tubby — youtube playlist search

• ⚡ demo: chasemoskal.com/tubby
• web component to display a youtube playlist with a search bar

using the <tubby-youtube-explorer> component on your webpage

1. first, you need access to the youtube api, and have your api key ready

   • login with the google developer console, and get yourself a youtube api key
   • you have to whitelist your domains with that api key
   • you can have an api key that allows 'localhost' for testing

2. install tubby on your page

   • include the following snippet on your page

     <script type="importmap-shim">
       {
         "imports": {
           "lit-html/": "https://unpkg.com/[email protected]/",
           "lit-html": "https://unpkg.com/[email protected]/lit-html.js",
           "lit-element/": "https://unpkg.com/[email protected]/",
           "lit-element": "https://unpkg.com/[email protected]/lit-element.js",
           "tubby/": "https://unpkg.com/[email protected]/",
           "tubby": "https://unpkg.com/[email protected]/dist/index.js"
         }
       }
     </script>
     <script type="module-shim">
       import "tubby/dist/register-all.js"
     </script>
     <script src="https://unpkg.com/[email protected]/dist/es-module-shims.js"></script>

     here's what's going on in this snippet

     • import maps
       • we provide an import map which locates tubby and its dependencies
       • of course, for the import map, we're using guy bedford's awesome polyfill: es-module-shims
       • if you're from the future, you won't need the -shim's
     • register the web components
       • we run import "tubby/dist/register-all.js" to register the <tubby-youtube-explorer> component

3. place the explorer html on your page

   <tubby-youtube-explorer
     search
     max-description-length="360"
     api-key="AIzaSyDeHpB9W14feQs8myoWgFAZOCrDeKMLRwE"
     playlist-id="UUL_f53ZEJxp8TtlOkHwMV9Q"
     canned="dist-demo/canned-videos.json"
     >
   </tubby-youtube-explorer>

   the youtube explorer fetches the specified youtube playlist, and displays video thumbnails in a grid formation, and optionally has a searchbar for filtering the video selection

   youtube's api limits us to 50 videos at a time, so tubby does the hard work and requests every video sequentially, using back-to-back calls until it finds every video in the playlist

   tubby has a simple dev-time caching mechanism i call "canning videos"

4. know how to configure the explorer for your use-case

   all of tubby's configurable properties are optional, but you probably want to use a combination that loads some videos

   • ⚓ string attributes [api-key] and [playlist-id]
     tubby will fetch all the videos in that playlist

   • ⚓ string attributes [api-key] and [channel-id]
     tubby will make an extra round trip to find the playlist called "uploads".
     it is faster to provide the [playlist-id] attribute instead

   • ⚓ string attribute [canned] url
     tubby will fetch these cached videos.
     if a playlist is also being fetched, the canned videos will provide a headstart that can prevent many round-trips and greatly improve the loading time.
     more on canning videos later in the readme

   • ⚓ boolean attribute [search]
     tubby will display a search bar that the user can use to narrow the video listing

   • ⚓ number attribute [max-description-length]
     character number limit for video descriptions

   optional javascript stuff

   
   // we can create the explorer programmatically
   const explorer = document.createElement("tubby-youtube-explorer")
   document.body.appendChild(explorer)
   
   // we can set the attributes as properties
   explorer["search"] = true
   explorer["max-description-length"] = 360
   explorer["api-key"] = "AIzaSyDeHpB9W14feQs8myoWgFAZOCrDeKMLRwE" 
   explorer["playlist-id"] = "UUL_f53ZEJxp8TtlOkHwMV9Q"
   explorer["canned"] = "dist-demo/canned-videos.json"
   
   // the explorer supports these callbacks
   explorer.onReady = () => console.log(`tubby loaded ${explorer.videos.length} videos`)
   explorer.onError = error => console.warn(error)
   
   // in addition to the two callbacks, there are matching custom events
   explorer.addEventListener("ready", () => console.info("ready"))
   explorer.addEventListener("error", () => console.info("error"))

5. know how to style the explorer

   <style>
     tubby-video-explorer {
       --focus-outline: 2px solid #0ef;
       --tubby-search-focus-outline-offset: -4px;
       --tubby-pending-bg: rgba(0,0,0, 0.2);
       --tubby-error-bg: rgba(128,0,0, 0.2);
       --tubby-pending-color: white;
       --tubby-error-color: yellow;
       --tubby-search-icon-color: rgba(0,0,0, 0.5);
       --tubby-results-blink-color: rgba(255,255,255, 0.5);
       --tubby-grid-bg: rgba(0,0,0, 0.2);
       --tubby-video-color: white;
       --tubby-video-bg: linear-gradient(to bottom right,
         rgba(25,25,25, 0.95) 32%,
         rgba(25,25,25, 0.6)
       );
     }
   </style>

6. canning videos with tubby's tubby-download commandline tool

   you can give tubby a headstart (greatly reducing load times for large playlists), by using the tubby-download cli to generate a cache of "canned" videos in a json file, here's how

   • npm install tubby
     install tubby into your npm package

   • tubby-download --api-key=AIzaSyDeHpB9W14feQs8myoWgFAZOCrDeKMLRwE --playlist-id=UUL_f53ZEJxp8TtlOkHwMV9Q > dist/canned-videos.json
     call the tubby binary in a package.json script, supply the api-key and playlist-id, and specify where tubby should write the canned videos json file

   • then you just provide the [canned] attribute like so: <tubby-youtube-explorer canned="dist/canned-videos.json">

tubby is open source and contributions are welcome!

please consider opening an issue or contributing pull requests :)