A “suggest-as-you-type” search widget in 57 lines of JavaScript code.

While developing Influen$e, I wanted to make the search function a bit more robust for people searching for an influencer's Instagram account by name.

Because discovery is a critical distinguishing feature of Influen$e, I also wanted to build in a simple recommendation function that targeted customers based on the relevance of influencers' content to their core brand values.

In this article, I'll share the code and walk you through how the function works.

Shortcut this

If you know what you're doing, you can get the code from GitHub.

Get it on GitHub

Let's begin

We'll start with our search form. It looks like this:

I'm using UIKit 3 for my CSS framework, so this part is easy. Here's the code:

<form id="SearchMemberForm">
  <div class="uk-margin-small">
      <div class="uk-inline uk-width-1-1">
          <button id="MemberSubmit" type="submit"
            class="uk-form-icon uk-form-icon-flip"
            uk-icon="icon: search"></button>
          <input id="MemberInput" type="search" name="ig_handle"
            class="member-search-input uk-input uk-form-large"
            placeholder="e.g., cristiano" autocapitalize="none"
            autocorrect="off" autocomplete="off" />
          <div id="PresetMember" class="uk-width-1-1 uk-margin-remove"
            uk-dropdown="mode: click; pos: bottom-justify;
            boundary: .member-search-input"></div>
      </div>
  </div>
</form>

Getting the data with Fetch

The first thing we need to do is get an Array of objects to populate suggestions, right? So, we'll make a request to an API or cloud function or whatever.

I'm using Fetch because all I need is a JSON object, but if you want more granular control you can use XHR just the same.

fetch('://<server>/api/', {
      method: "GET",
      cache: "no-cache",
      credentials: "same-origin",
      headers = {
          "Accept": "application/json",
          "Content-Type": "application/json; charset=utf-8",
          'X-Requested-With': 'XMLHttpRequest'
      };
  })
  .then(res => res.json())
  .then(json => {
      console.log(`JSON response ${json}`)
  })
  .catch(error => console.log(error));

Fetch is pretty straightforward:

I send a GET request to my API endpoint at ://<server>/api/.
.then(res => res.json()) renders the response as JSON.
.then(json => {}) is where we can use that JSON.
.catch(error =>...); catches any errors in our request.

One thing to bear in mind, Fetch doesn't do any error checking, it only returns a fail if something went wrong with the actual request to the server, like a 404 or 500. So, you'll need to write your own error catching code.

If you run that code, you'll get an array of objects. In my case it's an Array of social media influencer objects:

json-response

Coding the UI

Now we need to turn that Array into a nice drop down list. There are two types of lists I'm going to generate.

One will be for automated recommendations when the user clicks the search field.
The other will be our suggest-as-you-type list.

To do that, we'll pass our Array to a new function, called loadLists(). Let's update our Fetch response to call loadLists(json).

    fetch(endpoint, {
        method: "GET",
        cache: "no-cache",
        credentials: "same-origin",
        headers: headers,
    })
    .then(response => response.json())
    .then(json => loadLists(json))  // <-- Call loadLists() here
    .catch(error => console.log(error));

This passes our Array of member objects to a new function called loadLists():

let loadLists = (json) => {
    let members = [...json.members];

    let searchSuggest = ['input', function(e) {
            filterList(event.target.value, members);
        }, false];

    formInput.addEventListener(...searchSuggest);
    presetMembers.innerHTML = userList(members.slice(0, 3));
}

There's a couple of things going on here.

First, we're using a spread operator [...json.members] to get to the objects in our array.
Next, we're creating an Array called searchSuggest, and within calling the filterList() function.
Then, we add an event listener to the input field in our HTML form.
And finally, we're loading the top three recommended influencers into the presetMembers drop down menu (I'll get to that soon).

The reason we created the searchSuggest Array is because we want to pass the event and the function call to our event listener. It's just a cleaner way of separating our functions.

Okay, let's have a look at the filterList(); function.

let filterList = (key, members) => {
  let listEl = document.getElementById('MemberList');
  listEl.innerHTML = '';  // <-- Clear the field to avoid repeated entries.
  for (let i = 0; i < members.length; i++) {
    if ((members[i].ig_handle.toLowerCase()).indexOf(key.toLowerCase()) > -1) {
  let node = () => (`<li class="member-link">${members[i].ig_handle}</li>`);
  if (listEl.childNodes.length < 5) {
                                   listEl.insertAdjacentHTML('beforeend', node());
                }
            }
            if (key.length < 2) {
                listEl.innerHTML = '';
                presetMembers.innerHTML = userList(members.slice(0, 3));
            }
        }
    };

Wow, that's the biggest chunk of code so far. Let's walk through it.

First, we're looking at a couple of arguments key and members. So the argument members is obviously our Array of members we got from our Fetch API call.

But what is key?

Remember when we created that Array called searchSuggest? Reminder, the event we added is

searchSuggest = ['input', function(e) {...

We're adding an event listener for input, basically saying for every character the user types into our form field, fire the input event. Get it? The input event is a keystroke, which is our key argument.

In other words, every time a user types a character into our search field, fire the filterList() function.

Next, we're getting the MemberList element from our HTML. This is where we'll generate our member list.

Let's take a look at our for loop.

In this line we're converting all text to lowercase so our searches will be case insensitive:

if ((members[i].ig_handle.toLowerCase()).indexOf(key.toLowerCase()) > -1)

Then for each character entered, we iterate through our members list and create a new list item node:

let node = () => (
    `<li class="member-link">
        ${members[i].ig_handle}
    </li>`
);

${members[i].ig_handle} is how we're identifying our members. In my case I'm looking them up by the Instagram username ig_handle.

Then we get up to five members and add them to our list using insertAdjacentHTML:

if (listEl.childNodes.length < 5) {
    listEl.insertAdjacentHTML('beforeend', node());
}

And finally, if the user enters less than two characters -- or backspaces to less than two characters -- reload our presetMembers, hiding our suggestions:

if (key.length < 2) {
    listEl.innerHTML = '';
    presetMembers.innerHTML = userList(members.slice(0, 3));
}

Stitching it all together

Let get our elements using getElementById.

let formInput = document.getElementById('MemberInput')  // <-- form <input> field
let presetMembers = document.getElementById('PresetMember');  // <-- div for our list

Then we're going to use ES6 template literals to generate our <ul> markup that we're inserting into the presetmember div.

let userListItem = (item) => (
    `<li class="uk-width-1-1">
        <img class="uk-border-circle uk-width-1-4 uk-float-left" src="${item.profile_pic}" data-uk-image />
        <p class="uk-width-3-4 uk-text-lead uk-text-truncate uk-padding-small uk-padding-remove-vertical">${item.ig_handle}</p>
    </li>`
);

let userList = (list) => (
    `<ul id="MemberList" class="uk-list uk-list-divider uk-padding-remove-vertical">
        ${list.map(userListItem).join('')}
    </ul>`
);

Save the file and (hopefully) it should all work beautifully!