Creating JavaScript Functions in Active Reports

For those of us who love to (of just need to) expand Cognos beyond it’s built-in capabilities, one of the fundamental tools we use is the JavaScript function.  We use JavaScript to perform all kinds of reporting sorcery, and we create functions for those things that we call over and over again.For years this worked well.  Even after major releases, most of our functions continued to work with little or no modifications.

Then, one day IBM came out with Active Reports.  We were awed by it’s ability to make reports portable, and the fact that data was embedded into the reports also meant that the reports were super-fast when users changed filters or parameters.  But all of this came with a cost.  We lost the ability to create JavaScript functions and the <script> tags that should execute as soon as the report opened no longer functioned. We were still able to execute in-line JavaScript, but this had significant limitations.

Recently, I was reading a comment in LinkedIn by Paul Mendelson (aka CognosPaul) where he talked about creating a function when a user clicked a button.  Within the inline JavaScript, it checked to see if the function existed.  If it didn’t exist, it created it, otherwise it went on with the rest of the actions that the button-click demanded.  This did not satisfy my immediate requirement because I needed something to happen the moment the screen opened up, but it set me in the right direction.

How To Create Functions and Run JavaScript When the Report Opens

The key to this technique is the <img> tag which is one of the few to include an onLoad event.  This event fires as soon as the image is displayed.

Add an HTML item anywhere on the report, just make sure it’s somewhere that will always be rendered (on the page, but not inside any decks). Give the HTML item the following text:

<img src="hal/imagesirot/progress.gif" width="1" height="1" onload="
     while( ! alert) {} 
     /* Define the function */ 
     if (!document.testFunction ) { 
          testFunction = function() { 
               alert('We are running the test function'); 
     /* Run any startup actions */ 

The first line of the onload event is sure to raise hairs on the back of your head. “while( ! alert) {}” will endlessly loop until it detects the existence of the “alert” function.  This is necessary because in Active Reports, the image will load (and the onload event of the image will fire) before the document has completed loading.  JavaScript developers usually avoid this type of loop because JavaScipt runs synchronously, so the browser locks up until the loop is satisfied.  In my testing, this loop has never caused a problem, however, if someone can suggest a better alternative, I’m certainly open to trying it.

The image is hidden (style=”display:none;”), so it won’t affect your layout. I chose to use the progress.gif image because it’s always loaded.  You can use your own image, however I would not recommend a 1×1 transparent gif.  I tested this and it worked fine on Windows (IE, Firefox and Chrome), but did not work correctly on Cognos Mobile (iPad or Android).  If you do use your own image, it must also be somewhere in the report using a standard Cognos Image element from the toolbox, so that the image is properly registered.  Otherwise the image will only work while you are online.

The first thing you should do following the “while” loop is to define any functions you need.  In the example above I’ve demonstrated the “testFunction”.After that you can run any JavaScript that needs to run when the report first opens.Coming soon… getElementById in Active Reports.

2014-09-17 – Added details about the images that can be used and the placement of the HTML Item.

Changing a Value Prompt in Report Studio using JavaScript

There have been many occasions where I have needed to change the filters on a report by having the user click on data on the report.  For example if I had a listing of cars, class, and number sold, and the user clicked on “Sedan” under the Class column, I would want to filter the report to only show sedans.

In Cognos 8 Report Studio, I accomplished this by having  a drill-through on the Class column back to the same report, passing the data item into the Class prompt.  This was an imperfect solution for a number of reasons, but it was usually good enough.  However, Cognos 10’s introduction of Business Insight  made this technique a bit less than “good enough”.  If I were to add the report to a Business Insight workspace, clicking on “Sedan” would not update the report I was looking at.  Instead, it would open a new copy of the report in a new window.

So the drill-through technique is no longer an option. My new technique for accomplishing this is to use Javascript to change the prompt. The following technique works with Cognos 10.1.1 Report Studio and I believe it’ll work with Cognos 8 as well. However, remember that with any JavaScript used in Cognos, an update to another version of Cognos (even a minor one) could break it.

Set Up the Prompt

This technique requires that you add the prompt to your report page (regardless of whether or not you have a copy of the prompt on a  prompt page).  I like to put these near the top of the report so the user can immediately see what filters are being applied.  However, you can put it anywhere on the report page that you want. 

  1. Add the prompt to the report page
  2. Before the prompt, add an HTML item with the following HTML code:
    <div id=”VP_TestPrompt”>

    Give the div tag an appropriate ID.  If you have more than one prompt, give each one a unique name.

  3. After the prompt, add an HTML item to close the div tag:
  4. After the closing div tag, add a third HTML item with the following function:
    <script type=”text/javascript”>
    function selectPromptByValue(divId, val) {
       var oVPCont = document.getElementById(divId);
       var oVP = oVPCont.getElementsByTagName(“select”);
       var i = oVP[0].options.length;
          while (i) {
          if (oVP[0].options[–i].value==val) {

    If you have multiple prompts, you will only need this HTML item once on the report page.  The first parameter tells it which prompt to reference.

Set Up the Link

Now you just need to call the function selectPromptByValue, passing in the name of the div tag (set in step 2 above) and the value you want to pass into the value prompt.
Ther are a couple of ways to do this.  For most reports you could use a hyperlink to call the function (ex: javascript:selectPromptByValue(“VP_TestPrompt”, “A”);).  However, this does not work when running the report from a Business Insight workspace because BI incorrectly assumes that the hyperlink (anchor tag) will be taking you to another web page.  This results in a warning about navigating away from the page.

Instead I use a simple text object and wrap it in another div tag with an onclick event.

  1. Add your text.  To make it look more like a hyperlink, you may wish to underline it and set the color to blue.
  2. Add an HTML item before the text:
     <div style=”cursor:pointer”; onclick=’javascript:selectPromptByValue(“VP_TestPrompt”, “A”);’>

    The first parameter should be the id of the div tag you set before the prompt.  The second parameter is the value in the prompt that you want to select.
    If the link is to be data driven (ex: in a list), then you will need to make the ULR Source Type a report expression and parse it together so that it looks like the above example.

  3. Add an HTML item after the text to close the div tag:

Example Report

To see how this technique works, copy the text from this Word document into your clipboard, open Report Studio, and select Open Report from Clipboard.

How To Display Data From a Cognos 8 Report in Google Maps

IBM’s Cognos 8 Report Studio already has built-in maps that allow you to create geographic representations of your data, but they are not without limitations, the most notable being that there are only about a half dozen maps included.   The world-level maps contain every location, but not down to the level of granularity you often need (zip-code level for example).  There are U.S.-only maps that go down to the zip code, but they don’t include U.S. Posessions (ex: Puerto Rico) You can develop custom maps, but the software required to generate the MapInfo files is quite expensive and outside of the range of many project budgets.

One solution is to integrate your reports with Google Maps via the Google Maps API.  The API is capable of marking any location in the world, by a variety of means, including zip-codes, city names, or even latitude and longitude.  The interactivity is about as intuitive as it gets, and you can add markers to highlight specific map points or polygons to outline and color areas of any shape (a state for example)

This article follows-up on the example provided by Rayudu Vecha and Sridhar Munireddy in their developerWorks article; Integrate Google Maps with Cognos 8.  In it, they explain how to call the Google Maps API from within a Cognos 8 Report Studio report and insert hard-coded markers.  Here, I will show you how to expand that to dynamically populate markers on top of the Google Map.

The basic technique is fairly simple.  I take the single HTML Item that Vecha and Munireddy use (enhanced a bit for my particular needs) and then divide it into four separate HTML Items, with one of them nested inside a repeater object to handle an unknown number of rows from a query. Optionally, the third HTML can be embedded inside a singleton to allow it to retrieve from the database a value that will be put into a JavaScript variable.

When the report is run, the four separate HTML items act as a single block of HTML code, but displayed within the Cognos 8 report viewer.

What You’ll Need:
  • A Cognos 8 Business Intelligence server
  • A Google Maps API Key  for the Cognos 8 BI server’s domain (directions for obtaining this are below)
  • Basic development skills for Cognos 8 Report Studio
  • A fundamental understanding of JavaScript
  • A database that contains the latitude and longitute of the locations you wish to mark on the map.

The Steps:
  1. You are going to need a Google Maps API key.

    1. Go to
    2. Enter the domain name at which you access Cognos 8 (ex:
    3. Click the Generate API Key button.This should create a key that looks similar to this:

    4. Copy the API key displayed and save it for future use.
    5. If you can, store this key somewhere in your reporting database.  This will help if you plan to later migrate the report to another server (ex: from development to production)
  2. Open Report Studio and create a new report using the blank template.
  3. From the insertable objects toolbox, add a Repeater object to the body of the report.
  4. Right-click on the new repeater object and select Go To Query.
  5. Add to the query the Data Items you want to display.This data will be used to assemble a Javascript array (in a subsequent step).You will also need to include two columns containing the latitude and longitude of each marker.

  6. If you plan to have more than one row in the query (i.e. more than one marker on the map) you will need a column that contains a comma for all but the first row.  This comma will be used as a separator for the elements in the Javascript array.  To Do this:
    1. Add a Data Item to the query named Row. Set the expression to:

      rank ([Location])

      (replacing “[Location]” with a column you’ll be sorting on)

    2. Add another Data Item named ElementSeparator. Set the expression to:

      IF ( [Row] =  1 ) THEN
      ( ” )
      ( ‘, ‘ )

  7. Return to the page (Page1).
  8. Click on the repeater and edit the Data->Properties.Click check-boxes by each data item that you will want to use.  This will allow you to use them in a report expression.

  9. Change the repeater’s Data->Rows Per Page to 99999(otherwise, Report Studio will try to create a new page for every 20 rows in your repeater)

  10. You may also need to define the Data->Grouping and Sorting of the repeater.
  11. Add an HTML Item immediately to the left of the repeaterThis will contain the start of the Javascript

    1. The source type will be the default: Text
    2. Set the HTML text to the following:



      <title>Static -> Dynamic Google Map with Markers</title>


      <style type=”text/css”>

      .message {

      font-size: 80%;

      padding: 2px;

      font-weight: bold;

      text-align: center;

      width: 550px;

      height: 450px;

      background-color: #ffcc00;

      border: 1px solid white;

      position: relative;



      <script type=”text/javascript”>

      var locations = [

  12. Add a second HTML Item inside the repeaterThis will be the assembly of the array containing your data

    1. Set the source type to: Report Expression
    2. Set the report expression to the following, modifying it to fit your data:

      [Query1].[ElementSeperator] + ‘{name: ”’ +

      [Query1].[Location] + ”’, url: ”’ + [Query1].[URL] +  ”’, headcount: ‘ + number2string ( [Query1].[Headcount] ) + ‘, hires: ‘ + number2string ( [Query1].[Hires] ) + ‘, terminations: ‘ + number2string ( [Query1].[Terminations] ) + ‘, promotions: ‘ + number2string ( [Query1].[Promotions] ) + ‘, lat: ‘ + number2string ( [Query1].[Latitude] ) + ‘, lng: ‘ + number2string ( [Query1].[Longitude] ) + ‘} ‘

      The resulting data would look something like this (line breaks and indentation added for readability):

      name: ‘Location 1’,
      url: ‘http://url.for.location.1&#8217;,
      headcount: 123,
      hires: 1,
      terminations: 2,
      promotions: 3,
      lat: 38.88944,
      lng: -77.035341

      } ,


      name: ‘Location 2’,

      url: ‘http://url.for.location.2&#8217;,

      headcount: 456,

      hires: 7,

      terminations: 8,

      promotions: 9,

      lat: 43.878996,

      lng: -103.459811


  13. Immediately to the right of the repeater, add a third HTML Item
    (Note: if you stored the API Key in your database, add a Singleton with a query that retrieves it and put the HTML Item inside the singleton, and use a report expression to define the text)

    1. The source type will be the default: Text
    2. Set the HTML text to the following:


      var strMapKey = ‘


      (substituting your own API Key)

  14. Add a fourth HTML Item immediately to the right of the third one.This will contain the end of (and the bulk of) the Javascript

    1. The source type will be the default: Text
    2. Set the HTML Text to the following:

      var map = null;

      var bounds = null;

      var currentMarker = null;

      var mapDiv = null;

      var containerDiv = null;

      var clickedX = 0;

      var clickedY = 0;

      var isLoaded = false;


      * Called after script is asynchronously loaded in.

      * Creates the GMap2, GMarker objects and performs actions according to

      * what the user did to trigger the map load (search, zoom, click etc).


      function loadMap() {

      if (GBrowserIsCompatible()) { = ‘#fff’; = ”;

      map = new GMap2(mapDiv, {logoPassive: true});

      bounds = new GLatLngBounds();

      for (var i = 0; i < locations.length; i++) {

      bounds.extend(new GLatLng(locations[i].lat, locations[i].lng));


      var latSpan = bounds.toSpan().lat();

      map.setCenter(bounds.getCenter(), map.getBoundsZoomLevel(bounds));

      map.addControl(new GLargeMapControl());


      var mapControl = new GMapTypeControl();


      var newBounds = map.getBounds();

      var newLatSpan = newBounds.toSpan().lat();

      if (latSpan/newLatSpan > .90) { map.zoomOut(); }

      for (var i = 0; i < locations.length; i++) {

      var marker = createMarker(i);

      var latlng = marker.getLatLng();

      var pixel = map.fromLatLngToDivPixel(latlng);

      if (Math.abs(pixel.x -clickedX) < 12 && Math.abs(pixel.y -clickedY) < 20){

      //GEvent.trigger(marker, ‘click’);



      map.setCenter(bounds.getCenter(), map.getBoundsZoomLevel(bounds));

      }  } }

      function zoomToAll() {

      map.setCenter(bounds.getCenter(), map.getBoundsZoomLevel(bounds));



      * Creates a marker for the given location.

      * @param {Number} ind

      * @return {GMarker}


      function createMarker(ind) {

      var value= 103037;

      var location = locations[ind];

      var marker = new GMarker(new GLatLng(, location.lng));


      * When the user clicks on the marker while zoomed out, it zooms in (in case there is

      * another marker nearby that would not be visable when zoomed out)

      * When they click on the marker while zoomed in, it opens an Info Window that displays

      * details about the location


      GEvent.addListener(marker, ‘click’, function() {

      marker.html = [‘<b>’,, ‘</b>’,


      ‘<li>Headcount: ‘, location.headcount, ‘</li>’,

      ‘<li>New Hires: ‘, location.hires, ‘</li>’,

      ‘<li>Terminations: ‘, location.terminations, ‘</li>’,

      ‘<li>Promotions: ‘, location.promotions, ‘</li>’,


      ‘<a  href=”‘ + location.url + ‘”  target=”_blank “> Detail Listing</a>’].join(”);

      currentMarker = marker;

      if ( map.getZoom() <= 12 ) {

      map.setCenter( marker.getLatLng(), 15 );


      else {




      return marker;



      * Formats location info into a URL-friendly version for maps url.

      * @param {Object} location

      * @return {String}


      function formatAddressForMaps(location) {

      var address = location.street + ‘ ‘ + + ‘ ‘ + location.state + ‘ ‘


      return escape(address.replace(‘ ‘, ‘+’));


      function _cel(elementType, id) {

      var element = document.createElement(elementType); = id;

      return element;



      * Loads in the Maps API script. This is called after some sort of user interaction.

      * The script loads asynchronously and calls loadMap once it’s in.


      function loadScript() {

      if (!isLoaded) {

      isLoaded = true;

      var div = document.createElement(‘div’);

      div.className = ‘message’;

      div.innerHTML = ‘Loading…’; = (800/2 – 53) + ‘px’; = 800/2 + ‘px’;


      var script = document.createElement(‘script’);

      script.type = ‘text/javascript’;

      script.src = ‘; +

      ‘&async=2&callback=loadMap&key=’ + strMapKey;





      * Sets up the gadget by setting CSS and click events.


      function loadMapGadget() {

      containerDiv = document.getElementById(‘container’);

      mapDiv = document.getElementById(‘map’);

      mapDiv.onclick = function (e) {

      clickedX = (window.event && window.event.offsetX) || e.clientX;

      clickedY = (window.event && window.event.offsetY) || e.clientY;



      loadScript(); = ‘pointer’;

      var urlString = [‘’%5D;

      var markerString = [];

      for (var i = 0; i < locations.length; i++) {

      markerString.push(locations[i].lat + ‘,’ + locations[i].lng + ‘,red’);




      urlString.push(‘&key= ‘ + strMapKey); = ‘url(\” + urlString.join(”) + ‘\’)’;




      <body onload=”loadMapGadget();”>

      <div id=”container”>

      <div id=”map” style=”width: 800; height: 600; overflow:hidden”></div>




  15. Save and test the report.


Integrating Cognos 8 Report Studio data into Google Maps might not be quite as straight-forward as using the built in maps, but it’s not too difficult and the results are sure to impress your report consumers.  This is also a great starting point for learning to integrate Cognos 8 with other web services.