Sundog

Calling Amazon MWS - Part 2

terry.luschen@sundoginteractive.com (Terry Luschen) — 2015-01-28T05:23:37+00:00

In order to get a report out of Amazon MWS you have to request the report to be run. It does not return the report right away, but instead puts the report into a queue to be processed. Later we will have to make another call ‘GetReportRequestList’ to see if the report is done and then finally another call ‘GetReport’ to actually retrieve the report.

In my first blog post about Amazon MWS called ‘Calling Amazon MWS - Part 1’ we saw when using the Amazon MWS Scratchpad all the authentication strings needed to make these various calls. There are two main places these types of credential values can be stored.
1) In a Custom Object
2) In a Custom Setting
A custom object allows you to enter the data into an encrypted field, while the custom setting allows you to have only Admins have access to them so they each have their pros/cons. In this example we will see a custom object being used.

So here is the code to make the call…

public static string makeRequestReportCall(SecurityFields__c c){
	    string returnXML = '';

	    Http h = new Http();
	    HttpRequest req = new HttpRequest();
	    req.setEndpoint(c.End_Point__c);
	   
            Datetime now = Datetime.now().addDays(1);
	    String timestamp = now.formatGmt('yyyy-MM-dd')+'T'+now.formatGmt('HH:mm:ss')+'Z';
	    DateTime start_dt; 
	    DateTime end_dt;
	    if(c.Start_Date__c == null){
	    	start_dt = dateTime.Now().addDays(-30);
    	    }else{
    		start_dt = c.Start_Date__c;
    	    }
	    
	    if(c.End_Date__c == null){
	    	end_dt = dateTime.Now();
    	    }else{
    		end_dt = c.End_Date__c;
    	    }
	    
	    String start_date = start_dt.formatGmt('yyyy-MM-dd')+'T'+start_dt.formatGmt('HH:mm:ss')+'Z';
	    String end_date = end_dt.formatGmt('yyyy-MM-dd')+'T'+end_dt.formatGmt('HH:mm:ss')+'Z';
	 
	    String request = 'AWSAccessKeyId='+urlencode(c.Access_Key_ID__c) +	
				'&Action='+urlencode('RequestReport') +
	            '&EndDate=' + urlencode(end_date)+
	            '&MarketplaceIdList.Id.1='+urlencode(c.Marketplace_ID__c) +
	            '&Merchant='+urlencode(c.Seller_ID__c) +
	            '&ReportType=' + urlencode( '_GET_XML_ALL_ORDERS_DATA_BY_ORDER_DATE_')  +
	            '&SignatureMethod='+urlencode('HmacSHA256') +
	            '&SignatureVersion='+urlencode('2') +
	            '&StartDate='+urlencode(start_date) +
	            '&Timestamp='+urlencode(timestamp) +                                                                           
	            '&Version='+urlencode('2009-01-01');
	 	 
	    String canonical =  'POST'+'\n'+
	                      c.Service_URL__c + '\n'+
	                      '/\n'+
	                      request;
	    String key = c.Secret_Key__c;
	     
            String signature = getSignature(canonical, key);	 
	    String body = request + '&Signature=' + urlEncode(signature);
	   	    
            Http http = new Http();
            req.setCompressed(true);
            req.setHeader('Host', c.Service_URL__c);
            req.setHeader('Content-Type','text/xml');
            req.setHeader('x-amazon-user-agent','Salesforce');
            req.setEndpoint(c.End_Point__c + '/?'+body);
            req.setMethod('POST');
	    req.setTimeout(60000);
  
	    string bodyValue = '';
	    if(test.isRunningTest()){
	      bodyValue = '<?xml version="1.0"?>' +
			'' +
			  '' +
			    '' +
			      '9999999999' +
			      '_GET_XML_ALL_ORDERS_DATA_BY_ORDER_DATE_' +
			      '2009-01-21T02:10:39+00:00' +
			      '2009-02-13T02:10:39+00:00' +
			      'false' +
			      '2009-02-20T02:10:39+00:00' +
			      '_SUBMITTED_' +
			    '' +
			  '' +
			  '' +
			    '99faca76-b600-46d2-b53c-0c1a4533e22a' +
			  '' +
			'';
	    }else{
	      HttpResponse res = h.send(req);
	      bodyValue = res.getBody();
	    }
	    
	    System.debug(bodyValue);

	    returnXML = bodyValue;
	    return returnXML;
}

public static string getSignature(string canonical, string key){
		String signature;
                Blob bsig = Crypto.generateMac('HmacSHA256', Blob.valueOf(canonical), Blob.valueOf(key));
		signature = EncodingUtil.base64Encode(bsig);
		return signature;
}

There were two main posts on the Salesforce Developer Discussion Forums site that allowed me to reach the above solution. One and Two. Thank you so much to those contributors.

Notice the getSignature() method near the end of the code. Originally I was having some errors here because I was encoding my values twice. But the above code allows for the connection to succeed.

Now once you get your XML you just need to parse it. Here is an ‘oldie but a goodie’ blog post that can help you get started with that XML parsing: Parsing XML in Salesforce APEX

In my next two blog posts I look forward to diving into the call ‘GetReportRequestList’ to see if the report is done and then finally another call ‘GetReport’ to actually retrieve the report.

Happy Integrating!

Calling Amazon MWS - Part 1

terry.luschen@sundoginteractive.com (Terry Luschen) — 2015-01-27T11:42:56+00:00

As their site states: Amazon Marketplace Web Service (Amazon MWS) is an integrated Web service API that helps Amazon sellers to programmatically exchange data on listings, orders, payments, reports, and more.

The best way to get a sense of what is available with the API is to use their scratchpad page.

One of the most common API requests is the ‘RequestReport’ operation in the ‘Reports’ section of the API.

The base page for the Report piece of the API can be found here: Report Start Page. Notice that the API date at the top says “Version 2009-01-01”. That date will be used later in the code.

To get to the specific Request and Response pieces of the ‘RequestReport’ operation we go to the api documentation site for Amazon MWS. Notice again that API Version of 2009-01-01 is in the breadcrumb at the top.

But where do you get that Authentication information like the SellerId that is used on the Scratchpad page? You need to sign up to be an Amazon Professional (not Individual) seller. It seem counter-intuitive that you need to sign up to be a seller in order to do development, but that is how it works. This FAQ site answers most questions about how to sign up. When you are ready to sign up ( It does cost about $40 a month ), you go to this page and click on the ‘Sign up for MWS’ button. Then on the next page click on the ‘Register Now’ link. On this next page it is important that you click on ‘Sell as a Professional’.

The logic must be that if you are selling ‘As an Individual’ with fewer than 40 items a month that you do not need access to the API.

Now you can finish the Sign-up process and you will get your credentials. Store these in a safe place and never share them!

Yeah! You can now go back to the Scratchpad page and you can fill out your information so that you can try out some of the API calls.

In my next three blog posts I would like to look at the ‘RequestReport’(Request that Amazon MWS runs a report for you), ‘GetReportRequestList’(Check to see if the report is done) and ‘GetReport’(Get the actual report) API calls.

How To: Call an Apex Controller Method From Lightning Component

craig.isakson@sundog.net (Craig Isakson) — 2015-01-13T20:16:50+00:00

Lightning seems to be all the rage these days! How could it not? It is exciting stuff and changes the way that Salesforce developers need to think when coming up with their solutions on the platform. One of the most common things you will need to do with Lightning will be to call methods on an Apex controller for any server side logic you need to perform.

If this is something you are trying to figure out then you have come to the right place. If you are familiar with using Javascript Remoting in Visualforce, then calling Apex methods from a Lightning component will look very very similar.

The following assumes you have your namespace setup and you are able to create a Lightning app that you can hit from the preview in the developer console. Once you have a Lightning app created, you can create a component to add to the app. This component will be what makes the callout to the controller.

<aura:component controller="yournamespace.YourController">
    <aura:attribute name="yourAuraVariable" type="String"/>
    ...
</aura:component>

Here you can see that we have a controller called YourController. This controller will have the method that we use to populate the aura attribute yourAuraVariable. The controller code will look something like this.

public class YourController {
    @AuraEnabled
    public static String getMyAuraVariableFromTheController() {
        return 'Here is the text from my controller';
    }
}

Now that we have our controller code built, it is just a matter of calling this code from a little bit of Javascript in the Lightning component.

({
    getMyAuraVariable : function(component) {
        //this will be the method name on the controller contained within the 
        // controller property on our aura component
        var a = component.get("c.getMyAuraVariableFromTheController");
        var self = this;

        //build the callback since all controller calls are batched and sent
        //asynchronously
        a.setCallback(this, function(action){
            //action.getState() will let you know if it was successful or not. 
            //a good idea would be to check this to make sure everything
            //worked as expected.
            console.log(action.getState());

            //set the return value from the controller to the variable on our component
            component.set("v.yourAuraVaraible", action.getReturnValue());
            //do more stuff
        });

        //fire off the action
        $A.enqueueAction(a);
    }, 
    ...
})

This bit of code is contained within the helper of our component. You simply call this function from anywhere, either on init or another action on the page and it will populate our aura variable with the results that are returned from the controller. In this case ‘Here is the text from my controller’. Obviously in most cases you will want to return more complex data than just a string but this is just to get you off the ground and running with how calling controller methods work. Good luck and let me know if you have any questions!!

How To: Call Batch Apex from Batch Apex

craig.isakson@sundog.net (Craig Isakson) — 2015-01-13T20:14:56+00:00

Batch Apex is a powerful tool that can be used to do all sorts of cool stuff on the Salesforce platform. Batch Apex is typically used to perform larger operations that would normally be restricted by governor limits. It allows you to break that larger operation up into smaller bite-size chunks that are easier to deal with.

Since batch jobs are meant to deal with one query and processing records from that query, there may be times when you want to spin off another batch job when one is complete to finish your processing. An example of when you might want to do this would be if you needed to perform a batch job to delete thousands of records but once those records were deleted, you need to create thousands of records based off of another query or object.

Starting a batch job from a batch job is fairly straightforward. The key is placement of the code that actually starts the new batch. The code that starts the new batch needs to be placed in the finish method of the initial batch.

global void finish(Database.BatchableContext BC){
	//Add your start code for the other batch job here
	Database.executeBatch(new YourOtherBatchClass());
}

That’s pretty much it. Calling executeBatch from the finish method of one batch job will kick off the new batch job. There are a lot of fun things you can do with this. One thing I have done has been to create a recursive batch job that calls itself. If you are ever in a situation where you need to call a batch from a batch, I hope this helps!

Having trouble with animations in iOS?

craig.isakson@sundog.net (Craig Isakson) — 2015-01-13T20:11:27+00:00

I was recently doing a lot of animations for a couple of iOS apps I was building. For the most part, programming animations for iOS is straightforward and work really well. I kept running into some small issues that were really bugging me and I was unable to find the answer to my issue.

The issue I was having was using the UIView animateWithDuration method. I was taking a few views on my interface and moving/resizing/changing their alpha. The method is pretty straightforward. You basically call the method, tell it what you want the views to be when the animation is complete and set some other parameters. Here is an example:

[UIView animateWithDuration:0.5 delay:0.1
                            options: UIViewAnimationCurveEaseIn animations:^{
                                [self.viewSample setFrame:CGRectMake(-100, self.viewSample.frame.origin.y, self.viewSample.frame.size.width, self.viewSample.frame.size.height)];
                            }
                         completion:^(BOOL finished){
                         }];

This animation will take the sampleView and move it to to an x coordinate of -100 so it moves off the screen. The problem I was having with this is it would appear to move off the screen correctly but then pop back to it’s original position. As it turns out, this issue is due to AutoLayout being checked for the ViewController. Since -100 is not within the constraints of a valid view, AutoLayout takes over and tries to rearrange the views for you. This was obviously not what I wanted but was glad to see that simply not using AutoLayout on the ViewController fixed the issue.

If you run into this issue, try not using AutoLayout and your problems should be solved!!

Realtime Tracking with the Streaming API

craig.isakson@sundog.net (Craig Isakson) — 2015-01-13T20:09:53+00:00

At Dreamforce 2014 I had the opportunity to present some cool code on using the Streaming API to perform realtime tracking within VisualForce pages on Salesforce. It was a great experience! If you have been wondering how to use the Streaming API or how to incorporate Google Maps into VisualForce, check out the recording of the session and leave your comments below.

How to setup your Salesforce Organizations

terry.luschen@sundoginteractive.com (Terry Luschen) — 2014-12-31T12:02:15+00:00

If is always hard to decide within a business how many Salesforce organizations there should be. There are so many different factors at work within each business that will make the outcome unique for each company.

Here are three basic questions that we have come up with to help jump start the important pieces of this conversation.

1) Do I need a
a) Single Global Organization?
b) Set of Organizations by Region?
c) Set of Organizations by Business Unit?

2) How many pieces of the typical Salesforce environmental architecture do I need to support each individual organization?

3) What source control tools will I use to manage my Salesforce environments?.

Question 1
This question gets you thinking about your business in detail. If you are a smaller business that only works in the United States, then you should be able to do everything in one org so this question will be easily answered. As the size of your business increases, the number of business units multiple and your presence across the globe widens this becomes a very important conversation. You will need to talk about languages, number formats, business processes, speed of new features, reporting roll-up, portals/communities and license cost considerations. Sometimes the way organizations get structured just ‘sort of happens’ over time. But once a company makes a strategic investment in the Salesforce platform this question should be discussed at all levels of the company.

Question 2
This question gets into discussions about…
a) How many sandboxes should be used?
b) How are code changes promoted through the sandboxes of an organization?
c) Who is in charge of the changes at each level: Business Analyst, Developer, Release Manager or IT?
d) How are changes to the system documented? Are they through change sets? Deployment log files? Google Docs?

Question 3
This last question is pretty specific, but there are a lot of options.
a) Are you going to rely on Salesforce to be the main backup for your code? Salesforce will guarantee backups of production database, but not for your sandboxes.
b) Are you going to use a separate source control tool like SVN or GIT?
c) Are you going to force developers to check out metadata files?
d) Are you going to do a nightly batch commit in the background that will not get every change, but will get the bulk of the metadata changes for that day?

Just knowing a few solid questions to ask will frame the conversation to get you to your critical decision points. There are no right and wrong answers up front. The answers for your specific company will only come after many detailed and important conversions.

Good luck! Let us know at Sundog if we can ever help you during these discussions. These are some of the best and most important discussions of any new implementation.

How do you tie your files to Salesforce? Azure? Sharepoint?

terry.luschen@sundoginteractive.com (Terry Luschen) — 2014-12-29T11:15:48+00:00

Back in July, 2013 I wrote a whitepaper, Salesforce - Where to Save my Files?, detailing a couple of options for storing your files outside of Salesforce. I really enjoyed exploring some very specific ways on how to save files to Box or Amazon’s S3 service. To recap, the reason why you might want to save your files outside of Salesforce is because you only get so much space for files and that space is based on your number of licenses. This link describes all the details but basically for an Enterprise org you get 11GB of file storage just for having your Salesforce org and then you get 2GB of file storage for every license. This may seem like a lot, but you need to take into account how your files will grow over time. For example let’s say you have 20,000 accounts and each account receives a 500KB PDF report every year. So for just one year that will take up 20,000 x 500KB = 10,000,000KB = 10,000MB = 10GB. If you only had 10 licenses, you would have 31GB of storage space, so you can see how quickly your file storage space would run out. Note that you can purchase more file storage space from Salesforce separately.

Of course, the decision to keep these files outside of Salesforce should not be taken lightly. The ability to control access to these documents with standard Salesforce security makes life very easy. The ability to full-text search documents saved as Chatter Files or in Salesforce Content is also a great benefit. One limitation to keep in mind with Salesforce Content is that you can have up to 2000 libraries. If you were thinking of creating library for each account, that usually will not work. But in the end, sometimes there are just too many files to store and the expense of having them in Salesforce will not overcome the other benefits.

So this leads us back to my whitepaper where even two years ago putting these files in Box of Amazon S3 were really solid options. A couple of recent links about Amazon S3 are the Amazon Toolkit which allow for native hooks in Apex for working with AWS S3 and EC2 and an AppExchange product called Cloudpond S3 Connector recently build to do the AWS S3 integration for you.

With both of those options a folder was created in Box or S3 with the Salesforce ID as the name of the folder. That works well and is very clean.

I could also envision an integration where the file is pushed to Box or S3 and then the URL is returned and that link is saved as a Chatter link or a link in Salesforce Content. That type of solution would be the best of both worlds in a certain sense because the file itself would be kept outside of Salesforce, but the access to the link could be controlled by standard Salesforce security.

Another couple of options that have caught my eye recently are saving the files to Sharepoint and Microsoft Azure.

This article, Salesforce connects SharePoint files to its cloud with new tool, is really exciting, because it goes straight after solving this problem. The files are saved in Sharepoint, but they can be accessed through Salesforce with no custom coding. Here are links from Winter 2015 Release Notes and the Salesforce Files page.

Another option I have seen is to store all of the files in an Azure Blog Storage. This type of storage can be very cost effective and can leverage Azure’s infrastructure in the cloud. But anything involving Azure would involve a custom built solution that must be supported. From Salesforce a Canvas App or web service integration could be used to get to the files stored on the Azure Blog Storage.

What have you seen for ways to solve this common enterprise problem?

Happy Integrating!

Stopping a Visualforce Command Button from Firing the Click Event Twice - Part II

terry.luschen@sundoginteractive.com (Terry Luschen) — 2014-12-26T20:37:52+00:00

Back in September of 2013 I wrote this blog post, Stopping a Visualforce Command Button from Firing the Click Event Twice, about how to stop a Visualforce page button from posting back to the controller twice if the user double-clicks on it. The solution that I presented works in a certain situation and the other solutions posted by others also help. But I ran into a new use case where that solution does not work.

The scenario where my previous solution does not work is if the action behind the button has the potential to return an error or validation message. The problem is that the button will now be disabled when the message is displayed so the user will not be able to click the button again. As another potential solution I was wondering if we had the button call an apex actionFunction instead. In this case the button could be disabled because we could make it enabled again by using the oncomplete attribute of the actionFunction.

But I came across another way of doing this by just having some Javascript not allow the button to be clicked again for a couple of seconds after it is clicked. Here is how it works:

1) Create this Javascript at the top of your Visualforce page between some script tags

        var waitTime = 1;       
        var countDown = waitTime+1;
        var clickAllowed = true;
                
        function stopDoubleClick(){        
            countDown=waitTime+1;            
            if(clickAllowed==true){
                clickAllowed=false;
                runCountDown();
                return true;
            }else{
                return false;
            }
        }
        
        function runCountDown(){        
            countDown--;            
            if(countDown<=0){
                clickAllowed=true;
            }else{     
                window.setTimeout(runCountDown,1000);
            }
        }

2) Then in your commandButton call the stopDoubleClick() javascript function

<apex:commandButton onclick="return stopDoubleClick();" action="{!doWork}" value="Take Action"></apex:commandButton>

That solution has worked out pretty well. I have tried to click a whole bunch of times and I never get a duplicate call on my controller. But it is always hard to rely on a timing mechanism.

If you have an easier or better solution please let everyone know! I even wish Salesforce would make an attribute or present a best practice on how to do this consistently.

Happy Visualforcing!

Extend Your Brand Voice to Unexpected Places

dean.froslie@sundog.net (Dean Froslie) — 2014-12-23T14:32:59+00:00

For many organizations, it’s really challenging to master voice and tone.

Branding efforts often (justifiably) focus on strategy, archetypes, visual systems, brand standards and related areas. They’re necessary elements, and they’re usually easier to wrangle. As a result, solidifying voice and tone standards can receive less attention.

And even when voice and tone guidelines are clearly defined, it’s a seemingly endless effort to ensure they’re implemented across your various customer touchpoints. It’s one thing, for example, for product or service pages to reflect your desired voice and tone. It’s another, however, to extend your standards to success messages, confirmation emails, help pages, third-party applications and customer service communications (among others).

The enormity of the voice and tone effort makes brands like MailChimp even more impressive. They’ve adopted a “one voice, many tones” philosophy that’s simple, clear and accessible. Kate Kiefer Lee of MailChimp writes and speaks extensively on this topic, and they even freely share their voice and tone guide.

Today I discovered yet another great example from MailChimp. I was performing some quick research on the meta description practices of various brands, which involves peeking at the source code of their websites. Here’s what lurks “under the hood” of the MailChimp homepage:

It’s a witty, geeky example of discovering the MailChimp brand in the most unexpected places. The MailChimp voice, after all, is “fun but not childish” and “clever but not silly.”

Where (and how) can your content delight your audiences in similar ways?

To learn more, read these posts from Kate Kiefer Lee:
Tone and Voice: Showing Your Users That You Care
Vulnerable Branding: Keeping your company’s voice honest
Don’t Poke the Bear: Creating Content for Sensitive Situations