Node.js on Azure calling SAP Gateway

It’s hard to miss the growing excitement around Node.js, a server-side JavaScript environment, especially ever since Microsoft supports the Node.js project and even offers Node app hosting as a service on Windows Azure.

I was eager to get a first hands-on with Node on Azure, so in this post we are going to build a Node.js app running on Windows Azure, accessing SAP business data through SAP Gateway. I decided to go for ‘the Cloud’ all-the-way, so we will develop both a server and client (i.e. browser) application in JavaScript using Cloud9, a web based IDE. The browser app is based on SAPUI5. It displays a list of Sales Orders and the user can also request an overview in Excel, generated on-the-fly on the server. To easily work with the source code of the project, we will use a repository on GitHub.

Here are screenshots of the end result:

SNAGHTMLcac40b9

SNAGHTMLb1e5875

So if you are in for some exposure to cutting edge technologies, read on! Winking smile

The players

SAP Gateway is a SAP server add-on that allows you to access  business data in SAP systems through OData, an open web protocol for querying and updating data, built on common web technologies like HTTP, XML and REST.

SAPUI5 (officially known as “UI development toolkit for HTML5”) is a JavaScript library, currently in Beta, to build HTML5 apps.

Node.js is a server platform that is quickly gaining popularity. It is built on Chrome’s V8 JavaScript runtime and allows you to create server applications in JavaScript. Node.js uses an event-driven, non-blocking I/O model that makes it lightweight and efficient.

Cloud9 is a web-based IDE + node.js runtime environment offered as a service (partially open sourced). In Cloud9 you can develop, debug and test node.js applications, all from within your browser. Cloud9 can use a GitHub repository to manage and share the source code. Once you have completed your application, you can deploy the application to Windows Azure.

Windows Azure is Microsoft’s cloud platform. One of the services is hosting a Node.js application in a scalable environment.

GitHub hosts source code repositories,using the Git version control system. GitHub offers public and private repositories. To work with GitHub, you need to use Git client software.

The Big Picture

Before we go into the details, let’s sketch how it all fits together:

image

For development, we’ll create a code repository on GitHub. This way we can add/update source code and resource files both from our local machine (via Git) and on Cloud9. We will use the repository to add all SAPUI5 js and css files that are required to our project files. In Cloud9 we use a clone of the GitHub repository. This way we can push and pull Cloud9 project files in and out of the GitHub repository.

Cloud9 is a fully web-based development environment: developers use a browser to access the Cloud9 code editor, system command prompt, debugger, Node.js console output etc. We can run a Node.js application directly on the Cloud9 environment and use a special url to access the running server app as a (test) user.

Actually we need 2 applications: a server application which runs on the Node.js server and acts as middleman to SAP Gateway (and possibly other data sources) and a client application (HTML5, SAPUI5)  which runs in the browser, but uses html/css/js files which are hosted on the server. It is important to keep these two apart, even if we develop both in JavaScript.

The client and server app interact in the following way:

Display Sales Orders in Data Table

  • the user starts the app by requesting the server url
  • the server sends the client application html file (index.html) which instructs the browser to load additional SAPUI5 resource files and the client js code.
  • the client app starts up and creates a SAPUI5 OData Model and Data Table to fetch and display Sales Order data. The data isn’t fetched from the SAP Gateway directly (not allowed due to the Same Origin Policy). Instead, the request is sent to the server, which forwards it to SAP Gateway. The server acts as a reverse proxy.
  • The server passes the OData response from SAP Gateway to the browser. Here, the OData feed is parsed using SAPUI5’s built in OData support and the extracted data is shown in the Data Table.

Download as Excel

  • When the user clicks on ‘Download as Excel’ button, the client app requests a special url from the server.
  • This will trigger the server to fetch the Sales Order data from SAP Gateway, parse it (but now on the server side!), convert it into an Excel format and return this to the user. The user experience is identical to requesting a downloadable file.

Once we have the application working in the Cloud9 environment, we can request deployment to Windows Azure. Cloud9 will generate configuration files, package the apps, upload and deploy to Azure.

Caveats:

  • This demo project involves quite some platforms and services. Within the scope of this blog post it is not feasible to provide a very detailed, click-by-click walk-through. Instead I will refer to tutorials for the respective services which contain detailed instructions and I will point out the main route and highlights along the way.
  • I recommend that you use Safari, Chrome or Firefox on GitHub and Cloud9. IE support is lagging (GitHub) or non-existent (Cloud9).

Step 1: Create a GitHub repository

First we need to create a repository on GitHub to manage and share our source code across environments.

  1. Sign-up with a (free) account for GitHub. A free account let’s you create public repositories.
  2. Follow the instructions on Set up Git to install Git local software and set up SSH keys
  3. Create a repository on GitHub. We’ll assume it is named: NODEJS_SAPGW
  4. Clone the repository to local machine
  5. Try out pushing and pulling content from your local machine into the repository until you become familiar with the process

Here are the important Git commands you will need:

// Clone repository
$ git clone git@github.com:[yourname]/NODEJS_SAPGW.git

// push content
git add .
git commit -m "Commit message"
git push

// pull content
git pull

I think GitHub is a great tool for developers, so I encourage you to check out the GitHub help center to learn more.about the other features of Git and GitHub. For our demo project the above commands are all we need.

Step 2: Develop on Cloud9

Now we have a repository on GitHub, we can start a project on the Cloud9 IDE.

  1. Signup on http://c9.io. It is very convenient to signup with your GitHub account. Look for the small GitHub icon.
    image
  2. After signup, you will see your GitHub repository (I blurred out my other repositories):
    image
  3. Select the repository and click ‘Clone to edit’ and select Shared Development server.
  4. The Cloud9 environment will make a Clone of the GitHub repository available in your Cloud9 environment.
  5. Once the Clone is completed, you see the following screen. Click on ‘Start Editing’.
    image
  6. This will bring you into the project environment. Create a folder called ‘Public’. This folder will contain all the files required for the client application. Add the index.html and app.js (described below) and push the changes to GitHub and pull to your local machine. Add the SAPUI5 resources on your local working directory and push to GitHub and pull from Cloud9. All this pushing and pulling sounds like a workout, but trust me, you’ll get the hang of it!

Client application

The client application consists of an index.html file and app.js file and uses SAPUI5 to fetch and parse OData and render a data table widget.

The index.html includes the SAPUI5 core library (which dynamically will load all dependencies), the application js file and a placeholder for the data table.

<!DOCTYPE html>
<html>
    <head>
        <meta http-equiv='X-UA-Compatible' content='IE=edge' />
        <title>Node.js calling SAP Gateway</title>

        <script id="sap-ui-bootstrap" 
           src="resources/sap-ui-core.js"
           data-sap-ui-theme="sap_goldreflection"
           data-sap-ui-libs="sap.ui.commons, sap.ui.table"></script>

        <script src="app.js"></script>

    </head>
    <body class="sapUiBody">
         <img src="images/sap_logo.png" />
         <div id="salesorders"></div>
         
    </body>
</html>

The app.js contains the client code:

// Let's define some shortcuts to increase
// readability of the code
var ODataModel = sap.ui.model.odata.ODataModel,
    TextView = sap.ui.commons.TextView,
    Label = sap.ui.commons.Label,
    DataTable = sap.ui.table.DataTable,
    Toolbar = sap.ui.commons.Toolbar,
    Button = sap.ui.commons.Button,
    Column = sap.ui.table.Column,
    Right = sap.ui.commons.layout.HAlign.Right,
    Begin = sap.ui.commons.layout.HAlign.Begin,
    SelectionMode = sap.ui.table.SelectionMode;

// Specify the SAP Gateway SalesOrder service as an OData model.
// Please note: we do not connect directly to SAP Gateway (no host specified)
// Instead, the calls are made to the Node.js server,
// which will proxy the calls to SAP GW
// This way, we comply with Same Origin Policy of the browser. 
var salesOrderService =
        "/sap/opu/sdata/IWFND/SALESORDER",

    // SAP Gateway only supports XML, so don't use JSON
    asJson = false,
    salesOrderModel = new ODataModel(salesOrderService, asJson),
    salesOrderCollection = "SalesOrderCollection";

// Create a button to request an Excel workbook from server 
var button = new Button({
    text: "Download as Excel",
    icon: 'images/excel.png',
    iconFirst: false,
    height: '24px',
    press: function () {
        window.location = "/workbook"
    }
});

var toolbar = new Toolbar({ 
    items: [button]
});

// Create a master table with sales orders 
var salesOrders = new DataTable({
    title: "Sales Orders",
    width: "600px",
    visibleRowCount: 20,
    toolbar: toolbar,
    selectionMode: SelectionMode.None,
    editable: false
});

// define the relevant column options
var salesOrderColumns = [
    { header: "Sales Order ID", value: "{SalesOrderID}", width: '100px' },
    { header: "Customer Name", value: "{CustomerName}", width: '100%' },
    { header: "Net", value: "{NetSum}", width: '100px', hAlign: Right },
    { header: "Tax", value: "{Tax}", width: '100px', hAlign: Right },
    { header: "Total", value: "{TotalSum}", width: '100px', hAlign: Right }
];

// create the columns
salesOrderColumns.forEach(function (options) {
    var label = new Label({ text: options.header }),
        template = new TextView({ text: options.value }),
        column = new Column({
            label: label,
            template: template,
            width: options.width,
            hAlign: options.hAlign || Begin
        });

    salesOrders.addColumn(column);
});

// connect the data table to the SalesOrder service
salesOrders.setModel(salesOrderModel);

// An OData request for the SalesOrderCollection 
// will return the sales orders. 
// Each sales order should result in a table row.
salesOrders.bindRows(salesOrderCollection);

// Put table in the DOM.
// placeAt will automatically defer if 
// DOM is not ready yet (like in this demo).
salesOrders.placeAt("salesorders");

With that, our client application is in place.

Server application

Node.js includes core library functions to create an HTTP server and client. You have probably seen the famous 6 lines of code to create a ‘Hello world’ http server in Node.js. The core library is intentionally kept light-weight and (relatively) low-level. But is very easy to extend the base capabilities of Node by including modules in your code. The Node.js community has built a wide range of modules that handle common tasks like XML parsing, static file serving, request routing etc.

To keep our server application compact, we will use 2 of these modules:

  • express.js: a web application framework on top of Node, to provide request routing (connecting incoming requests to JS functions) and serving the static files (the html, js and css files for the client application).
  • xml2js: to parse the SAP Gateway OData response into a JavaScript object

To install the modules in your Cloud9 project, you can use NPM (Node Package Manager). Just issue the following commands in the command field:

npm install express
npm install xml2js

image

This will install the modules (and their dependencies) in the node_modules folder of your Cloud9 project. Click the refresh button next to Project Files and you can see the results.

Now we have the required modules installed, we can coding our server.js file which contains the server application code. Based on the code of our client application, the server application needs to respond to the following requests:

  • / : redirect requests without pathname (i.e. server name only) to index.html in the ‘public folder
  • /sap/*: forward (proxy) all requests for SAP Gateway content and include authentication
  • /workbook: generate an Excel workbook on the fly and deliver as download
  • otherwise: serve static files for the client app if contained in the ‘public’ folder

So here is server.js, the code to implement the server application

var http = require('http'),
    express = require('express'),
    xml2js = require('xml2js'),
    app = express.createServer(),
    port = process.env.PORT;

// The SalesOrder service requires authentication
// get the username/password from the SCN page. 
var username = 'USERNAME',
    password = 'PASSWORD';

var sapgw = {
  host: 'gw.esworkplace.sap.com',

  // set up Basic authentication
  headers: {
    'Authorization': 'Basic ' + new Buffer(username + ':' + password).toString('base64')
  }
}

// Action: Proxy an incoming (ie. from the user's browser) request to the SAP Gateway
function proxy(user_request, user_response) {

  // We're proxying calls,
  // so copy the path from the user request
  sapgw.path = user_request.url;

  // Use a client request to call SAP Gateway
  http.get(sapgw, function (sapgw_response) {

    // Include a content type in the response header
    user_response.header('Content-Type', 'application/atom+xml;type=feed');

    // In Node, http responses are streams. You can just
    // pipe the response data from the Gateway to the user.
    sapgw_response.pipe(user_response);
  });
}

// Action: Generate an Excel workbook containing SalesOrders on the fly  
function workbook(req, res) {

  // We will fetch the SalesOrderCollection from SAP Gateway
  sapgw.path = '/sap/opu/sdata/IWFND/SALESORDER/SalesOrderCollection';

  // Kick-off by fetching the SalesOrderCollection..       
  http.get(sapgw, function (sapgw_response) {
    var xml = '';

    // Every time Node receives a chunk of data
    // from SAP Gateway, the 'data' event fires.
    // We just collect all chunks into a string
    sapgw_response.on("data", function (chunk) {
      xml += chunk
    });

    // The 'end' event fires when the SAP Gateway response is done 
    // We can start processing the xml string...
    sapgw_response.on("end", function () {

      // Node.js doesn't automatically parse the XML (like XmlHttpRequest),
      // so we need to do that explicitly.
      // We will use the xml2js module to parse the XML string
      // into a JavaScript object
      // Create a parser to convert the XML to JavaScript object
      var parser = new xml2js.Parser();

      // The 'end' event fires when the parser is done.
      // The resulting JS object is passed as parameter.
      parser.on('end', function (result) {
        // The result parameter is a complete representation 
        // of the parsed XML string.
        // We need to extract the values we need to render the workbook
        var rows = [],

        // columns is the subset of properties we want to include
        // in the Excel list
        columns = ['SalesOrderID', 'CustomerName', 'NetSum', 'Tax', 'TotalSum'];

        // the value of result['atom:entry'] is an array of objects, 
        // representing 's in the XML string.
        // Each entry represents a SalesOrder
        result['atom:entry'].forEach(function (entry) {
          var row = {},
            // properties points to the parsed properties 
            properties = entry['atom:content']['m:properties'];

          columns.forEach(function (property) {
            // get the value for the property
            row[property] = properties['d:' + property]['#'] || 
                            properties['d:' + property];
          });

          // add it to the rows 
          rows.push(row)

        })

        // We're done processing the response and have extracted all data required to 
        // create the workbook.
        // First, include the proper HTTP headers
        res.header('content-disposition', 'attachment;filename=SalesOrders.xml');
        res.header('Content-Type', 'application/vnd.ms-excel');

        // Next, render the contents of Excel workbook as
        // Excel 2003 XML format, using the Workbook.ejs template
        res.render('Workbook.ejs', {
          layout: false,
          rows: rows
        });

      });

      // Set the parser in motion with the xml data received
      // from SAP Gateway. The parser will turn the XML string into
      // a JS object. When done, the 'end' event will fire
      // and the 'end' event handler defined above will extract the values
      // and render the Workbook.
      parser.parseString(xml);
    });

  });
};

// Action: perform a redirect to the home page (index.html)   
function home(req, res) {
  res.redirect('/index.html')
}

// Now we will set up the routing for the server...
// Look if the request is for a static file in the public directory
// this is where the client side html, css, js and SAPUI5 resources are located
app.use(express.static(__dirname + '/public'));

// If you just call the server (root), redirect to homepage
app.get('/', home);

// Requests starting '/sap/' are proxied to the SAP Gateway
app.get('/sap/*', proxy);

// Route for workbook action
app.get('/workbook', workbook);

// Routes are set up,
// Now start the app server...
app.listen(port);

console.log('Server started on port ' + port);

You can find the username/password for the SAP Gateway demo environment on this SCN page.

The server application uses an ejs view template to generate the Excel workbook content. Express expects these views by default in the ‘views’ directory. An ejs template is a text file interspersed with JavaScript (very similar to classic asp or jsp pages). The view renderer uses the template and provided parameter to output a text string, in our case the Excel content. Here’s the relevant part of the view template (rows are passed as a parameter):

<Worksheet ss:Name="Sheet1">
  <Table ss:ExpandedColumnCount="5" 
         ss:ExpandedRowCount="<%= rows.length+1 %>" 
          x:FullColumns="1"
          x:FullRows="1" 
         ss:DefaultRowHeight="15">
   <Column ss:AutoFitWidth="0" ss:Width="77.25"/>
   <Column ss:AutoFitWidth="0" ss:Width="138.75"/>
   <Column ss:StyleID="s67" ss:AutoFitWidth="0" ss:Width="56.25" ss:Span="2"/>
    <Row ss:StyleID="s62">  
            <Cell><Data ss:Type="String">SalesOrderID</Data></Cell>
            <Cell><Data ss:Type="String">CustomerName</Data></Cell>
            <Cell><Data ss:Type="String">Net</Data></Cell>
            <Cell><Data ss:Type="String">Tax</Data></Cell>
            <Cell><Data ss:Type="String">Total</Data></Cell>
    </Row>
    <% rows.forEach(function(row){ %>
        <Row>   
            <Cell><Data ss:Type="String"><%= row.SalesOrderID %></Data></Cell>
            <Cell><Data ss:Type="String"><%= row.CustomerName %></Data></Cell>
            <Cell><Data ss:Type="Number"><%= row.NetSum %></Data></Cell>
            <Cell><Data ss:Type="Number"><%= row.Tax %></Data></Cell>
            <Cell><Data ss:Type="Number"><%= row.TotalSum %></Data></Cell>
        </Row>
    <% }) %>
</Table>
</Worksheet">

Now, let’s run the application in debugging mode on Cloud9. The Output tab will show the output from the application initialization and any console.log statements you include in the code. It also provides the url where you can access the running application.

image

Click on the server url and our application opens up:

SNAGHTMLb193d38

Click on the ‘Download as Excel’ button and Excel workbook with the Sales Orders will open:

SNAGHTMLb1e5875

OK, we now have a working application in Cloud9! There’s a lot more to explore in the IDE. For example, try the debugging options: you can set breakpoints, just like in desktop IDEs. Code execution on remote server will halt and you can explore variable and parameter values, etc. You can also push your changes to the central repository on GitHub.

Here, we will continue our Quick Tour by deploying the application we just completed to Windows Azure.

Step 3: Deploy to Windows Azure

Cloud9 makes deployment to Windows Azure very easy. This tutorial shows all the steps in detail. You can skip the first steps because we already have a working app on Cloud9. Please start at “Create a Windows Azure account’.

After the initial set up, you can request deployment to Azure:

image

Cloud9 will offer to create default configuration files for you. Please accept.

Deploying can take a few minutes, while the application is packaged, uploaded and deployed on Azure:

SNAGHTMLb3a5c8b

Once deployment is active, you can click on it and see the details image

The Windows Azure management tool provides full details on the deployment:

image

Our application is now available on Windows Azure:

SNAGHTMLfdb040

http://nodesapgw.cloudapp.net

You can try it! It’s live.

You’ll notice that it takes quite some time to load the page. This is mainly caused by dynamic loading of SAPUI5 resource files. Optimizations are certainly possible, but outside the scope of this article.

This concludes our demo project to show how you can front-end SAP Gateway using Node.js on Azure.

‘Just because you can, doesn’t mean you should’

That’s a profound adage, so let’s wrap-up with a number of suggestions  why you may consider front-ending SAP Gateway in the Cloud:

  • Provide access to multiple backend systems through a single domain (origin). This way, you comply with the Same Origin Policy.
  • Integrate SAP data in an Azure based cloud solution (E.g. a hosted .Net, MS-SQL solution)
  • Offload non-business data http traffic of SAP Gateway to a more cost-effective platform.
  • Avoid exposing the SAP Gateway end-point to the public internet
  • Low-latency traffic by using caching to prevent repeatedly asking the same data via SAP Gateway. E.g. for type-ahead and suggestion/select drop-boxes in HTML5 apps, low-latency response are required. This is becoming more important with mobile apps where quick selections/suggestion lists are preferred of manual input of texts
  • Scale-out on demand
  • Protocol conversion: you may want to build client applications using other protocols (e.g. WebSockets) to expose business data. E.g. consider real-time auctions.
  • Content conversion: for example turning business data into Excel sheet as demonstrated in our application
  • Identity conversion: a consumer facing web application may allow consumers to login with their Facebook or Twitter account using OAuth2. The web application server can connect to SAP Gateway with a functional account.

Let me know if you have other suggestions/use cases.

Thanks for reading!

Helpful resources

Visualize SAP Gateway OData services

SAP NetWeaver Gateway OData services can contain a comprehensive set of entity types and associations. A visual representation can help you understand and communicate the contents of a SAP Gateway OData service.

In this post I will describe how you can explore SAP Gateway OData services, using the Open Data Protocol Visualizer, a free extension for Visual Studio 2010.

Installing the OData Protocol Visualizer in VS2010

The Open Data Protocol Visualizer is a Visual Studio 2010 extension, which you can install via the Extension manager.

Launch the “Extension Manager” from the Visual Studio “Tools” menu.

image

The visualizer is available from the online gallery, in the Tools>Data section:

SNAGHTML30ffac2.

Click to download and install. Restart Visual Studio and you’re done!

Configure a SAP Gateway service

Let’s use the ubiquitous Flight service on the SAP Gateway demo system as an example. You will need the demo username/password, provided on this page.

The service URL for the Flight service is:

http://gw.esworkplace.sap.com/sap/opu/odata/IWBEP/RMTSAMPLEFLIGHT_2/

To use the OData visualizer, you need to create a project in VS and include a Service Reference to an OData service:

First, create (for example) a Console project. In the Service References section, right-click and select ‘”Add Service Reference”:

image

In the dialog box, enter the service Url in the Address field and provide a meaningful name in the Namespace field. Click on Go and provide the demo username and password. Visual Studio will now download the service information ($metadata). Once completed, you should now be able to expand the Service header and see the collections available through the Flight service.

SNAGHTML3273d1f

Click OK, and the Service is available for use in your project

Visualize SAP Gateway service

To start using the visualizer, right-click the Service reference and click ‘View in Diagram”

image

This opens the visualizer window. In the left pane, you will find a model browser, which lists Entity types, Associations, Complex Types and their Attributes. The center pane contains a  ‘canvas’ where you select and layout the visual representation of the model.

SNAGHTML33dadcc

A detailed view of the Flight and Booking entity types:

image

The OData visualizer has several useful features:

  • layout: determine how the shapes are visually arranged
  • zoom in/out: canvas can be zoomed in and out.
  • collapsed view: just shows entity types and associations, no attributes
  • select/deselect: determine which part of the ‘model’ you want to visualize
  • view in browser: automatically open the collection for a selected Entity type. E.g. View ‘Booking’ in browser will request the BookingCollection from the Flight service.
  • export to XPS: export the content to XPS file format (unfortunately no other file format is supported…).

Here’s an example of a collapsed (partial) view of the Activity Management service (based on SAP CRM):

image

 

Finally, here’s an example of the properties pane showing the attribute properties of the Value attribute of a Sales Order Item.

SNAGHTML38012a7

 

The OData visualizer can be a valuable part of your toolbox. Give it a try!

Thanks for reading.

Getting started with SAPUI5

There are some great HTML5 UI frameworks available in the marketplace: Sencha Touch, KendoUI and Twitter Bootstrap to name just a few. And now SAP has unveiled the Beta version of SAPUI5, officially known as “UI development toolkit for HTML5”.

As a follow up on my post on Building a SAP mobile app with Sencha Touch 2, I was keen on having a hands-on with SAPUI5. These helpful posts helped me to get started.

In this post we’ll set up the Beta version of SAPUI5 for use in Visual Studio and create a demo SAPUI5 app that fetches SAP ERP Sales Orders from SAP Gateway and combines them with issues tracked in a SharePoint list.

Installing the SAPUI5 SDK on Visual Studio 2010

  1. Visit the homepage for SAPUI5 on SDN and download the trial version (SAPUI5 beta)
  2. Unzip the downloaded file HTML5Beta_complete.zip
  3. Open the folder with extracted content. Unzip ‘sapui5-static.zip’. You will need the ‘resources’ folder for development
  4. Rename demokit.war to demokit.zip and unzip. This archive contains the documentation which you can install locally.
  5. In Visual Studio, create a solution and within the solution ‘Add a new web site’
  6. Copy the contents of the demokit into the web site
  7. Copy the resource folder of sapui5-static.zip to the web site
  8. Rename index.html to default.html (the VS dev server opens default.html instead of index.html when browsing a directory).
  9. Your file/folder structure should now look like this:image
  10. Now select the web site and ‘View in Browser’

You now have access to the documentation of SAPUI5:

SNAGHTML379228f

A first test

Now we have the documentation up and running, we create our first test to see if we have all the required files.

To run this test, setup a new project or website in VS. Copy the ‘resources’ folder from sapui5-static into the site and create the following test.html file:

<!DOCTYPE html>
<html>
    <head>
        <meta http-equiv='X-UA-Compatible' content='IE=edge'>
        <title>SAPUI5 test</title>
        <script id="sap-ui-bootstrap"
           src="resources/sap-ui-core.js"
           data-sap-ui-theme="sap_platinum"
           data-sap-ui-libs="sap.ui.commons"&gt;&lt;/script&gt;
        <script>
            $(function () {
                // show an SAPUI5 alert box
                sap.ui.commons.MessageBox.alert("SAPUI5 ready for action.");
            })
        </script>
    </head>
    <body>
    </body>
</html>

The test.html file instructs the browser to load the sap-ui-core.js file which contains jQuery and also a dynamic loader which will load further required js and css files. To show that everything loads correctly, we simply create an alert box (one of the SAPUI5 controls) when the DOM is ready.

View the test.html file in the browser and you see:

SNAGHTML7a1fad9

OK, we’re all set to do something a little more exciting!

Building a demo app

Consider the following business scenario: a company is managing Sales Orders in SAP ERP. To improve order fulfillment, they use a issue tracking workflow system in SharePoint. Their requirement is to see the info from SAP and SharePoint in a single screen. Let’s build that with SAPUI5!

We will connect to the online SAP Gateway demo system to fetch Sales Orders from SAP ERP. Using the demo SAP Gateway system, you can try out SAPUI5 without having to install/configure any server side components.

Our issue tracking list is part of a SharePoint Online site (SharePoint Online is part of Microsoft’s Office365 offering). Each Issue contains a field Sales Order ID which we will use to filter. Here’s how the Issues look inside the SharePoint site:

image

Both SAP Gateway and SharePoint are OData providers, so we will use SAPUI5’s OData model to connect to the backend systems and parse the responses. Our application page will present three tables:

  • Sales Orders: the master table showing the available Sales Orders
  • Line items: a details table showing the line items belonging to the selected Sales Order
  • Issues: a details table showing the issues tracked in SharePoint related to the selected Sales Order.

This structure is already pre-defined in the body of the index.html file:

<!DOCTYPE html>
<html>
    <head>
        <meta http-equiv='X-UA-Compatible' content='IE=edge'>
        <title>SAPUI5 demo</title>
        <script id="sap-ui-bootstrap"
           src="/resources/sap-ui-core.js"
           data-sap-ui-theme="sap_platinum"
           data-sap-ui-libs="sap.ui.commons, sap.ui.table"></script>
        <script src="app/app.js"></script>
    </head>
    <body class="sapUiBody">
         <img src="images/sap_logo.png" >
         <div id="salesorders"></div>
         <div id="lineitems"></div>
         <img src="images/o365_logo.jpg" style="margin:30px 0 10px 0" />
         <div id="issues"></div>
    </body>
</html>

The JavaScript of our application is contained in app/app.js.

// Let's define some shortcuts to increase
// readability of the code
var ODataModel = sap.ui.model.odata.ODataModel,
    TextField = sap.ui.commons.TextField,
    TextView = sap.ui.commons.TextView,
    Label = sap.ui.commons.Label,
    DataTable = sap.ui.table.DataTable,
    Column = sap.ui.table.Column,
    SelectionMode = sap.ui.table.SelectionMode;
// Specify the SAP Gateway SalesOrder service as an OData model
var salesOrderService =
        "https://gw.esworkplace.sap.com/sap/opu/sdata/IWFND/SALESORDER",
    // The SalesOrder service requires authentication
    // get the username/password from the SDN page.
    username = "[username]",
    password = "[password]",
    // SAP Gateway only supports XML, so don't use JSON
    asJson = false,
    salesOrderModel = new ODataModel(salesOrderService, asJson, username, password)
    salesOrderCollection = "SalesOrderCollection";
// specify the SharePoint site containing the Sales Order Issues as an OData model
// we will assume there a SharePoint site called 'demo'
// which has an issues list called 'SalesOrderIssues'
var issueService =
        "http://[SharePoint_server]/demo/_vti_bin/ListData.svc",
    issueCollection = "SalesOrderIssues"  // name of SP List
    issueModel = new ODataModel(issueService);
// Create a master table with sales orders
var salesOrders = new DataTable({
    title: "Sales Orders",
    width: "100%",
    visibleRowCount: 5,
    selectionMode: SelectionMode.Single,
    editable: false
});
// define the relevant column properties
var salesOrderColumns = [
    { header: "Sales Order ID", value: "{SalesOrderID}", width: '100px' },
    { header: "Customer Name", value: "{CustomerName}", width: '50%' },
    { header: "Amount", value: "{TotalSum}", width: '50%' }
];
// create the columns
salesOrderColumns.forEach(function (column) {
    var label = new Label({ text: column.header }),
        template = new TextView({ text: column.value }),
        column = new Column({
            label: label,
            template: template,
            width: column.width
        });
    salesOrders.addColumn(column);
});
// connect the data table to the SalesOrder service
salesOrders.setModel(salesOrderModel);
// An OData request for the SalesOrderCollection
// will return the sales orders.
// Each sales order should result in a table row.
salesOrders.bindRows(salesOrderCollection);
// Put table in the DOM.
// placeAt will automatically defer if
// DOM is not ready yet (like in this demo).
salesOrders.placeAt("salesorders");
// At this point the Sales Order master table is done
// Creating the lineItems and issues table is very similar
// Creating the line items datatable
var lineItems = new DataTable({
    title: "Line items",
    width: "100%",
    visibleRowCount: 10,
    selectionMode: SelectionMode.Single,
    editable: false
});
lineItemColumns = [
    { header: "Line item #", value: "{SalesOrderItem}", width: '100px' },
    { header: "Product Name", value: "{ProductName}", width: '50%' },
    { header: "Amount", value: "{NetSum}", width: '50%' }
]
lineItemColumns.forEach(function (column) {
    var label = new Label({ text: column.header }),
    template = new TextView({ text: column.value }),
    column = new Column({
        label: label,
        template: template,
        width: column.width
    });
    lineItems.addColumn(column);
})
lineItems.setModel(salesOrderModel);
lineItems.placeAt("lineitems");
// Create the issues datatable
var issues = new DataTable({
    title: "Issues",
    width: "100%",
    visibleRowCount: 5,
    selectionMode: SelectionMode.Single,
    editable: false
});
issuesColumns = [
    { header: "Id", value: "{Id}", width: '30px' },
    { header: "Title", value: "{Title}", width: '40%' },
    { header: "Status", value: "{IssueStatusValue}", width: '10%' },
    { header: "Comments", value: "{Comments}", width: '50%' }
]
issuesColumns.forEach(function (column) {
    var label = new Label({ text: column.header }),
        template = new TextView({ text: column.value }),
        column = new Column({
            label: label,
            template: template,
            width: column.width
        });
    issues.addColumn(column);
});
issues.setModel(issueModel);
issues.placeAt("issues");
// The three data tables are ready!
// Now we need to define what should happen when
// the user selects a row in the sales order (master) table
salesOrders.attachRowSelect(function (event) {
    var Filter = sap.ui.model.Filter,
        FilterOperator = sap.ui.model.FilterOperator,
        selectedRowContext = event.getParameter("rowContext"),
        selectedSalesOrderID = salesOrderModel.getProperty("SalesOrderID", selectedRowContext),
        selectedSalesOrderLineItems = selectedRowContext + "/salesorderlineitems";
    // load the line items for the selected sales order
    lineItems.bindRows(selectedSalesOrderLineItems);
    // create a filter based on sales order ID
    var filter = new Filter("SalesOrderID", FilterOperator.EQ, 'SO:' + selectedSalesOrderID);
    // load the issues table using the filter
    issues.bindRows(issueCollection, null, [filter]);
});

Now, fire up your Chrome browser with the “disable-web-security” command line. This will suppress the Same Origin Policy which normally doesn’t allow you to do XHR request across domains:

SNAGHTML253f39d

Now you can view index.html in Chrome:

SNAGHTML24dc1e3

We have just build our first mash-up of SAP ERP and SharePoint data using SAPUI5!

Themes

SAPUI5 provides a number of themes. The theme used in this example is called ‘sap_platinum’. You can find other themes in the /resource/sap/ui/*/themes folders. Here are screenshots of the same page using the other themes (click the images to enlarge):

imageTheme: sap_goldreflection imageTheme: sap_ux
imageTheme: sap_hcb imageTheme: base

My first impressions on SAPUI5:

  • Offers a comprehensive set of UI widgets and layouts
  • Huge code base. Definitely not a micro-framework Smile.
  • Amazingly little code is required to get a demo app with OData sources described above up and running.
  • Includes jQuery (version 1.4.4. included in sap-core-ui.js)
  • Includes datajs.js(OData library by Microsoft)
  • Style and idiom is a bit verbose and seems to be influenced by Java. This may be a pro or a con, depending on where you are coming from.

Overall SAPUI5 is a very interesting entrant in the marketplace and sure to further boost the position as HTML5 as UI platform.

I hope this post is helpful in starting to explore the SAPUI5 Beta yourself.

Thanks for reading!

UPDATE (Oct 2012): SAPUI5 now includes mobile controls. See comment below from Ruben.