Sunday, 27 November 2016

Force CLI Part 5 - Node Wrapper

Force CLI Part 5 - Node Wrapper



In earlier posts in this series I shared the odd example bash script that carried out some combination of commands in conjunction with the Force CLI to achieve things like creating a zip backup of Salesforce metadata. The reason I went with bash scripts is that I know the syntax really well and was able to quickly build some tools for managing deployment of our BrightMedia Full Force solution. In essence it allowed us to get big quick.

However, there are a couple of downsides to this approach:

  • Not that many people know bash these days - I spent many years building systems on Linux so I’m pretty familiar with it, but I didn’t want to be a bottleneck
  • Bash isn’t the easiest tool to process XML or JSON format data. You either end up running the files through additional commands such as XMLStarlet, which gives you a whole new syntax/configuration to learn, or you start parsing it using tools like sed and awk, which works for the basics but quickly becomes unusable with complex data.
I’m pushing most of our people towards JavaScript now, with the advent of lightning components, so JavaScript running on the Node platform seemed like the obvious choice to rewrite the scripts in. I then found out that the CLI for SalesforceDX will be built in JavaScript on Node, and will have a plugin architecture to allow us to extend it, which more than justified my choice. 

JSON takes Salesforce Data

While my code behind the tools is now a lot more flexible and easier to maintain, the key use case I’ve found is when extracting Salesforce record data for processing - e.g. carrying out a data fix as part of a release. The Force CLI can output the results of a query formatted as a JSON string, and JavaScript can parse this into an object using the built in JSON.parse() function.

The following example reflects a real-world scenario I encountered a few weeks ago. I was bitten by a known issue with the Winter 17 release of Salesforce, which meant that I couldn’t login to the Lightning Experience. I needed to flip the users in this org to the classic interface, which I started doing via the data loader before creating a script to do the same thing with the Force CLI wrapped in a Node script.

Getting Started

First, download and install Node from the NodeJS org website.

Then create a new directory for the project - in this example a subdirectory named cli.

  • mkdir cli
  • cd cli
Next, create a Node package file for the project:
  • npm init

Accept the defaults for everything and add a description if you so desire.

Note that you don’t strictly need to do this, for the simple example in this post, but it’s a really good habit to get into for when you need to include dependencies on other packages.

Show Me the Code

Save the following code into a file named index.js :

var child_process=require('child_process');

var instance=process.argv[2];
if (instance===undefined) {
var loginOpts=['login', '-i=' + instance];
child_process.execFileSync('force', loginOpts);

// extract the user records as JSON
var queryOpts=['query', 'select id, Username, UserPreferencesLightningExperiencePreferred from User',

var uJSON=child_process.execFileSync('force', queryOpts);

// turn into a JS object
var users=JSON.parse(uJSON);

// dump the users out to the console
for (var idx=0, len=users.length; idx<len; idx++) {
    var user=users[idx];
    console.log('User = ' + JSON.stringify(user, null, 4));

 Executing this produces the following (edited) output: 

kbowden@Keirs-MacBook-Air cli $ node index.js
User = {
    "Id": "00524000003vqIgAAI",
    "UserPreferencesLightningExperiencePreferred": true,
    "Username": "",
    "attributes": {
        "type": "User",
        "url": "/services/data/v36.0/sobjects/User/00524000003vqIgAAI"
User = {
    "Id": "00524000000FctAAAS",
    "UserPreferencesLightningExperiencePreferred": true,
    "Username": "",
    "attributes": {
        "type": "User",
        "url": "/services/data/v36.0/sobjects/User/00524000000FctAAAS"

So I can see that a number of users are setup for the Lightning Experience.

Explain Yourself

Breaking down the code a little, the first line loads the child_process module, which allows a script to spawn other processes:

var child_process=require('child_process');

The next lines figure out the correct Salesforce instance to connect to - the default is login, but you can override this by passing an additional argument after index.js, such as test to connect to the sandbox login endpoint. Once this is done, a child process is spawned to execute the login. 

var instance=process.argv[2];
if (instance===undefined) {
var loginOpts=['login', '-i=' + instance];
child_process.execFileSync('force', loginOpts);

Note that the chid process function takes two parameters - the first is the command to execute (force in this case) and the second an array of parameters to pass to the command. Note also that I use the execFileSync variant to execute the command synchronously and return the output as the result which I promptly ignore!

Next, a query is executed via the Force CLI to extract details of all users in JSON format:

// extract the user records as JSON
var queryOpts=['query', 'select id, Username, UserPreferencesLightningExperiencePreferred from User',

var uJSON=child_process.execFileSync('force', queryOpts);

and this is converted to an array of JavaScript objects:

var users=JSON.parse(uJSON);

Updating the Users

Once I verified that I could extract the users successfully, I could then update each user to use Salesforce Classic, by replacing the final loop with the following code which executes the Force CLI record update command for each returned user:

for (var idx=0, len=users.length; idx<len; idx++) {
    var user=users[idx];
    if (user.UserPreferencesLightningExperiencePreferred) {
        console.log('Reverting user ' + user.Username + ' to classic');
        var updateOpts=['record', 'update', 'user', user.Id,
        child_process.execFileSync('force', updateOpts);

Running the revised script gives the following (edited) output:

kbowden@Keirs-MacBook-Air cli $ node index.js login
Reverting user to classic Reverting user to classic

Logging in after executing this script confirms that the users have been updated correctly as I am taken straight to the classic UI rather than a sad snowman with an error message! 

Related Posts



Thursday, 10 November 2016

Lightning Component Actions with Signature Capture Part 2

Lightning Component Actions with Signature Capture Part 2


In the first part of this blog I covered how to set up a quick action using the Lightning Component Action in the Signature Capture package. In this post I’ll show how to create a custom lightning component action that wraps the Signature Capture component and allows overriding of the default text and labels.


The following list shows the Signature Capture message/label attributes and their default values:

  • startMsg - “Click the ‘Capture Signature’ button to begin 
  • enterMsg - “Sign here please"
  • completeMsg - “Here is the captured signature"
  • captureButtonLabel - “Capture Signature"
  • saveButtonLabel - “Save”     (new in Winter 17)
  • clearButtonLabel - “Clear”    (new in Winter 17)

Text Literals

The packaged component action, SignatureCaptureAction, leaves the default attribute values as-is. A custom action, on the other hand, allows some or all of these to be overridden with different text literals:

<BGSIGCAP:SignatureCapture recordId="{!v.recordId}"
          startMsg="Let's Go - Click the button to start” />


Screen Shot 2016 11 10 at 08 11 42

Custom Labels

Attributes may also be defined as custom labels, which have the added advantage of automatically picking up the correct translation for the user:

Screen Shot 2016 11 10 at 08 16 03

<BGSIGCAP:SignatureCapture recordId="{!v.recordId}"
         saveButtonLabel="{!$Label.c.SigCapSave}" />

Screen Shot 2016 11 10 at 08 25 01


Example Custom Action

Here’s a couple of screenshots from an example custom component with all text content overridden :

Screen Shot 2016 11 10 at 08 33 41

Screen Shot 2016 11 10 at 08 33 53

Screen Shot 2016 11 10 at 08 34 03

The source for this can be found in the Signature Capture Samples GITHUB repository.

Related Posts


Monday, 7 November 2016

Lightning Component Actions with Signature Capture

Lightning Component Actions with Signature Capture

Screen Shot 2016 11 07 at 18 36 19


As some readers of this blog may be aware, a year or so ago I wrote a Lightning Component for the Component Exchange that was soon to be launched at Dreamforce. The component is called Signature Capture and unsurprisingly allows a user to capture a signature image and attach it to a Salesforce record. Up until the Winter 17 release of Salesforce, the only way to associate this with a record declaratively was to use the Lightning App Builder to include it in a custom record home page. This works fine but is a slightly odd user experience, as the component is always present even after the user has captured the signature. This all changed once Winter 17 was live thanks to Lightning Component Actions.

Lightning Component Actions

As the name suggests, Lightning Component Actions are custom actions that invoke a lightning component. Many lightning components are called, but few are chosen for custom actions. Specifically those that implement one of two interfaces:

  • force:LightningQuickAction
  • force:LightningQuickActionWithoutHeader

The difference being that if you implement force:LightningQuickActionWithoutHeader the entire user interface is down to you.

You can use Lightning Component Actions in Salesforce1 and the Lightning Experience.


The latest version of Signature Capture package contains a simple component that implements the force:LightningQuickAction interface and includes the Signature Capture component itself. To create a quick action that uses this is a matter of a few clicks. 

Setting Up

First choose which sObject type you want to associate the quick action with - in the screenshots below I’m using account.

Navigate to the Buttons, Links, and Actions section of the setup for the sObject and choose New Action. Select ‘Lightning Component’ as the action type and ‘BGSIGCAP:SignatureCaptureAction' as the Lightning Component. I find that a height of 500px works best, but your mileage may vary. Then enter a Label and Name and click the Save button:

Screen Shot 2016 11 07 at 18 14 31

Once the quick action is defined, it must be added to the page layout in the Salesforce1 and Lightning Experience Actions section (if you haven’t added any quick actions to this section before you’ll need to override the default via the link provided):

Screen Shot 2016 11 07 at 18 24 09

Take it for a Spin

After setup is complete, navigate to any account record in the Lightning Experience and you will see the quick action in the set of buttons on the top right:

Screen Shot 2016 11 07 at 18 26 04

Clicking the button opens the action in its own dialog, allowing a signature to be drawn via mouse, trackpad etc:

Screen Shot 2016 11 07 at 18 27 45

To my mind this is a much better user experience, as the component only appears on demand, so there is no confusion around a component that is always present. The slight downside is that once the signature has been captured the user still has to close the dialog via the standard Cancel button, but that’s a minor drawback to my mind.

Mobility Included

Adding the quick action to the page layout also makes it available to the Salesforce1 application on mobile devices. A word of warning around this - the locker service currently doesn’t surface touch events to a component, so if you enable the locker service the Signature Capture component won’t work on tablets or phones.

Repo Man

The Signature Capture component has its own github repo with sample apps (or app at the time of writing!). This is also where feature requests are tracked, so if there’s anything you’d like to see added, let me know and I’ll see if it can be accommodated. Just to set expectation, this is something that I work on mainly in my spare time (I’m currently on vacation, hence having time for this post) so I may take a while to respond.

Related Posts




Saturday, 29 October 2016

Lightning Components - New Tab, New You

Lightning Components - New Tab, New You



This is something I’ve been meaning to blog about for a while now, but things like Dreamforce, preparing for Dreamforce and the Winter 17 release kept getting in the way. I didn’t get as far as setting up a nice example, due to the sheer volume of code involved, so I haven’t been back and tried it out recently. Therefore it’s possible that you’ll never have the same experience I did, but as it’s not exactly onerous it seemed worth putting it out there anyway. Your mileage may vary.

Signature move

While building out the MVP for our BrightMedia Lightning booking page to demonstrate at Dreamforce, I needed to change the signature of one of my Apex controller methods, so to get a simple example from something like:

public static Account GetAccount(String idStr)


public static Account GetAccount(String idStr, String name)

and changed the parameters that the Lightning Component passed on the action from:

var params={idStr:'0018000001QtQVA', name:'Blog Post Account'};

That went well!

Once I’d made the various changes and saved all the artefacts, I refreshed the page and starting seeing some unexpected behaviour. The name parameter came through as null regardless of what I set it to. Debug statements in the lightning component showed that the name had a value (it wasn’t a hardcoded string unlike the example above!), but some time after I added it to the action it went bad. Making it a hardcoded string didn’t help, nor did juking around the other parameters. I even tried adding more parameters that definitely didn’t exist server side, but nothing helped.

After a some time and about and thousand debug statements later, I found that while I might be adding the new parameter to my params object, when that was sent to the server it had disappeared, hence the server always reported null.

Cache is king

I then tried what I should have as soon as this manifested itself, the browser equivalent of turning it off and on again. I closed the browser tab and opened a new one. Lo and behold everything worked as expected. We all know that the lightning framework aggressively caches on the front end to improve performance, and it looks like the client retained a metadata view of the server side signature after I changed it.

Now I need to stress that for once I’m not complaining about the behaviour of the framework! This is exactly the kind of thing that I’d expect when caching metadata. It’s not always possible to check everything in a cache is still valid - you quickly start to lose the benefit if you go back and check the source on every request. 

The key takeaway here is that when your application involves caching and you change something which doesn’t work as you expected, open a new tab (or window), or even a new browser. It only takes a few seconds and will save time and heartache in the long run.

Related posts


Friday, 14 October 2016

Visualforce Development Cookbook Second Edition

Visualforce Development Cookbook Second Edition



As regular readers of this blog will know, I wrote a book three years ago - the Visualforce Development Cookbook - and I really haven't stopped banging on about it since. Well things just got worse as I finished the second edition in ?September? and I'm obviously keen to shift as many copies as possible!

What’s a Second Edition?

A second edition isn't a new book. Instead it's an iteration on the previous version, minus content that is no longer relevant and including new features, so the first job is to identify what should be cut and what should be added.

One of the nice things about working on Salesforceis that over time new features are added that remove the need for custom code, so I was able to chop a few recipes out that are now handled natively by the platform. The next job is to update all existing code to the latest API version - something that typically happens several times as Salesforce produces releases faster than I can write a book (even a second edition!). 

Then its back to the writing - for existing recipes making sure that the text is still appropriate and creating new recipes from scratch. This was pretty compressed this time around, as Packt proposed a pretty aggressive timescale to complete before Dreamforce which I was more than happy to sign up to, as I knew that would take up all of my time through to mid-October. I wouldn't recommend this if you don't have some vacation time you can dedicate to writing though, as it would have been pretty tough to complete with only evenings and weekends available.

What’s changed?

The main changes in the second edition are:

  • The chapter on sites has been reworked to use the Salesforce Lightning Design System
  • The chapter on jQueryMobile has been replaced with Salesforce1 recipes.
  • A new chapter on Troubleshooting has been added, covering debugging, managing the view state and advice on avoiding common problems.
  • All the retained recipes have been revised as appropriate.

You want me to buy it again?

Now I’m not like the music industry that expects you to buy the same  music every time they think up a different format, and I can understand that for someone who already purchased the first edition, this might not represent enough of a change to justify buying the second edition (although it was for Fabrice Cathala, who is clearly going for the full set - thanks Fabrice!).

I raised this with the publishers and they gave me a couple of discount codes to use, both valid until 17th October, which you can use on the Packt website:

  • Use code VISDSE50 for 50% off the Ebook
  • Use code VISDSE15 for 15% off the print book

Note that you don't need to have purchased the first edition to use these codes, you can simply use them to scoop up a bargain.

Show me the code!

For those who just need the code and no explanation, its available free of charge on github at:

A word of advice though - best not come to me asking for help around adapting the recipes or producing unit tests if you've decided not to purchase the book - that would be asking me to cannibalize my own sales!

Related Posts


Friday, 23 September 2016

Force CLI Part 4 - Deploying Metadata

Force CLI Part 4 - Deploying Metadata



In previous posts in this series I’ve covered Getting Started, Extracting Metadata and Accessing Data via the Force CLI. In this post I’ll cover the functionality that I use probably 95% of the time that I use the Force CLI - deploying metadata from the local filesystem to a Salesforce instance.

The Building Blocks

Like other tools for deploying to Salesforce (e.g. migration tool, Eclipse IDE) there are a two items required to deploy via the Force CLI

Project Manifest (aka package.xml)

The package.xml file defines the metadata components that will be deployed - you can find out more about this in the Salesforce Metadata API Developer’s Guide

Metadata Components

The metadata components need to appear in the directory structure expected by the metadata API - e.g. Apex classes in a directory named ‘classes’, Visualforce pages in a directory named ‘pages'

I don’t make the rules (but I do make the deployment directory)

The rule when deploying is that the components specified in the package.xml manifest file need to match those present in the metadata components directory structure. If you specify components that are missing you will get an error, and if you provide additional components that aren’t mentioned in the package.xml, you’ll get a different flavour of error.

My metadata comes from GIT, so the metadata components structure of my entire project is defined there. However, I don’t always want to carry out a full deployment so I create a new structure in a temporary directory and create a package.xml file reflecting that. I also use a naming convention for my temporary directory so that I can look back at the order that I did things in - drop_<YYYYMMDDhhmmss>, where YYYYMMDD is the year/month/day of the date and hmmss is the hours/minutes/seconds of the time. The reason I go with this naming convention is that when I use the ls command, my directories are listed in order of creation:

$ ls -l

drwxr-xr-x  6 kbowden  wheel    204 16 Sep 14:39 drop_20160916141152
drwxr-xr-x  6 kbowden  wheel    204 17 Sep 10:11 drop_20160916184125
drwxr-xr-x  6 kbowden  wheel    204 17 Sep 12:45 drop_20160917121232
drwxr-xr-x  6 kbowden  wheel    204 18 Sep 12:56 drop_20160918092957
drwxr-xr-x  6 kbowden  wheel    204 18 Sep 19:21 drop_20160918185148
drwxr-xr-x  6 kbowden  wheel    204 20 Sep 15:39 drop_20160920152837

To use a very simple example, here’s the directory structure to deploy just the User object. My starting directory is /tmp/drop_20160923080432:


 My package.xml file just contains an entry to deploy the objects directly (it’s called CustomObject, but covers both standard and custom):

<?xml version="1.0" encoding="UTF-8"?>
<Package xmlns="">

Deploy all the Metadata

Once the directory structure and manifest are in place, use the force import command to deploy your metadata, using the -d switch to specify the directory containing the metadata:

force import -d /tmp/drop_20160923080432/target

The Force CLI will update you every five seconds on progress

Not done yet: Queued  Will check again in five seconds.
Not done yet: InProgress  Will check again in five seconds.
Not done yet: InProgress  Will check again in five seconds.
Not done yet: InProgress  Will check again in five seconds.
Not done yet: InProgress  Will check again in five seconds.
Not done yet: InProgress  Will check again in five seconds.

Once the deployment is complete, you’ll get details of successes/failures - the default behaviour is simply the to output the number of each

Failures - 0

Successes - 24
Imported from /tmp/drop_20160923080432/target

While if you specify the -v switch, you'll get full details of each success/failure:

Successes - 24
	status: unchanged
	status: unchanged
	status: unchanged


Deploying to Production

As everyone in the world knows, in order to deploy metadata to production you need to run some unit tests. The Force CLI provides support for this via the -l switch, which takes the same options as the migration tool:

  • NoTestRun 
  • RunSpecifiedTests
  • RunLocalTests
  • RunAllTestsInOrg

Verification Only

Another very useful switch is -c. This carries out a verification only deployment and rolls back all changes upon completion. I use this as part of our continuous integration environment to make sure I can successfully deploy to a clean development environment.

Related Posts


Sunday, 28 August 2016

Stand and Deliver with Salesforce Geocoding

Stand and Deliver with Salesforce Geocoding


The Summer 16 release of Salesforce introduced Automatic Geocodes for Addresses which, as the name suggests, allows you to geocode records with addresses automatically. To enable this just go to Setup, type ‘Clean' into the search and click on Clean Rules from the resulting setup nodes.  Then activate one or more of the clean rules and wait a few minutes. Much easier than setting up a batch job to integrate with Google and then finding yourself in breach of the license terms! 

In conjunction with the Geolocation API, this new feature allows you to build some cool applications without too much effort. In this post I’ll show how to combine these two technologies to build an application that assigns a delivery to the nearest user (driver, bike messenger etc) to the collection location.

User Check In

The functionality to allow users to check in (record their location) is provided by a Lightning Component. The component includes the Lightning Design System (yeah, I know it's part of the platform now, but including it myself means that I can also use this component in lightning out) and after the CSS has loaded, uses the geolocation API to capture the current location:

    doInit : function(cmp, ev) {
        var self=this;
        if (navigator.geolocation) {
                function(result) {self.gotGeoCode(result, cmp, self);},
                maximumAge: 0,
                enableHighAccuracy: true

The parameters passed to the geolocation function are as follows:

  • maximumAge - the maximum age of a cached position. Setting this to zero ensures that a new position is captured
  • timeout - the maximum number of milliseconds to wait for a response
  • enableHighAccuracy - requests that the most accurate position is captured. A request specifying this typically takes longer and consumes more power, and may be denied by the hardware. It’s just a suggestion really.

Once captured, the a server side Apex action is invoked to store the position in a custom field on the user record, along with a timestamp of when it was captured:

    public static String StoreUserLocation(Decimal latitude, Decimal longitude)
        String result='SUCCESS';
	    User u=new User(id=UserInfo.getUserId(),
	    update u;
        catch (Exception e)
        return result;

Note that I don’t need to retrieve the user record from the database to update it, I can just instantiate a new User object in memory and specify the id based on the UserInfo global - always worth saving a SOQL call if possible.

The component in action is shown in the screen capture video below:

Delivery Request

The flip side of the user check in is the delivery request. This is where a customer requires something to be collected from one (their) account and delivered to another. As this isn’t something that is likely to be requested on the fly, this is handled by a Visualforce page, although still using the Lightning Design System to style a Visualforce form.

To create a new delivery, the customer (me in this case) specifies the From and To account and any specific delivery instructions. I’m sending a package to Universal Containers as they seem to have a new social media team and I am keen to engage with them:

Screen Shot 2016 08 28 at 17 55 11

The package is being sent from Sea Palling Beach, where myself and the Jobs Admin have recently checked in:

Screen Shot 2016 08 28 at 17 56 44

When the record is saved, the controller sets the owner to the user who has most recently checked in to a location within 50 miles - if no such user can be found, the owner will be left as the user that created the record:


    public PageReference save()
        Account acc=[select ShippingLatitude, ShippingLongitude
                     from Account
                     where id=:delivery.From__c];
	List users=[SELECT id, Location__c, Location_Timestamp__c
	 	    FROM User
	 	    WHERE DISTANCE(Location__c, GEOLOCATION(:acc.ShippingLatitude, :acc.ShippingLongitude), 'mi') < 50
                    ORDER BY Location_Timestamp__c DESC];
	if (users.size()>0)
        insert delivery;
        ApexPages.StandardController std=new ApexPages.standardController(delivery);
        return std.view();

 So even though I created the delivery request, it was assigned to the job admin user as they checked in more recently:

Screen Shot 2016 08 28 at 18 18 31

That’s It?

That is indeed  all there is to it - a couple of custom fields on the user object, a custom object, one lighting component, one Visualforce page and two Apex classes. If you are so inclined, you can view the full code at the github repository.