Experience Sitecore !

There was much of recent buzz around containers as technology and Docker in particular, especially now where more and more community efforts focused at Docker in conjunction with Sitecore. Plenty of articles do explain how it works on the very top level, what the benefits are but very rare do precise guidance. As for an ultimate beginner - I know how it is important to get a quick start, the minimal positive experience as a further point of development. This blog exactly about how to achieve the bare minimal Sitecore running in Docker.

Content

• Terminology
• Installing Docker
• Docker registry
• Preparing images
• Building images
• Running containers
• Stopping and clean-up
• Afterthoughts

1. Terminology

• Docker images - a blueprint for creating containers, is what you pull from remote registry
• Docker container - an implementation of a specific image, you run and work with containers, not images
• Docker repository - a logical unit containing one or many built images (specified by tags)
• Docker registry - works like a remote git repo, it's a hosted solution to you push your built images into

There are plenty more of terminology, but these are the essentials for a demo below

2. Installing Docker

If you have it up and running, you may skip to the next part.

In order to operate the repository for given walk-through, you need to have Windows 10 x64 with at least build 1809.

The simplest way is to install it from Chocolatey gallery:

cinst docker-desktop

Once done, you'll need to switch a mode. Docker for Windows can work in either Windows or Linux mode at the same time - which means you cannot mix types of containers.

One of the biggest issues at the moment is the size of Windows base images - the minimal Nano Server with almost everything cut off is already 0.5Gb and Server Core (this one either does not have UI - just a console) goes to 4Gb. That's too much comparing to minimal Linux images starting from as little as 5 megs. That's why it may seem very attractive to run Solr from Linux image (as both Solr and Java it requires are cross-platform and there are ready-to-use images there), same for Ms SQL Server which also has been ported on Linux and its images are also available.

Until the very-very recent the short answer was no - one could manage only single type of container at a time (while already running Linux container will keep up running unmanageable, there are also few workarounds how to make them work in parallel, but that's out of scope for now), but as from April 2019 it is doable (from Linux mode on Windows), I managed to combine NGINX on Linux with IIS on Windows.

Switching mode from UI is done by right-clicking Docker icon context menu from a system tray:

3. Docker registry

What you should do next is to provide a docker registry. Docker Hub is probably the first option for any docker beginner.

Docker Hub however allows only one private repository for free. You need to ensure sure all your repositories are private. Images you're building will contain your license file, having them in a public access will also be treated as sharing Sitecore binaries on your own, which you're not allowed - only Sitecore can distribute that publicly.

"YOUR_DOCKER_REGISTRY_PASSWORD" | docker login --username martinmiles --password-stdin

# Load module
Import-Module (Join-Path $PSScriptRoot "\modules\SitecoreImageBuilder") -Force

# Build and push
SitecoreImageBuilder\Invoke-Build `
    -Path (Join-Path $PSScriptRoot "\images") `
    -InstallSourcePath "c:\Docker\Install\9.1.1" `
    -Registry "martinmiles" `
    -Tags "*" `
    -PushMode "WhenChanged"

set-executionpolicy unrestricted

version: '2.4'

services:

  sql:
    image: sitecore-xm1-sqldev:9.1.1-windowsservercore-ltsc2019
    volumes:
      - .\data\sql:C:\Data
    mem_limit: 2GB
    isolation: hyperv
    ports:
      - "44010:1433"

  solr:
    image: sitecore-xm1-solr:9.1.1-nanoserver-1809
    volumes:
      - .\data\solr:C:\Data
    mem_limit: 1GB
    isolation: hyperv
    ports:
      - "44011:8983"

  cd:
    image: sitecore-xm1-cd:9.1.1-windowsservercore-ltsc2019
    volumes:
      - .\data\cd:C:\inetpub\sc\App_Data\logs
    isolation: hyperv
    ports:
      - "44002:80"
    links:
      - sql
      - solr

  cm:
    image: sitecore-xm1-cm:9.1.1-windowsservercore-ltsc2019
    volumes:
      - .\data\cm:C:\inetpub\sc\App_Data\logs
    isolation: hyperv
    ports:
      - "44001:80"
    links:
      - sql
      - solr

docker-compose up

docker exec -it CONTAINER_HASH cmd

docker-compose down

docker ps

docker image ls

docker prune -a

Always having productivity and effectiveness as the major criteria of measuring my work, I have identified most of the time-wasters and came up with the list of things that slow down my development process. It is important to differ productivity from performance - the first applies to your personal bottlenecks while the second - to your solution or configuration. Performance tuning has been covered by numerous blog posts so far so I will be mentioning only those affecting my personal productivity. I am also not going to cover things like CI / CD and Application Insights / KUDOs at this post, while all of the mentioned is the proven great tools, they are not related to pure development productivity adn more tend to be DevOps.

I am sharing this list with you also accompanied by some improvements, tips & tricks that can help to decrease time to market for your products. Despite it is Sitecore-focused, there are more-less generic recommendations at the bottom as well.

Content

1. Sitecore Productivity
   
   • Use XM topology
   • Disabling EXM
   • Unwanted languages
   • Notification table
   • Links Database
   • Publishing
   • Virtual Machines and Docker
   • Fighting cold starts
   • Content Editor performance
2. Software
   
   • Visual Studio
   • Total Commander
   • PowerShell
   • Other great tools
   • Source control
   • Browser
3. Hardware
   
   • My working system
4. Organizational
   
   • Approaching technical debt
   • General productivity thoughts

1. SITECORE PRODUCTIVITY

If you are on Sitecore 9.1 and onwards - use XM topology for starting up, prototyping and coming up with an early PoC or MVP. XM topology is now shipped with all analytics configs, DLLs and the rest of unused stuff physically cut out of the provisioned system, resulting in quicker operational times. I am assuming you very unlikely need analytics features at early development stages. however, please be aware of personalization limitations.

If you are using XP - you may disable EXM unless you develop for it:

<appSettings>
    <add key="exmEnabled:define" value="no"/>
</appSettings>

Use the trick with cutting out unwanted languages from core database. Do you really need all these languages for Sitecore interface? The way Sitecore is built is that it uses Sitecore items for translating itself. That create unwanted loops and leads to unwanted performance losses. The items for deletion are:

/sitecore/system/Languages/da
/sitecore/system/Languages/de-DE
/sitecore/system/Languages/ja-JP
/sitecore/system/Languages/zh-CN

Be careful to avoid deletion of English by mistake as you'll then have to reinstall your Sitecore instance. By default these language items are greyed-out and the message says "You cannot edit this item because it is protected" so you need to "Unprotect it" firstly. You'll be double-asked for confirmation:

Agree OK and that's it! Of course, you can do things quicker and unobtrusively from Sitecore PowerShell:

cd core:/
Get-Item "/sitecore/system/Languages/da" | remove-item

Get rid of Notification table. This table is known for supporting clones, but you don't need it unless you're actually using this functionality. In that case, you can at least remove it from if from web database, as clones only work at CM helping editors with managing numerous clones of the specific item being modified, before getting published. Also, there was an alternative solution coming from Sitecore KnowledgeBase.

Anyway, disabling clones is as simple as the setting:

<setting name="ItemCloning.Enabled" value="false"/>

Disable Links Database (it's a table, in fact) on CD database. It is normally used for identifying links between items, but there's no need of it on web database where items turn into URLs.

Publishing productivity tips:

• May sound obvious but publish only what you've changed and rebuild only what you need.
• Consider using Publishing Service - that's really quick and saves batches directly into the database.
• Or even better - build it encapsulated in Docker or at least VM so that you can just have it referenced from any of your dev-environment.
• Running Sitecore in Live Mode vs. default publishing mode may successfully save some time on publishing for you. For those lucky who are developing with SXA it is way more simple to switch for using both master database and index: just select master from Database field of /sitecore/content/Tenant/Site/Settings/Site Grouping/Site item - no rebuild, restart or "re-whatever" is required.

Use Virtual Machines with snapshots and/or Docker. You may consider a nice triple-combo of Hyper-V - Vagrant - Boxstarter. Wisely configuring and using allows save plenty of time on switching between VMs, restoring, experimenting the code - in other words changing the state, which in Sitecore world is an expensive operation. You may also run an entire farm of VMs configured together, and also (partly or entirely) remotely. Microsoft even gives totally free Hyper-V Server to manage your VMs.

As for Docker, нou may use it as a non-production-unit-of-deployment - it can save plenty of time for some cases for example when working Sitecore-agnostic but very good front-end developers on non-JSS website; I want them, however, be able to have their copy temp of Sitecore instance without all the mess of setup and which they "cannot break".

Fight instance cold starts that happen after you change config or DLLs! There are several things you may do in order to improve your development environment:

• Consider switching <compilation optimizeCompilations="true"> but before make sure you understand what is dynamic ASP.NET compilation and how it works. This is the biggest save for cold starts.
• Tune up your prefetch cache for master database to the minimum
• Disable content testing Sitecore.ContentTesting.config as
• Not a silver bullet, but when starting a new project, why not consider working with SXA or even better - with JSS? While the first several times reduces the number of cold starts, the second eliminates them entirely!
• Reduce time ListManagement agent to run every hour rather than every 10 seconds, used for EXM mostly:

  <scheduling>
      <agent type="Sitecore.ListManagement.Operations.UpdateListOperationsAgent, Sitecore.ListManagement">
          <patch:attribute name="interval">01:00:00</patch:attribute>
      </agent>
  </scheduling>

• Do the same frequency change for IndexingStateSwitcher - from 10 seconds to, let's say, 1 hour:

  <scheduling>
      <agent type="Sitecore.ContentSearch.SolrProvider.Agents.IndexingStateSwitcher, Sitecore.ContentSearch.SolrProvider">
          <patch:attribute name="interval">01:00:00</patch:attribute>
      </agent>
  </scheduling>

• Also, turn off rebuilding indexes automatically:

  <scheduling>
      <agent name="Core_Database_Agent">
          <patch:attribute name="interval">00:00:00</patch:attribute>
      </agent>
      <agent name="Master_Database_Agent">
          <patch:attribute name="interval">00:00:00</patch:attribute>
      </agent>
  </scheduling>

Content Editor

2. SOFTWARE

It's important to have some sort of master productivity tool. For example, I am using Total Commander, that far more than just a great two-panel file manager - I made it as a power pack so it includes:

• Diff tool (I configured to use Beyond Compare with Total Commander) but it comes with a free and fair built-in tool as well
• Built-in FTP client with encrypted password storage
• Hotkeys on almost everything you can do.
• Rapid access to the most important folder you define (and yes - hotkeys for that)
• true and reliable search by content, regex, ... and also in archives (ex compare with windows)
• Ooverride system file associations and assign your tool of choice, with parameters
• I integrated TeraCopy into Commander, so that I have the best and fastest copying tool as well
• I also integrated PowerShell console into TC so helps a lot save much time opening it in the right context
• plenty of plugins and much more useful stuff that I struggle to remember at the moment

Since I've just mentioned PowerShell, nowadays it helps you automating almost everything. This includes:

• managing Windows server, all its dependencies, all types of activities on a filesystem, registry, MMC, etc.
• installing, modifying, deploying Sitecore and all the dependencies
• building images, starting, stopping, deploying containers
• do all that same with Hyper-V virtual machines (and I assume, that also should be possible with VMs from other vendors)
• all the type of management and configuration with IIS and SQL

Other tools that significantly save my day-to-day life:

Source control

cinst sourcetree
cinst tortoisegit
cinst git.install

git config --global alias.lga "log --decorate --graph --oneline --all"

Browser today is still our primary target application, so it's also a point of tuning the productivity:

I previously wrote a post about having a rendering variant holding an inline JavaScript one might need along adding some basic JS functionality into your components. 

This is useful when you're early developing your pages and have no possibility or capacity of recompiling entire frontend and updating Creative Exchange package into your solution because of adding/changing few lines; however, given approach is not SEO-friendly as search engines penalize sites for excessive inline scripts and styles. So use it considering to be technical debt, that should be addressed prior to going to production.

The very minimal change one can do is to replace the inline script with a reference to that same script stored in Media Library - same that SXA does itself with themes. This blog post below reveals an approach:

Firstly, create a template:

Then reference given template IDs within Constants.cs file:

using Sitecore.Data;

namespace Platform.Foundation.Variants.Pipelines.VariantFields.ScriptReferenceTag
{
    public static partial class Constants
    {
        public static partial class RenderingVariants
        {
            public static partial class Templates
            {
                public static ID ScriptReferenceTag { get; } = new ID("{0EC036D7-384D-4CF6-AD1F-FE949E96126A}");
            }

            public static partial class Fields
            {
                public static class ScriptReferenceTag
                {
                    public static ID ScriptMedia { get; } = new ID("{F1497AF9-7DD3-4B38-BE22-5F092007F929}");
                }
            }
        }
    }
}

using Sitecore.Data.Items;
using Sitecore.XA.Foundation.RenderingVariants.Fields;

namespace Platform.Foundation.Variants.Pipelines.VariantFields.ScriptReferenceTag
{
    public class VariantScriptReferenceTag : RenderingVariantFieldBase
    {
        public string ScriptMedia { get; set; }

        public VariantScriptReferenceTag(Item variantItem) : base(variantItem)
        {
        }
    }
}

using Sitecore.Data;
using Sitecore.XA.Foundation.Variants.Abstractions.Pipelines.ParseVariantFields;

namespace Platform.Foundation.Variants.Pipelines.VariantFields.ScriptReferenceTag
{
    public class ParseScriptReferenceTag : ParseVariantFieldProcessor
    {
        public override ID SupportedTemplateId =>  Constants.RenderingVariants.Templates.ScriptReferenceTag;
        
        public override void TranslateField(ParseVariantFieldArgs args)
        {
            ParseVariantFieldArgs variantFieldArgs = args;

            var variantHtmlTag = new VariantScriptReferenceTag(args.VariantItem) { Tag = "script" };
            variantHtmlTag.ScriptMedia = args.VariantItem[Constants.RenderingVariants.Fields.ScriptReferenceTag.ScriptMedia];
            variantFieldArgs.TranslatedField = variantHtmlTag;
        }
    }
}

using System;
using Sitecore.Data;
using System.Web.UI.HtmlControls;
using Sitecore.XA.Foundation.RenderingVariants.Pipelines.RenderVariantField;
using Sitecore.XA.Foundation.Variants.Abstractions.Pipelines.RenderVariantField;
using Sitecore.Resources.Media;

namespace Platform.Foundation.Variants.Pipelines.VariantFields.ScriptReferenceTag
{
    public class RenderScriptReferenceTag : RenderVariantField
    {
        public override Type SupportedType => typeof(VariantScriptReferenceTag);

        public override void RenderField(RenderVariantFieldArgs args)
        {
            var variantField = args.VariantField as VariantScriptReferenceTag;
            if (variantField != null)
            {
                var id = variantField?.ScriptMedia;
                if (string.IsNullOrWhiteSpace(id))
                {
                    return;
                }

                var scriptItem = Context.Database.GetItem(new ID(id));
                if(scriptItem == null)
                {
                    return;
                }

                var url = MediaManager.GetMediaUrl(scriptItem);

                var tag = new HtmlGenericControl(variantField.Tag);
                tag.Attributes.Add("type", "text/javascript");
                tag.Attributes.Add("defer", String.Empty);
                tag.Attributes.Add("src", url);

                args.ResultControl = tag;
                args.Result = RenderControl(args.ResultControl);
            }
        }
    }
}

Example of usage:

This rendering variant field generates the following output:

<script src="/-/media/Project/Platform/Other/Scripts/Header-script.js" type="text/javascript" defer="" ></script>

This approach works perfectly well. But once again for a second, have you ever considered moving such scripts into a Theme along with related component (if any) instead of leaving it like that? Hope this helps!

Previously I wrote about adding Unicorn to Sitecore LaunchPad to be available for admin users. This time I am adding one more tool icon - Show Config that leads admin users to the page where they can see patched and merged configuration that is used by given Sitecore instance

No more need to remember and manually type https://instance.hostname/sitecore/admin/showConfig.aspx URL in order to access it.

You may download the installation package at the bottom of this blog post.

The actual button item is located in core database under the following path: /sitecore/client/Applications/Launchpad/PageSettings/Buttons/Tools/Show Config.

Download ready to use package (19.7kb, keep in mind that Unicorn icon will be shown to admin users only).

One day I was working on a page that had way too many Content Editor section opened, plus those plenty coming out from Standard fields also added frustration. I thought it would be great to have Collapse All button implemented, that closes all the sections in order to help navigation. 

I went checking the way Content Editor works this out and later wrote a JavaScript snippet, that implements and wires desired functionality. I also added Expand All button bringing the reverse behavior. Here's the code:

scContentEditor.prototype.onDomReady = function (evt) {
    this.addCollapser(window.jQuery || window.$sc);
};
scContentEditor.prototype.addCollapser = function ($) {
    $ = $ || window.jQuery || window.$sc;
    if (!$) { return; }

    $('#EditorTabs').append("<style>.toggler { border: 1px solid #bdbdbd; box-shadow: 0 1px #ffffff inset; cursor: pointer; height: 35px; margin: 16px 1px 0; }</style>");
    $('#EditorTabs').append("<button id='expander' class='toggler'>Expand all</button><button id='collapser' class='toggler'>Collapse all</button>");
    $('#EditorTabs').on("click", "#collapser", function () {

        $('.scEditorSectionCaptionExpanded').each(function () {
            var script = $(this).attr("onclick");
            eval(script);
        });
        return false;
    });
    $('#EditorTabs').on("click", "#expander", function () {

        $('.scEditorSectionCaptionCollapsed').each(function () {
            var script = $(this).attr("onclick");
            eval(script);
        });
        return false;
    });
};

All you need to do is to append to the bottom of <your_web_root>\sitecore\shell\Applications\Content Manager\Content Editor.js and that's it!

For those like me who like automation, I am attaching this JavaScript file, right click this link and save it, and then use PowerShell:

$file2 = Get-Content "ExpanderCollapser.js"
Add-Content "C:\inetpub\wwwroot\<YOUR_WEB_ROOT>\sitecore\shell\Applications\Content Manager\Content Editor.js" $file2

Once done, you'll see the result in Content:

There is also package available for download (compatible with Sitecore 9.0.* - 9.2)

I was working on implementing a subscription model system, where authenticated users visit website with a specific role coming from Identity Server (or, unauthenticated - anonymous, of course), so that I can apply personalization of content, as we normally do.

The difference was, however, that subscription level were logical units, more complicated and not matching IDS roles. They also should be adjustable from Sitecore by business users. That made using personalization by these users type quite complicated, due to complex rules creation, especially those with inverted logic except when. But even with that in mind, I could not simply use personalization for preventing unauthorised users (for example, those registered and logged, but still having insufficient permissions) from accessing specific types of content. The business requirement demands all the pages to be accessible by anyone, but when users don't have required access level - most of content apart from few teasing paragraphs in the beginning, needs to be greyed out by a components encouraging them to increase their subscription level in order to get full access.

So, in order to address these requirements, I decided to implemented a simple role-mapping Subscription Model, something as could be described by this template:

But wait! There is no possibility to use Sitecore security roles in an item!

So I decided to implement the one. After quick googling I came across Mike Reynold's experiments with fields and templates and went similar way on implementing Role Multilist Selector field. 

Once done, core database needs to get a new field type registered - Roles, which is implemented in a way of traditional multi-select field:

So now, I can use it as an ordinary Sitecore item field. Please note, that Source column at first screenshot above contains Domain=ids - that is a set of parameters passed in a format of URL string (UrlString is .NET class that accepts these parameters in the code). I've implemented that as a Sitecore domain filtering parameter, where ids is the domain name.

Now we can select roles - they will be stored in pipe-separated format in given field:

Finally, after implementing a logical layer of Subscription Model, I also had to create custom rules conditions to apply personalization operating these logical subscriptions, but that made business users' life way easier.

Hope this helps!

Image Link rendering variant field

This is quite powerful and at the same time very simple rendering variant field - it nicely renders <img> tag surrounded with <a> anchor tag without any of other unwanted wrappings normally coming when nesting components in SXA, as below:

<a href="http://link.to/internal-or-external-item" class="individual-class-for-anchor">
    <img src="/-/Media-item-from-sitecore" class="individual-class-for-image"/>
</a>

When starting working with SXA I faced some lack of good guidance and walkthrough (with an exception of excellent series by Adam Najmanowicz). Today I am going to eliminate this gap by adding my own walkthrough of implementing a footer, what could be simpler, I thought. After reading some official tutorials I expected it to be an exercise of dropping structure components (ie. splitters with rows and columns) and assigning link lists into them so that it could be later styled by front-end team. Wrong! Things appeared to be not as easy.

To start with, my strict front-end team came to me with quite a precise requirement for a footer, and below is they demand from me.

Requirements

They send me an image with all the blocks assigned:

It was accompanied by the HTML output expected for me to achieve:

Minimal and effective, nice job, front-enders! Now the ball is on my side in order to get all that implemented.

Implementation

1. As per Sitecore SXA recommendations, I create Footer partial design and open it in Experience Editor for editing. Footer component will be created for that partial design, which itself will be used to construct a resulting page.

2. Not to mess with any OOB components, I create Footer rendering by cloning one of the existing renderings with datasource, so that I ensured a copy of datasource template and folder created. Also make sure this rendering stays outside of Experience Accelerator folder, but on feature layer where serialization enabled.

3. Assign new rendering to Available Renderings (/sitecore/content/Tenant/Site/Presentation/Available Renderings/Module) so that it appears in my custom components sectionand also in Toolbox

4. Now it is a good time to adjust a Template for Footer,it was created automatically by cloning, but this is how I defined it:

What is important to explain here - Elements field is pointing to Footer Elements Folder. At a first glance, that folder contains just a collection of link lists. But that's not right! There are at least 3 types of footer blocks: first four blocks are indeed Link List blocks,however,followed by Rich Text blocks, and finally, there is Social Presence block and all three can be added intoFooter Elements Folder. These three types are defined along with Footer template, you may see on the screenshots below.

5. Elements templates. This is how I defined them, Link ist footer element:

Rich text footer element:

Social presence footer element:

6. Insert Options to be configured for Footer folder to accept both Footer items and Footer Elements Folder. Configure Footer Elements Folder to accept these 3 types. Now one can insert the data. Once done, that will be how the data folder looks like:

First 4 items (About Us, Hot topics, Other sites, Help and Support) reference corresponding Link Lists as defined as below:

Next two items point to reusable Rich Text items under /Data folder. And the last one is a reference to a Social Presence I have implemented previously.

After items of all three types are created, we can assign them into footer datasource item itself:

7. Rendering variants. In the previous steps, we have defined rendering, templates, folders and actual data. Now it is time to make it all together work to produce the output by creating rendering variants. This is the most tricky part of the current walkthrough

Default is the main and only rendering variant to be called or footer. Other three variants are "service" variants and designed to be called from Defaultinternally in a loop, being assigned to a component rendering field with personalisation applied (see image above).

Here's how these three other variants look like:

Link list footer element switches to a referenced Link List item and uses Query variant field to iterate its children.

Rich Text footer element simply references to a Rich Text (reusable) item under /Data folder in the same manner.

As for Social presence footer element rendering variant, it defines a component that references Social buttons rendering with Social presence rendering variant, that I have described in one of my previous posts.

Lastly, footer-info__copyright field renders copyright lines at the very bottom.

8. In order to avoid confusion for your editors, it makes sense to restrict rendering variant by selecting Allowed in templates field leaving only Default variant since other three variants are internally called from Default using Component variant field. 

9. Apply new rendering to a Partial Design it in Experience Editor. Do not forget to configure footer placeholder to accept only Footer component by creating a placeholder setting,

Result

Save the page and enjoy the result:

No single line of back-end code!

When dealing with a rendering variant field, it is not a big deal to set few data-attributes to it - those inputs are located at the very bottom of Variant Details section. You can do it like that:

But what if you need data attributes to the top level of component, which it Rendering Variant item itself? There isn't such an option!

Requirements are

1. An id attribute (ie. section-1, section-2, ... section-N)
2. One or many data-attributes (ie. " User-friendly title", "Another user-friendly title", etc.)
3. CSS class section-with-anchor on those instances, which have both previous requirements implemented

Solution

public static MvcHtmlString RenderAllDataAttributes(this HtmlHelper helper)
{
    var rendering = Sitecore.Mvc.Presentation.RenderingContext.Current.Rendering;

    string additionalAttributes = String.Empty;
    if (rendering?.Parameters != null)
    { 
        foreach (KeyValuePair<string, string> parameter in rendering.Parameters)
        {
            if (parameter.Key.ToLower().StartsWith("data-"))
            {
                additionalAttributes += $"{parameter.Key}=\'{parameter.Value}\' ";
            }
        }
    }

    return new MvcHtmlString(additionalAttributes.Trim());
}

<div @Html.RenderAllDataAttributes() @Html.Sxa().Component(Model.Rendering.RenderingCssClass ?? "default-class", Model.Attributes)>
    <div class="component-content">
        ...
    </div>
</div>

Result

Finally, I got it all rendered as expected:

Yet another rendering variant field came to my to-do list for implementation - Script rendering variant field. Why would I need one at all? 

I came across two uses cases where implementing this field type made my job done and that's just recently. Some developers have (reasonable) biases against having JavaScript code inline instead of referencing JS-file at the bottom of a page, and I do understand their point - Google may penalise your site for overusing inline JS. So use it responsibly, don't overuse. Keeping in mind that given field comes up as a part of a component dynamically added to a page - there isn't a big choice on how to extend running website with some additional client-side functionality. I am presenting both cases below, let's take a look at them. Would you know the better way of achieving these goals - please let me know via Slack or Twitter. Also, the code of Script Variant Field is at the very bottom of the page.

Use case 1: implementing in-page navigation panel

As usual, I got a precise requirement from my strict front-end team to implement such a piece of code as a component. It has a UL-tag that will keep a link list to other components of this same page (prefixed with #) created client-side dynamically, and a script that does the actual job. When rendering the in-page navigation component by the backend, we're now aware of other components and their attributes, so the walk-around was handling page loaded event and identifying all the components that have IDs set and class section-with-anchor.

<div class="component content col-12 content-section">
    <div class="component-content">
        <div class="anchor-panel">
            <ul class="anchor-panel__list"></ul>
            <script defer>
                document.addEventListener('DOMContentLoaded', function(){
                    if (!$('.section-with-anchor').length || !$('.section-with-anchor').length) return;
                    $('.section-with-anchor').each(function(index, el) {
                        var anchor = '#' + $(el).attr('id');
                        var text = $(el).attr('data-text');
                        var $anchorList = $('.anchor-panel__list');
                        var $anchorItem = $('<li class="anchor-panel__item"></li>');
                        var $anchorLink = $('<a href=""></a>')
                        $anchorItem.append($anchorLink.text(text).attr('href', anchor));
                        $anchorList.append($anchorItem);
                    });
                });
            </script>
        </div>
    </div>
</div>

Quite obviously, my rendering variant will contain 2 fields: a UL-tag section field for in-page navigation that takes the links dynamically and script variant field, containing the client-side logic. You can test this script in action by this link. That is how it looks in Content Editor:

Since I have 5 other sections qualifying script requirements - they all have been identified by the script and added to in-page navigation panel. Here's how the result looks like on a styled page for me:

Once again, I decided to implement a new script variant reference field because the component is subject to some seldom minor JavaScript changes, but should be configurable. Also, at given use case it can be dropped only once into a page, so there's no problem with multiple instances of the same script for me, but using it you might need to check if that applies for your scenarios. But even with that in mind, I'd probably not bothered creating yet another new rendering variant field, if not few more potential usages I have in backlog.

Use case 2: accessing client side URL hash-key parameter and presenting it on a page

Recently I implemented a URL query string parameter variant field, where one can set any parameter and the field present its URL-decoded value on a page in any given tag and style. A colleague of mine who develops search results page with SXA asked if that's doable to extract a hash-key URL parameter and show it on a page along with other content, however that a fully client-side parameter that never got posted to the server.

I decided to give a try and managed to confirm with a small proof of concept. I quickly wrote this script (so please be forgiving for it being just a quick PoC and me not being the proper FED).

window.addEventListener("hashchange", function () {
    var h1 = document.getElementsByClassName("updatedHashValue");
    if (h1.length > 0) {
        h1[0].innerHTML = getHashValue("param");
    }
    function getHashValue(parameter) {
        var hashValues = window.location.hash.substr(1);
        var result = hashValues.split('&').reduce(function (result, item) {
                var parts = item.split('=');
                result[parts[0]] = parts[1];
                return result;
        }, {});
        return result[parameter];
    };
});

That's how it looks implemented as the part of rendering variant. It creates an H1 element to store the value, extracted out of hash parameters stored at URL bar and updated without any postback to the server, and of source script variant field:

Testing. After adding a component to a page, saving and selecting the above rendering variant, I got the page reloaded as anticipated with no visual changes. Then I open a console from browser dev.tools and enter:

window.location.hash = "param=Successful!"

From dev.tools that was easy to confirm that there was no postback done to the server, however, browser navigation bar predictably changed, appending new hash parameters pair: 

And guess what? H1 element immediately got that value displayed. Love that magic!

The code is very simple, similar to other variant fields I've blogged previously. Create the model and implement property for a field:

public class VariantScript : VariantField
{
    public string Script { get; set; }
}

Reference ID of that field from a template and ID of a template itself:

public static class Constants
{
    public static class RenderingVariants
    {
        public static class Templates
        {
            public static ID Script = new ID("...");
        }
        public static class Fields
        {
            public static class Script
            {
                public static ID ScriptField { get; } = new ID("...");
            }
        }
    }
}

Here's a parser

public class ParseScript : ParseField
{
    public override ID SupportedTemplateId => Constants.RenderingVariants.Templates.Script;

    public override void TranslateField(ParseVariantFieldArgs args)
    {
        ParseVariantFieldArgs variantFieldArgs = args;

        variantFieldArgs.TranslatedField = new VariantScript
        {
            Script = args.VariantItem[Constants.RenderingVariants.Fields.Script.ScriptField]
        };
    }
}

and a renderer:

public class RenderScript : RenderVariantField
{
    public override Type SupportedType => typeof(VariantScript);

    public override RendererMode RendererMode => RendererMode.Html;

    public override void RenderField(RenderVariantFieldArgs args)
    {
        var variantField = args.VariantField as VariantScript;
        if (variantField != null)
        {
            args.ResultControl = RenderScriptField(variantField, args);
            args.Result = RenderControl(args.ResultControl);
        }
    }

    protected virtual Control RenderScriptField(VariantScript variantScript, RenderVariantFieldArgs args)
    {
        if (!string.IsNullOrWhiteSpace(variantScript.Script))
        {
            var tag = new HtmlGenericControl("script") { InnerHtml = variantScript.Script };
            tag.Attributes.Add("defer", String.Empty);
            return tag;
        }

        return new LiteralControl();
    }
}

New Script Rendering Variant field, that enhances your tooling for implementing more modern-looking websites. And as usual - please use it responsibly!