', styles);   watch('src/images/**/*.{jpg,jpeg,png,svg,gif}', images);   watch(['src/**/*','!src/{images,js,scss}','!src/{images,js,scss}/**/*'], copy); }

Try adding any file or folder to the src directory and it should be copied over to the the /dist directory. If, however, we were to add a file or folder inside of /images, /js or /scss, it would be ignored since we already handle these folders in separate tasks.

We still have a problem here though. Try to delete the added file and it won’t happen. Our task only handles copying. This problem could also happen for our /images, /js and /scss, folders. If we have old images or JavaScript and CSS bundles that were removed from the src folder, then they won’t get removed from the dist folder. Therefore, it’s a good idea to completely clean the dist folder every time to start developing or building a theme. And that’s what we are going to do in the next task.

Composing tasks for developing and building

Let’s now install a package that will be responsible for deleting the dist folder. This package is called del:

npm install --save-dev del

Import it at the top:

import del from 'del';

Create a task that will delete the dist folder:

export const clean = () => {   return del(['dist']); }

Notice that del returns a promise. Thus, we don’t have to call the cb() function. Using the new JavaScript features allows us to refactor this to:

export const clean = () => del(['dist']);

The folder should be deleted now when running gulp clean. What we need to do next is delete the dist folder, run the images, copy and styles tasks, and finally watch for changes every time we start developing. This can be done by running gulp clean, gulp images, gulp styles, gulp copy and then gulp watch. But, of course, we will not do that manually. Gulp has a couple of functions that will help us compose tasks. So, let’s import these functions from Gulp:

import { src, dest, watch, series, parallel } from 'gulp';

series() will take some tasks as arguments and run them in series (one after another). And parallel() will take tasks as arguments and run them all at once. Let’s create two new tasks by composing the tasks that we already created:

export const dev = series(clean, parallel(styles, images, copy), watchForChanges) export const build = series(clean, parallel(styles, images, copy)) export default dev;

Both tasks will do the exact same thing: clean the dist folder, then styles, images and copy will run in parallel one the cleaning is complete. We will start watching for changes as well for the dev (short for develop) task, after these parallel tasks. Additionally, we are also exporting dev as the default task.

Notice that when we run the build task, we want our files to be minified, images to be compressed, and so on. So, when we run this command, we will have to add the --prod flag. Since this can easily be forgotten when running the build task, we can use npm scripts to create aliases for the dev and build commands. Let’s go to package.json, and in the scripts field, we will probably find something like this:

"scripts": {   "test": "echo "Error: no test specified" && exit 1" }

Let’s change it to this:

"scripts": {   "start": "gulp",   "build": "gulp build --prod" },

This will allow us to run npm run start in the command line, which will go to the scripts field and find what command corresponds to start. In our case, start will run gulp and gulp will run the default gulp task, which is dev. Similarly, npm run build will run gulp build --prod. This way, we can completely forget about the --prod flag and also forget about running the Gulp tasks using the gulp command. Of course, our dev and build commands will do more than that later on, but for now, we have the foundation that we will work with throughout the rest of the tasks.

Creating the scripts task

As mentioned, in order to bundle our JavaScript files, we are going to need a module bundler. webpack is the most famous option out there, however it is not a Gulp plugin. Rather, it’s a plugin on its own that has a completely separate setup and configuration file. Luckily, there is a package called webpack-stream that helps us use webpack within a Gulp task. So, let’s install this package:

npm install --save-dev webpack-stream

webpack works with something called loaders. Loaders are responsible for transforming files in webpack. And to transform new Javascript versions into ES5, we will need a loader called babel-loader. We will also need @babel/preset-env but we already installed this earlier:

npm install --save-dev babel-loader

Let’s import webpack-stream at the top of the Gulpfile:

import webpack from 'webpack-stream';

Also, to test our task, lets add these lines in src/js/bundle.js and src/js/components/slider.js:

// bundle.js import './components/slider'; console.log('bundle');   // slider.js console.log('slider')

Our scripts task will finally look like so:

export const scripts = () => {   return src('src/js/bundle.js')   .pipe(webpack({     module: {       rules: [         {           test: /.js$ /,           use: {             loader: 'babel-loader',             options: {               presets: []             }           }         }       ]     },     mode: PRODUCTION ? 'production' : 'development',     devtool: !PRODUCTION ? 'inline-source-map' : false,     output: {       filename: 'bundle.js'     },   }))   .pipe(dest('dist/js')); }

Let’s break this down a bit:

  • First, we specify bundle.js as our entry point in the src() function.
  • Then, we pipe the webpack plugin and specify some options for it.
  • The rules field in the module option lets webpack know what loaders to use in order to transform our files. In our case we need to transform JavaScript files using the babel-loader.
  • The mode option is either production or development. For development, webpack will not minify the output JavaScript bundle, but it will for production. Therefore, we don’t need a separate Gulp plugin to minify JavaScript because webpack can do that depending on our PRODUCTION constant.
  • The devtool option will add source maps, but not in production. In development, however, we will use inline-source-maps. This kind of source maps is the most accurate though it can be a bit slow to create. If you find it too slow, check the other options here. They won’t be as accurate as inline-source-maps but they can be pretty fast.
  • Finally, the output option can specify some information about the output file. In our case, we only need to change the filename. If we don’t specify the filename, webpack will generate a bundle with a hash as the filename. Read more about these options here.

Now we should be able to run gulp scripts and gulp scripts --prod and see a bundle.js file created in dist/js. Make sure that minification and source maps are working properly. Let’s now enqueue our JavaScript file in WordPress, which can be in the theme’s functions.php file, or wherever you write your functions.

<?php function _themename_assets() {   wp_enqueue_style( '_themename-stylesheet', get_template_directory_uri() . '/dist/css/bundle.css', array(), '1.0.0', 'all' );      wp_enqueue_script( '_themename-scripts', get_template_directory_uri() . '/dist/assets/js/bundle.js', array(), '1.0.0', true ); } add_action('wp_enqueue_scripts', '_themename_assets');

Now, looking at the console, let’s confirm that source maps are working correctly by checking the file that the console logs come from:

Without the source maps, both logs will appear coming from bundle.js.

What if we would like to create multiple JavaScript bundles the same way we do for the styles? Let’s create a file called admin.js in src/js. You might think that we can simply change the entry point in the src() to an array like so:

export const scripts = () => {   return src(['src/js/bundle.js','src/js/admin.js'])   .   . }

However, this will not work. webpack works a bit differently that normal Gulp plugins. What we did above will still create one file called bundle.js in the dist folder. webpack-stream provides a couple of solutions for creating multiple entry points. I chose to use the second solution since it will allow us to create multiple bundles by passing an array to the src() the same way we did for the styles. This will require us to install vinyl-named:

npm install --save-dev vinyl-named

Import it:

import named from 'vinyl-named';

…and then update the scripts task:

export const scripts = () => {   return src(['src/js/bundle.js','src/js/admin.js'])   .pipe(named())   .pipe(webpack({     module: {       rules: [         {           test: /.js$ /,           use: {             loader: 'babel-loader',             options: {               presets: ['@babel/preset-env']             }           }         }       ]     },     mode: PRODUCTION ? 'production' : 'development',     devtool: !PRODUCTION ? 'inline-source-map' : false,     output: {       filename: '[name].js'     },   }))   .pipe(dest('dist/js')); }

The only difference is that we now have an array in the src(). We then pipe the named plugin before webpack, which allows us to use a [name] placeholder in the output field’s filename instead of hardcoding the file name directly. After running the task, we get two bundles in dist/js.

Another feature that webpack provides is using libraries from external sources rather than bundling them into the final bundle. For example, let’s say your bundle needs to use jQuery. You can run npm install jquery --save and then import it to your bundle import $ from 'jquery'. However, this will increase the bundle size and, in some cases, you may already have jQuery loaded via a CDN or — in case of WordPress — it can exist as a dependency like so:

wp_enqueue_script( '_themename-scripts', get_template_directory_uri() . '/dist/assets/js/bundle.js', array('jquery'), '1.0.0', true );

So, now WordPress will enqueue jQuery using a normal script tag. How can we then use it inside our bundle using import $ from 'jquery'? The answer is by using webpack’s externals option. Let’s modify our scripts task to add it in:

export const scripts = () => {   return src(['src/js/bundle.js','src/js/admin.js'])     .pipe(named())     .pipe(webpack({       module: {         rules: [           {             test: /.js$ /,             use: {               loader: 'babel-loader',               options: {                 presets: []             }           }         }       ]     },     mode: PRODUCTION ? 'production' : 'development',     devtool: !PRODUCTION ? 'inline-source-map' : false,     output: {       filename: '[name].js'     },     externals: {       jquery: 'jQuery'     },   }))   .pipe(dest('dist/js')); }

In the externals option, jquery is the key that identifies the name of the library we want to import. In our case, it will be import $ from 'jquery'. And the value jQuery is the name of a global variable where that the library lives. Now try to import $ from ‘jquery’ in the bundle and use jQuery using the $ — it should work perfectly.

Let’s watch for changes for JavaScript files as well:

export const watchForChanges = () => {   watch('src/scss/**/*.scss', styles);   watch('src/images/**/*.{jpg,jpeg,png,svg,gif}', images);   watch(['src/**/*','!src/{images,js,scss}','!src/{images,js,scss}/**/*'], copy);   watch('src/js/**/*.js', scripts); }

And, finally, add our scripts task in the dev and build tasks:

export const dev = series(clean, parallel(styles, images, copy, scripts), watchForChanges); export const build = series(clean, parallel(styles, images, copy, scripts));

Refreshing the browser with Browsersync

Let’s now improve our watch task by installing Browsersync, a plugin that refreshes the browser each time tasks finish running.

npm install browser-sync gulp --save-dev

As usual, let’s import it:

import browserSync from "browser-sync";

Next, we will initialize a Browsersync server and write two new tasks:

const server = browserSync.create(); export const serve = done => {   server.init({     proxy: "http://localhost/yourFolderName" // put your local website link here   });   done(); }; export const reload = done => {   server.reload();   done(); };

In order to control the browser using Browsersync, we have to initialize a Browsersync server. This is different from a local server where WordPresss would typically live. the first task is serve, which starts the Browsersync server, and is pointed to our local WordPress server using the proxy option. The second task will simply reload the browser.

Now we need to run this server when we are developing our theme. We can add the serve task to the dev series tasks:

export const dev = series(clean, parallel(styles, images, copy, scripts), serve, watchForChanges);

Now run npm start and the browser should open up a new URL that’s different than the original one. This URL is the one that Browsersync will refresh. Now let’s use the reload task to reload the browser once tasks are done:

export const watchForChanges = () => {   watch('src/scss/**/*.scss', series(styles, reload));   watch('src/images/**/*.{jpg,jpeg,png,svg,gif}', series(images, reload));   watch(['src/**/*','!src/{images,js,scss}','!src/{images,js,scss}/**/*'], series(copy, reload));   watch('src/js/**/*.js', series(scripts, reload));   watch("**/*.php", reload); }

As you can see, we added a new line to run the reload task every time a PHP file changes. We are also using series() to wait for our styles, images, scripts and copy tasks to finish before reloading the browser. Now, run npm start and change something in a Sass file. The browser should reload automatically and changes should be reflected after refresh once the tasks have finished running.

Don’t see CSS or JavaScript changes after refresh? Make sure caching is disabled in your browser’s inspector.

We can make even one more improvement to the styles tasks. Browsersync allows us to inject CSS directly to the page without even having to reload the browser. And this can be done by adding at the very end of the styles task:

export const styles = () => {   return src(['src/scss/bundle.scss', 'src/scss/admin.scss'])     .pipe(gulpif(!PRODUCTION, sourcemaps.init()))     .pipe(sass().on('error', sass.logError))     .pipe(gulpif(PRODUCTION, postcss([ autoprefixer ])))     .pipe(gulpif(PRODUCTION, cleanCss({compatibility:'ie8'})))     .pipe(gulpif(!PRODUCTION, sourcemaps.write()))     .pipe(dest('dist/css'))     .pipe(; }

Now, in the watchForChanges task, we won’t have to reload for the styles task any more, so let’s remove the reload task from it:

export const watchForChanges = () => {   watch('src/scss/**/*.scss', styles);   .   . }

Make sure to stop watchForChanges if it’s already running and then run it again. Try to modify any file in the scss folder and the changes should appear immediately in the browser without even reloading.

Packaging the theme in a ZIP file

WordPress themes are generally packaged up as a ZIP file that can be installed directly in the WordPress admin. We can create a task that will take the required theme files and ZIP them up for us. To do that we need to install another Gulp plugin: gulp-zip.

npm install --save-dev gulp-zip

And, as always, import it at the top:

import zip from "gulp-zip";

Let’s also import the JSON object in the package.json file. We need that in order to grab the name of the package which is also the name of our theme:

import info from "./package.json";

Now, let’s write our task:

export const compress = () => { return src([   "**/*",   "!node_modules{,/**}",   "!bundled{,/**}",   "!src{,/**}",   "!.babelrc",   "!.gitignore",   "!gulpfile.babel.js",   "!package.json",   "!package-lock.json",   ])   .pipe(zip(`$ {}.zip`))   .pipe(dest('bundled')); };

We are passing the src() the files and folders that we need to compress, which is basically all files and folders (**/), except a few specific types of files, which are preceded by !. Next, we are piping the gulp-zip plugin and calling the file the name of the theme from the package.json file ( The result is a fresh ZIP file an a new folder called bundled.

Try running gulp compress and make sure it all works. Open up the generated ZIP file and make sure that it only contains the files and folders needed to run the theme.

Normally, though, we only need to ZIP things up *after* the theme files have been built. So let’s add the compress task to the build task so it only runs when we need it:

export const build = series(clean, parallel(styles, images, copy, scripts), compress);

Running npm run build should now run all of our tasks in production mode.

Replacing the placeholder prefix in the ZIP file

One step we need to do before zipping our files is to scan them and replace the themename placeholder with the theme name we plan to use. As you may have guessed, there is indeed a Gulp plugin that does that for us, called gulp-replace.

npm install --save-dev gulp-replace

Then import it:

import replace from "gulp-replace";

We want this task to run immediately before our files are zipped, so let’s modify the compress task by slotting it in the right place:

export const compress = () => { return src([     "**/*",     "!node_modules{,/**}",     "!bundled{,/**}",     "!src{,/**}",     "!.babelrc",     "!.gitignore",     "!gulpfile.babel.js",     "!package.json",     "!package-lock.json",   ])   .pipe(replace("_themename",   .pipe(zip(`$ {}.zip`))   .pipe(dest('bundled')); };

Try to building the theme now with npm run build and then unzip the file inside the bundled folder. Open any PHP file where the _themename placeholder may have been used and make sure it’s replaced with the actual theme name.

There is a gotcha to watch for that I noticed in the replace plugin as I was working with it. If there are ZIP files inside the theme (e.g. you are bundling WordPress plugins inside your theme), then they will get corrupted when they pass through the replace plugin. That can be resolved by ignoring ZIP files using a gulp-if statement:

.pipe(   gulpif(     file => file.relative.split(".").pop() !== "zip",     replace("_themename",   ) )

Generating a POT file

Translation is a big thing in the WordPress community, so for our final task, we let’s scan through all of our PHP files and generate a POT file that gets used for translation. Luckily, we also have a gulp plugin for that:

npm install --save-dev gulp-wp-pot

And, of course, import it:

import wpPot from "gulp-wp-pot";

Here’s our final task:

export const pot = () => {   return src("**/*.php")   .pipe(       wpPot({         domain: "_themename",         package:       })     )   .pipe(gulp.dest(`languages/$ {}.pot`)); };

We want the POT file to generate every time we build the theme:

export const build = series(clean, parallel(styles, images, copy, scripts), pot, compress);

Summing up

Here’s the complete Gulpfile, including all of the tasks we covered in this post:

import { src, dest, watch, series, parallel } from 'gulp'; import yargs from 'yargs'; import sass from 'gulp-sass'; import cleanCss from 'gulp-clean-css'; import gulpif from 'gulp-if'; import postcss from 'gulp-postcss'; import sourcemaps from 'gulp-sourcemaps'; import autoprefixer from 'autoprefixer'; import imagemin from 'gulp-imagemin'; import del from 'del'; import webpack from 'webpack-stream'; import named from 'vinyl-named'; import browserSync from "browser-sync"; import zip from "gulp-zip"; import info from "./package.json"; import replace from "gulp-replace"; import wpPot from "gulp-wp-pot";   const PRODUCTION =;   const server = browserSync.create();   export const serve = done => {     server.init({       proxy: "http://localhost:8888/starter"     });     done();   };   export const reload = done => {     server.reload();     done();   };   export const clean = () => del(['dist']);        export const styles = () => {   return src(['src/scss/bundle.scss', 'src/scss/admin.scss'])     .pipe(gulpif(!PRODUCTION, sourcemaps.init()))     .pipe(sass().on('error', sass.logError))     .pipe(gulpif(PRODUCTION, postcss([ autoprefixer ])))     .pipe(gulpif(PRODUCTION, cleanCss({compatibility:'ie8'})))     .pipe(gulpif(!PRODUCTION, sourcemaps.write()))     .pipe(dest('dist/css'))     .pipe(;   }   export const images = () => {   return src('src/images/**/*.{jpg,jpeg,png,svg,gif}')     .pipe(gulpif(PRODUCTION, imagemin()))     .pipe(dest('dist/images'));   }   export const copy = () => {     return src(['src/**/*','!src/{images,js,scss}','!src/{images,js,scss}/**/*'])     .pipe(dest('dist'));   }     export const scripts = () => {       return src(['src/js/bundle.js','src/js/admin.js'])       .pipe(named())       .pipe(webpack({         module: {         rules: [           {             test: /.js$ /,             use: {               loader: 'babel-loader',               options: {                 presets: []                 }               }             }           ]         },         mode: PRODUCTION ? 'production' : 'development',         devtool: !PRODUCTION ? 'inline-source-map' : false,         output: {           filename: '[name].js'         },         externals: {           jquery: 'jQuery'         },       }))       .pipe(dest('dist/js'));     }     export const compress = () => {       return src([         "**/*",         "!node_modules{,/**}",         "!bundled{,/**}",         "!src{,/**}",         "!.babelrc",         "!.gitignore",         "!gulpfile.babel.js",         "!package.json",         "!package-lock.json",       ])       .pipe(         gulpif(           file => file.relative.split(".").pop() !== "zip",           replace("_themename",         )       )       .pipe(zip(`$ {}.zip`))       .pipe(dest('bundled'));     };     export const pot = () => {       return src("**/*.php")         .pipe(           wpPot({             domain: "_themename",             package:           })         )       .pipe(dest(`languages/$ {}.pot`));     };     export const watchForChanges = () => {       watch('src/scss/**/*.scss', styles);       watch('src/images/**/*.{jpg,jpeg,png,svg,gif}', series(images, reload));       watch(['src/**/*','!src/{images,js,scss}','!src/{images,js,scss}/**/*'], series(copy, reload));       watch('src/js/**/*.js', series(scripts, reload));       watch("**/*.php", reload);     }      export const dev = series(clean, parallel(styles, images, copy, scripts), serve, watchForChanges);     export const build = series(clean, parallel(styles, images, copy, scripts), pot, compress);     export default dev;

Phew, that’s everything! I hope you learned something from this series and that it helps streamline your WordPress development flow. Let me know if you have any questions in the comments. If you are interested in a complete WordPress theme development course, make sure to check out my course on Udemy with a special discount for you. 😀

The post Gulp for WordPress: Creating the Tasks appeared first on CSS-Tricks.


, , ,

The Dark and Colorful Art of Michael Reeder


Gulp for WordPress: Initial Setup

This is the first part of a two-part series on creating a Gulp workflow for WordPress theme development. This first part covers a lot of ground for the initial setup, including Gulp installation and an outline of the tasks we want it to run. If you’re interested in how the tasks are created, then stay tuned for part two.

Earlier this year, I created a course for building premium WordPress themes. During the process, I wanted to use a task runner to concatenate and minify JavaScript and CSS files. I ended up using that task runner to automate a lot of other tasks that made the theme much more efficient and scalable.

The two most popular task runners powered by Node are Gulp and Grunt. I went with Gulp after a good amount of research, it appeared to have an intuitive way to write tasks. It uses Node streams to manipulate files and JavaScript functions to write the tasks, whereas Grunt uses a configuration object to define tasks — which might be fine for some, but is something that made me a little uncomfortable. Also, Gulp is a bit faster that Grunt because of those Node streams and faster is always a good thing to me!

So, we’re going to set Gulp up to do a lot of the heavy lifting for WordPress theme development. We’ll cover the initial setup for now, but then go super in-depth on the tasks themselves in another post.

Article Series:

  1. Initial Setup (This Post)
  2. Creating the Tasks (Coming Tomorrow!)

Initial theme setup

So, how can we use Gulp to power the tasks for a WordPress theme? First off, let’s assume our theme only contains the two files that WordPress requires for any theme: index.php and styles.css. Sure, most themes are likely to include many more files that this, but that’s not important right now.

Secondly, let’s assume that our primary goal is to create tasks that help manage our assets, like minify our CSS and JavaScript files, compile Sass to CSS, and transpile modern JavaScript syntax (e.g. ES6, ES7, etc..) into ES5 in order to support older browsers.

Our theme folder structure will look like this:

themefolder/ ├── index.php ├── style.css └── src/     ├── images/     │   └── cat.jpg     ├── js/     │   ├── components/     │   │   └── slider.js     │   └── bundle.js     └── scss/         ├── components/         │   └── slider.scss         └── bundle.scss

The only thing we’ve added on top of the two required files is a src directory where our original un-compiled assets will live.

Inside of that src directory, we have an images subdirectory as well as others for our JavaScript and Sass files. And from, there, the JavaScript and Sass subdirectories are organized into components that will be called from their respective bundle file. So, for example, bundle.js will import and include slider.js when our JavaScript tasks run so all our code is concatenated into a single file.

Identifying Gulp tasks

OK, next we want Gulp tasks to a create a new dist directory where all of our compiled, minified and concatenated versions of our assets will be distributed after the tasks have completed. Even though we’re calling this directory dist in this post because it is short for “distribution,” it could really be called anything, as long as Gulp knows what it is called. Regardless, these are the assets that will be distributed to end users. The src folder will only contain the files that we edit directly during development.

Identifying which Gulp tasks are the best fit for a project will depend on the project’s specific needs. Some tasks will be super helpful on some projects but completely irrelevant on others. I’ve identified the following for us to cover in this post. You’ll see that one or two are more useful in a WordPress context (e.g. the POT task) than others. Yet, most of these are broad enough that you’re likely to see them in many projects that use Gulp to process Sass, JavaScript and image assets.

  • Styles Task: This task is responsible for compiling the bundle.scss file in the scss subdirectory to bundle.css in a css directory located in the dist directory. This task will also minify the generated CSS file so that its is it’s smallest possible size when used in production.

We will talk about production vs. development modes during the article. Note that we will not create a task to concatenate CSS files. The bundle.scss file will act as an entry point for all . <code>scss files that we want to include. In other words; any Sass or CSS files you want to include in your project, just import them in the bundle.scss file using @import statements. For instance, in our example folder, we can use @import ./components/slider'; to import the slider.scss file. This way in our task we will have to compile and minify only one file (bundle.css).

  • Scripts Task: Similar to the Styles task, this task will transpile bundle.js from ES6 syntax to ES5, then minify the file for production.

We will only compile bundle.js. Any other JavaScript files we want to include will be done using ES6 import statements. But in order for those import statements to work on all browsers, we will need to use a module bundler. We’re going to use webpack as our bundler. If this is your first time working with it, this primer is a good place to get an overview of what it is and does.

  • Images Task: This task will simply copy images from src/images and send them to dist/images after the files have been compressed to their smallest size.
  • Copy Task: This task will be responsible for copying any other files or folders that are not in /src/images, /src/js or /src/scss and post them to the dist directory.

Remember. the src folder will contain the files that are only used during development and that will not be included in the final theme package. Thus, any assets other than our images, JavaScript and Sass files need to be copied posted to the dist folder. For instance, if we have a /src/fonts folder, we would want to copy the files in there into the dist directory so they get included in the final deliverable.

  • POT Task: As the name suggests, this task will scan all your theme’s PHP files and generate a .pot (i.e. translation) file from gettext calls in the files. This is the most WordPress-centric of all the tasks we’re covering here.
  • Watch Task: This task will literally watch for changes in your files. When a change is made, certain tasks will be triggered and executed, depending on the type of file that changed.

For instance, if we change a JavaScript file, then the Scripts task should do its magic and then it would be ideal if the browser refreshed for us automatically so we can see those changes. Further, If we change a PHP file, then let’s simply refresh the browser since PHP files don’t rely on any other tasks in our project. We’ll be using a Gulp plugin called Browsersync to handle browser refreshes, but we’ll get to that and other plugins a little later.

  • Compress Task: As you might expect, all the plugins that we use to write our tasks will be managed using npm. So, our theme folder will contain another folder, like node_modules, that in turn, contains files like package.json and other configuration files that define our project’s dependencies — and these files and folders are only needed during development. During production, we can take out the necessary files for our theme and leave the unneeded development files behind. That’s what this task will do; it will create a ZIP file that only contains the necessary files to run our theme.

As a bonus step for the compress task, if you are creating a theme that you intend to publish on or maybe on a website like ThemeForest, then you probably already know that all functions in your theme must be prefixed with a unique prefix:

function mythemename_enqueue_assets() {   // function body }

So, if you are creating a lot of themes. you’ll need to easily reuse functions in different themes but change the prefix to the name of the theme, to prevent conflicts. We can prefix our functions with a placeholder prefix and then replace all instances of that placeholder in the compress task. For instance, we can choose the string _themename as a place holder, and when we compress our theme we will replace all ‘_themename’ strings to the actual theme name:

function _themename_enqueue_assets() {   // function body }

This can also apply to anywhere we use the theme name for example in the text domain:

<?php _e('some string', '_themename'); ?>
  • Develop Task: This task does nothing new. It runs when as we develop our theme. It cleans the dist folder, runs the Styles, Scripts, Images and Copy tasks in development mode (i.e. without minifying any of the assets), then watches for file changes to refresh the browser for us.
  • Build Task: This task is intended to build our files for production. It will do all the same cleaning and tasks as the Develop task, but in production mode (i.e. minify the assets in the process) and generate a new POT file for translation updates. After it runs, our dist folder should contain the the files that are ready for distribution.
  • Bundle Task: This task will simply run the build task, will making sure that all the files in the dist folder are minified and ready for distribution. Then, it will run the Compress task, which bundles all of the production-ready files and folders into a ZIP file. We want a ZIP file because that is the format WordPress recognizes to extract and install a theme.

Here’s how our file structure looks after our tasks complete:

themefolder/ ├── index.php ├── style.css ├── src/ └── dist/     ├── images/     │   └── cat.jpg // after compression     ├── js/     │   └── bundle.js // bundled with all imported files (minified in prodcution)     └── scss/         └── bundle.scss // bundled with all imported files (minified in production)

Now that we know what tasks we’re going to use on our project and that they do, let’s get into the process for installing Gulp into the project.

Installing Gulp

Before we install Gulp, we should make sure that we have Node and npm installed on our machines. We can do that by running these commands in the command line:

node --version
npm --version

…and, we should get some version number as seen here:

Now, let’s point the command line to the theme folder:

cd path/to/your/theme/folder

…and then run this command to initialize a new npm project:

npm init

This will prompt us with some options. The only important option in our case is the package name option. This is where the name of the theme can be provided — everything else can stay at their default setting. When choosing the theme name, make sure to only use lowercase characters and underscores while avoiding dashes and special characters since this theme name will be used to replace the functions placeholder that we mentioned earlier.

On to installing Gulp! First, we’ve got to install Gulp’s command line interface (gulp-cli) globally so we can use Gulp in the command line.

npm install --global gulp-cli

After that, we run this command in order to install Gulp itself in the theme directory:

npm install --save-dev gulp@next

The current stable release of Gulp is 3.9.1 at the time of this writing, but version 4.0 is already available in the project repository. The command above uses the @next which installs version 4.0. Removing that will install 3.9.1 (or whatever the current version is) instead.

To make sure everything is installed correctly, we’ll run this command:

gulp --version

Nice! Looks like we’re running version 4.0, which is the latest version at the time of this writing.

Writing Gulp tasks

Gulp tasks are defined in a a file in called gulpfile.js that we’ll need to create and place into the root of our theme.

Gulp is JavaScript at its core, so we can define a quick example task that logs something to the console.

var gulp = require('gulp'); gulp.task('hello', function() {   console.log('First Task'); })

In this example, we’ve defined a new task by calling gulp.task. The first argument for this function is the task’s name (hello) and the second argument is the function we want to run when that name is entered into the command line which, in this case, should print “First Task” into the console.

Let’s do that now.

gulp hello

Here’s what we get:

As you can see, we do indeed get the console.log('First Task') output we want. However, we also get an error saying that our task did not complete. All Gulp tasks require telling Gulp where to end the task, and we do that by calling a function that is passed as the first argument in our task function like so:

var gulp = require('gulp'); gulp.task('hello', function(cb) {   console.log('First Task');   cb(); })

Let’s try running gulp hello again and we should get the same output, but without the error this time.

There are some cases where we won’t have to call the cb() function, such as when a task returns a promise or a node stream. A node stream is what we will use in the tasks in this post, which means we will see it a lot throughout our article.

Here’s an example of a task that returns a promise. In this task, we won’t have to call the cb() function because gulp already knows that the task will end when the promise resolves or returns an error:

gulp.task('promise', function(cb) {   return new Promise(function(resolve, reject) {     setTimeout(function() {       resolve();     }, 300);   }); });

Now try and run ‘gulp promise’ and the task will complete without returning any errors.

Finally, it’s worth mentioning that Gulp accepts a default task that runs by typing gulp alone in the command line. All it takes is using “default” as the task name argument.

gulp.task('default', function(cb) {   console.log('Default Task');   cb(); });

Now, typing gulp by itself in the command line will run the task.

Whew! Now we know the basics of writing Gulp tasks.

There is one more thing we can do to improve things and that’s enabling ES6 syntax in the Gulpfile. This will allow us to use features like destructuring, import statements, and arrow functions, among others.

Using ES6 in the Gulpfile

The first step to use ES6 syntax in the Gulpfile is to rename it from gulpfile.js to gulpfile.babel.js. As you may already know, Babel is the compiler that compiles ES6 to ES5.

So, let’s install Babel and some of its required packages by running:

npm install --save-dev @babel/register @babel/preset-env @babel/core

After that, we have to create a file called .babelrc inside of our theme folder. This file will tell Babel which preset to use to compile our JavaScript. The contents of the .babelrc file will look like this:

{   "presets": ["@babel/preset-env"] }

Now we can use ES6 in our Gulpfile! Here’s how that would look if we were to re-write it:

import gulp from 'gulp'; export const hello = (cb) => {   console.log('First Task');   cb(); }  export const promise = (cb) => {   return new Promise((resolve, reject) => {     setTimeout(() => {       resolve();     }, 300);   }); };  export default hello

As you can see, we are importing Gulp using import and not require. In fact, there’s no longer any need to import Gulp at all! I included the import statement anyway to show it can be used instead of require. We’re allowed to skip importing Gulp because we don’t have to call gulp.task — instead, we only need to export a function, and the name of this function will be the name of the task. Further, all that’s needed to define a default function is use export default. And notice those arrow functions, too! Everything is so much more concise.

Let’s move on and start coding the actual tasks.

Development vs. Production

As we covered earlier, we need to create two modes: development and production. The reason we need to delineate between the two is that some details in our tasks will be time and memory-consuming, which only make sense in a production environment. For instance, the styles task needs to minify the CSS. However, the minification can take both time and memory — and if that process runs every single time something changes during development, that is not only unnecessary, but is very inefficient. It’s ideal for tasks to be as fast as possible during development.

We need to set a flag that specifies whether a task should run in one mode or the other. We can use a package called yargs that allows us to define these types of arguments while running a command. So, let’s install it and put it to use:

npm install --save-dev yargs

Now, we can add arguments to our command like so:

gulp hello --prod=true

…and then retrieve these argument in the Gulpfile:

import yargs from 'yargs'; const PRODUCTION =;  export const hello = (cb) => {   console.log(PRODUCTION);   cb(); }

Notice that the values we define in the command are available inside the yargs.argv object in the Gulpfile and in console.log(PRODUCTION). In our case this will output true, so PRODUCTION will be our flag that decides whether or not a function runs inside the tasks.

We’re all set up!

We covered a lot of ground here, but we now have everything we need to start writing tasks for our WordPress theme development. That just so happens to be the sole focus of the next part of this series, so stay tuned for tomorrow.

The post Gulp for WordPress: Initial Setup appeared first on CSS-Tricks.


, , ,

An Initial Implementation of clip-path: path();

One thing that has long surprised (and saddened) me is that the clip-path property, as awesome as it is, only takes a few values. The circle() and ellipse() functions are nice, but hiding overflows and rounding with border-radius generally helps there already. Perhaps the most useful value is polygon() because it allows us to draw a shape out of straight lines at arbitrary points.

Here’s a demo of each value:

See the Pen clip-path examples by Chris Coyier (@chriscoyier) on CodePen.

The sad part comes in when you find out that clip-path doesn’t accept path(). C’mon it’s got path in the name! The path syntax, which comes from SVG, is the ultimate syntax. It allows us to draw literally any shape.

More confusingly, there already is a path() function, which is what properties like offset-path take.

I was once so flabbergasted by all this that I turned it into a full conference talk.

The talk goes into the shape-outside property and how it can’t use path(). It also goes into the fact that we can change the d property of a literal <path>.

I don’t really blame anyone, though. This is weird stuff and it’s being implemented by different teams, which inevitably results in different outcomes. Even the fact that SVG uses unit-less values in the <path> syntax is a little weird and an anomaly in CSS-land. How that behaves, how values with units behave, what comma-syntax is allowed and disallowed, and what the DOM returns when asked is plenty to make your head spin.

Anyway! Along comes Firefox with an implementation!

Here’s that flag in Firefox (layout.css.clip-path-path.enabled):

And here’s a demo… you’ll see a square in unsupported browsers and a heart in the ones that support clip-path: path(); — which is only Firefox Nightly with the flag turned on at the time of this writing.

See the Pen clip-path: path()! by Chris Coyier (@chriscoyier) on CodePen.

A screenshot of clip-path: path() working in Firefox Nightly

Now, all we need is:

  • clip-path to be able to point to the URL of a <clipPath> in SVG, like url("#clip-path");
  • shape-outside to be able to use path()
  • shape-outside to be able to use a <clipPath>
  • offset-path to take all the other shape functions
  • Probably a bunch of specs to make sure this is all handled cleanly (Good luck, team!)
  • Browsers to implement it all


The post An Initial Implementation of clip-path: path(); appeared first on CSS-Tricks.


, , ,

People Talkin’ Shapes

Codrops has a very nice article on CSS Shapes from Tania Rascia. You might know shape-outside is for redefining the area by which text is floated around that element, allowing for some interesting design opportunities. But there are a couple of genuine CSS tricks in here:

  1. Float shape-outside elements both right and left to get text to flow between them.
  2. You can set shape-outside to take an image and use shape-image-threshold to adjust where the text flows, meaning you could even use a gradient!

Shapes are in the water recently, as Heydon Pickering recently published a short video on using them. He also covers things like clip-path and canvas and such:

We recently moved our long-time page on (basically faking) CSS shapes over to a blog post so it’s easier to maintain.

Robin also wrote Working with Shapes in Web Design that digs into all this. So many tricks!

See the Pen 10c03204463e92a72a6756678e6348d1 by CSS-Tricks (@css-tricks) on CodePen.

When we talk about CSS shapes, it’s almost like we’re talking about values moreso than properties. What I mean is that the value functions like polygon(), circle(), ellipse(), offset(), path(), etc. are more representative of “CSS shapes” than the properties they are applied to. Multiple properties take them, like shape-outside, clip-path, and offset-path.

I once did a whole talk on this:

The only thing that’s changed since then is that Firefox started allowing clip-path: path() behind the flag layout.css.clip-path-path.enabled (demo).

And don’t forget Jen Simmons was talking about the possibilities of CSS Shapes (in her lab demos) years earlier!

The post People Talkin’ Shapes appeared first on CSS-Tricks.


, ,

Animating Between Views in React

You know how some sites and web apps have that neat native feel when transitioning between two pages or views? Sarah Drasner has shown some good examples and even a Vue library to boot.

These animations are the type of features that can turn a good user experience into a great one. But to achieve this in a React stack, it is necessary to couple crucial parts in your application: the routing logic and the animation tooling.

Let’s start with animations. We’ll be building with React, and there are great options out there for us to leverage. Notably, the react-transition-group is the official package that handles elements entering and leaving the DOM. Let’s explore some relatively straightforward patterns we can apply, even to existing components.

Transitions using react-transition-group

First, let’s get familiar with the react-transition-group library to examine how we can use it for elements entering and leaving the DOM.

Single components transitions

As a simple example of a use case, we can try to animate a modal or dialog — you know, the type of element that benefits from animations that allow it enter and leave smoothly.

A dialog component might look something like this:

import React from "react";  class Dialog extends React.Component {   render() {     const { isOpen, onClose, message } = this.props;     return (       isOpen && (         <div className="dialog--overlay" onClick={onClose}>           <div className="dialog">{message}</div>         </div>       )     );   } }

Notice we are using the isOpen prop to determine whether the component is rendered or not. Thanks to the simplicity of the recently modified API provided by react-transition-group module, we can add a CSS-based transition to this component without much overhead.

First thing we need is to wrap the entire component in another TransitionGroup component. Inside, we keep the prop to mount or unmount the dialog, which we are wrapping in a CSSTransition.

import React from "react"; import { TransitionGroup, CSSTransition } from "react-transition-group";  class Dialog extends React.Component {   render() {     const { isOpen, onClose, message } = this.props;     return (       <TransitionGroup component={null}>         {isOpen && (           <CSSTransition classNames="dialog" timeout={300}>             <div className="dialog--overlay" onClick={onClose}>               <div className="dialog">{message}</div>             </div>           </CSSTransition>         )}       </TransitionGroup>     );   } }

Every time isOpen is modified, a sequence of class names changes will happen in the dialog’s root element.

If we set the classNames prop to "fade", then fade-enter will be added immediately before the element mounts and then fade-enter-active when the transition kicks off. We should see fade-enter-done when the transition finishes, based on the timeout that was set. Exactly the same will happen with the exit class name group at the time the element is about to unmount.

This way, we can simply define a set of CSS rules to declare our transitions.

.dialog-enter {   opacity: 0.01;   transform: scale(1.1); }  .dialog-enter-active {   opacity: 1;   transform: scale(1);   transition: all 300ms; }  .dialog-exit {   opacity: 1;   transform: scale(1); }  .dialog-exit-active {   opacity: 0.01;   transform: scale(1.1);   transition: all 300ms; }

JavaScript Transitions

If we want to orchestrate more complex animations using a JavaScript library, then we can use the Transition component instead.

This component doesn’t do anything for us like the CSSTransition did, but it does expose hooks on each transition cycle. We can pass methods to each hook to run calculations and animations.

<TransitionGroup component={null}>   {isOpen && (     <Transition       onEnter={node => animateOnEnter(node)}       onExit={node => animateOnExit(node)}       timeout={300}     >       <div className="dialog--overlay" onClick={onClose}>         <div className="dialog">{message}</div>       </div>     </Transition>   )} </TransitionGroup>

Each hook passes the node to the callback as a first argument — this gives control for any mutation we want when the element mounts or unmounts.


The React ecosystem offers plenty of router options. I’m gonna use react-router-dom since it’s the most popular choice and because most React developers are familiar with the syntax.

Let’s start with a basic route definition:

import React, { Component } from 'react' import { BrowserRouter, Switch, Route } from 'react-router-dom' import Home from '../views/Home' import Author from '../views/Author' import About from '../views/About' import Nav from '../components/Nav'  class App extends Component {   render() {     return (       <BrowserRouter>         <div className="app">           <Switch>             <Route exact path="/" component={Home}/>             <Route path="/author" component={Author} />             <Route path="/about" component={About} />           </Switch>         </div>       </BrowserRouter>     )   } }

We want three routes in this application: home, author and about.

The BrowserRouter component handles the browser’s history updates, while Switch decides which Route element to render depending on the path prop. Here’s that without any transitions:

Don’t worry, we’ll be adding in page transitions as we go.

Oil and water

While both react-transition-group and react-router-dom are great and handy packages for their intended uses, mixing them together can break their functionality.

For example, the Switch component in react-router-dom expects direct Route children and the TransitionGroup components in react-transition-group expect CSSTransition or Transition components to be direct children of it too. So, we’re unable to wrap them the way we did earlier.

We also cannot toggle views with the same boolean approach as before since it’s handled internally by the react-router-dom logic.

React keys to the rescue

Although the solution might not be as clean as our previous examples, it is possible to use the libraries together. The first thing we need to do is to move our routes declaration to a render prop.

<BrowserRouter>   <div className="app">     <Route render={(location) => {       return (         <Switch location={location}>           <Route exact path="/" component={Home}/>           <Route path="/author" component={Author} />           <Route path="/about" component={About} />         </Switch>       )}     /> </BrowserRouter>

Nothing has changed as far as functionality. The difference is that we are now in control of what gets rendered every time the location in the browser changes.

Also, react-router-dom provides a unique key in the location object every time this happens.

In case you are not familiar with them, React keys identify elements in the virtual DOM tree. Most times, we don’t need to indicate them since React will detect which part of the DOM should change and then patch it.

<Route render={({ location }) => {   const { pathname, key } = location    return (     <TransitionGroup component={null}>       <Transition         key={key}         appear={true}         onEnter={(node, appears) => play(pathname, node, appears)}         timeout={{enter: 750, exit: 0}}       >         <Switch location={location}>           <Route exact path="/" component={Home}/>           <Route path="/author" component={Author} />           <Route path="/about" component={About} />         </Switch>       </Transition>     </TransitionGroup>   ) }}/>

Constantly changing the key of an element — even when its children or props haven’t been modified — will force React to remove it from the DOM and remount it. This helps us emulate the boolean toggle approach we had before and it’s important for us here because we can place a single Transition element and reuse it for all of our view transitions, allowing us to mix routing and transition components.

Inside the animation function

Once the transition hooks are called on each location change, we can run a method and use any animation library to build more complex scenes for our transitions.

export const play = (pathname, node, appears) => {   const delay = appears ? 0 : 0.5   let timeline    if (pathname === '/')     timeline = getHomeTimeline(node, delay)   else     timeline = getDefaultTimeline(node, delay) }

Our play function will build a GreenSock timeline here depending on the pathname, and we can set as many transitions as we want for each different routes.

Once the timeline is built for the current pathname, we play it.

const getHomeTimeline = (node, delay) => {   const timeline = new Timeline({ paused: true });   const texts = node.querySelectorAll('h1 > div');    timeline     .from(node, 0, { display: 'none', autoAlpha: 0, delay })     .staggerFrom(texts, 0.375, { autoAlpha: 0, x: -25, ease: Power1.easeOut }, 0.125);    return timeline }

Each timeline method digs into the DOM nodes of the view and animates them. You can use other animation libraries instead of GreenSock, but the important detail is that we build the timeline beforehand so that our main play method can decide which one should run for each route.


I’ve used this approach on lots of projects, and though it doesn’t present obvious performance issues for inner navigations, I did notice a concurrency issue between the browser’s initial DOM tree build and the first route animation. This caused a visual lag on the animation for the first load of the application.

To make sure animations are smooth in each stage of the application, there’s one last thing we can do.

Profiling the initial load

Here’s what I found when auditing the application in Chrome DevTools after a hard refresh:

You can see two lines: one blue and one red. Blue represents the load event and red the DOMContentLoaded. Both intersect the execution of the initial animations.

These lines are indicating that elements are animating while the browser hasn’t yet finished building the entire DOM tree or it’s parsing resources. Animations account for big performance hits. If we want anything else to happen, we’d have to wait for the browser to be ready with these heavy and important tasks before running our transitions.

After trying a lot of different approaches, the solution that actually worked was to move the animation after these events — simple as that. The issue is that we can’t rely on event listeners.

window.addEventListener(‘DOMContentLoaded’, () => { })

If for some reason, the event occurs before we declare the listener, the callback we pass will never run and this could lead to our animations never happening and an empty view.

Since this is a concurrency and asynchronous issue, I decided to rely on promises, but then the question became: how can promises and event listeners be used together?

By creating a promise that gets resolved when the event takes place. That’s how.

window.loadPromise = new Promise(resolve => {   window.addEventListener(‘DOMContentLoaded’, resolve) })

We can put this in the document head or just before the script tag that loads the application bundle. This will make sure the event never happens before the Promise is created.

Plus, doing this allows us to use the globally exposed loadPromise to any animation in our application. Let’s say that we don’t only want to animate the entry view but a cookie banner or the header of the application. We can simply call each of these animations after the promise has resolved using then along with our transitions.

window.loadPromise.then(() =>

This approach is reusable across the entire codebase, eliminating the issue that would result when an event gets resolved before the animations run. It will defer them until the browser DOMContentLoaded event has passed.

See now that the animation is not kicking off until the red line appears.

The difference is not only on the profiling report — it actually solves an issue we had in a real project.

Wrapping up

In order to act as reminders, I created a list of tips for me that you might find useful as you dig into view transitions in a project:

  • When an animation is happening nothing else should be happening. Run animations after all resources, fetching and business logic have completed.
  • No animation is better than crappy animations If you can’t achieve a good animation, then removing it is a fair sacrifice. The content is more important and showing it is the priority until a good animation solution is in place.
  • Test on slower and older devices. They will make it easier for you to catch spots with weak performance.
  • Profile and base your improvements in metrics. Instead of guessing as you go, like I did, see if you can spot where frames are being dropped or if something looks off and attack that issue first.

That’s it! Best of luck with animating view transitions. Please post a comment if this sparked any questions or if you have used transitions in your app that you’d like to share!

The post Animating Between Views in React appeared first on CSS-Tricks.


, , ,


(This is a sponsored post.)

I just read a nicely put together story about WooCommerce over on the CodeinWP blog. WooCommerce started life as WooThemes, sort of a “premium themes” business started by just a couple of fellas who had never even met in person. Two years and a few employees later they launch WooCommerce, and 2 years after that it hits a million downloads. A major success story, to be sure, but a collaborative and remote-work based one that wasn’t exactly overnight. Another 2 years and Automattic picks them up and the WooThemes part is spun down.

Now we’re 3-4 years into WooCommerce being an Automattic project and it’s looking at nearly 60 million downloads, 4 million of which are active. A number they are saying is about 30% of all eCommerce on the web. Daaaaang. I’ve used WooCommerce a number of times and it always does a great job for me.

Direct Link to ArticlePermalink

The post WooCommerce appeared first on CSS-Tricks.