Home > Articles > Web Services > XML

XML Reference Guide

📄 Contents

  1. XML Reference Guide
  2. Overview
  3. What Is XML?
  4. Informit Articles and Sample Chapters
  5. Books and e-Books
  6. Official Documentation
  7. Table of Contents
  8. The Document Object Model
  9. Informit Articles and Sample Chapters
  10. Books and e-Books
  11. Official Documentation
  12. DOM and Java
  13. Informit Articles and Sample Chapters
  14. Books and e-Books
  15. Implementations
  16. DOM and JavaScript
  17. Using a Repeater
  18. Repeaters and XML
  19. Repeater Resources
  20. DOM and .NET
  21. Informit Articles and Sample Chapters
  22. Books and e-Books
  23. Documentation and Downloads
  24. DOM and C++
  25. DOM and C++ Resources
  26. DOM and Perl
  27. DOM and Perl Resources
  28. DOM and PHP
  29. DOM and PHP Resources
  30. DOM Level 3
  31. DOM Level 3 Core
  32. DOM Level 3 Load and Save
  33. DOM Level 3 XPath
  34. DOM Level 3 Validation
  35. Informit Articles and Sample Chapters
  36. Books and e-Books
  37. Documentation and Implementations
  38. The Simple API for XML (SAX)
  39. Informit Articles and Sample Chapters
  40. Books and e-Books
  41. Official Documentation
  42. SAX and Java
  43. Informit Articles and Sample Chapters
  44. Books and e-Books
  45. SAX and .NET
  46. Informit Articles and Sample Chapters
  47. SAX and Perl
  48. SAX and Perl Resources
  49. SAX and PHP
  50. SAX and PHP Resources
  51. Validation
  52. Informit Articles and Sample Chapters
  53. Books and e-Books
  54. Official Documentation
  55. Document Type Definitions (DTDs)
  56. Informit Articles and Sample Chapters
  57. Books and e-Books
  58. Official Documentation
  59. XML Schemas
  60. Informit Articles and Sample Chapters
  61. Books and e-Books
  62. Official Documentation
  63. RELAX NG
  64. Informit Articles and Sample Chapters
  65. Books and e-Books
  66. Official Documentation
  67. Schematron
  68. Official Documentation and Implementations
  69. Validation in Applications
  70. Informit Articles and Sample Chapters
  71. Books and e-Books
  72. XSL Transformations (XSLT)
  73. Informit Articles and Sample Chapters
  74. Books and e-Books
  75. Official Documentation
  76. XSLT in Java
  77. Java in XSLT Resources
  78. XSLT and RSS in .NET
  79. XSLT and RSS in .NET Resources
  80. XSL-FO
  81. Informit Articles and Sample Chapters
  82. Books and e-Books
  83. Official Documentation
  84. XPath
  85. Informit Articles and Sample Chapters
  86. Books and e-Books
  87. Official Documentation
  88. XML Base
  89. Informit Articles and Sample Chapters
  90. Official Documentation
  91. XHTML
  92. Informit Articles and Sample Chapters
  93. Books and e-Books
  94. Official Documentation
  95. XHTML 2.0
  96. Documentation
  97. Cascading Style Sheets
  98. Informit Articles and Sample Chapters
  99. Books and e-Books
  100. Official Documentation
  101. XUL
  102. XUL References
  103. XML Events
  104. XML Events Resources
  105. XML Data Binding
  106. Informit Articles and Sample Chapters
  107. Books and e-Books
  108. Specifications
  109. Implementations
  110. XML and Databases
  111. Informit Articles and Sample Chapters
  112. Books and e-Books
  113. Online Resources
  114. Official Documentation
  115. SQL Server and FOR XML
  116. Informit Articles and Sample Chapters
  117. Books and e-Books
  118. Documentation and Implementations
  119. Service Oriented Architecture
  120. Web Services
  121. Informit Articles and Sample Chapters
  122. Books and e-Books
  123. Official Documentation
  124. Creating a Perl Web Service Client
  125. SOAP::Lite
  126. Amazon Web Services
  127. Creating the Movable Type Plug-in
  128. Perl, Amazon, and Movable Type Resources
  129. Apache Axis2
  130. REST
  131. REST Resources
  132. SOAP
  133. Informit Articles and Sample Chapters
  134. Books and e-Books
  135. Official Documentation
  136. SOAP and Java
  137. Informit Articles and Sample Chapters
  138. Books and e-Books
  139. Official Documentation
  140. WSDL
  141. Informit Articles and Sample Chapters
  142. Books and e-Books
  143. Official Documentation
  144. UDDI
  145. UDDI Resources
  146. XML-RPC
  147. XML-RPC in PHP
  148. Informit Articles and Sample Chapters
  149. Books and e-Books
  150. Official Documentation
  151. Ajax
  152. Asynchronous Javascript
  153. Client-side XSLT
  154. SAJAX and PHP
  155. Ajax Resources
  156. JSON
  157. Ruby on Rails
  158. Creating Objects
  159. Ruby Basics: Arrays and Other Sundry Bits
  160. Ruby Basics: Iterators and Persistence
  161. Starting on the Rails
  162. Rails and Databases
  163. Rails: Ajax and Partials
  164. Rails Resources
  165. Web Services Security
  166. Web Services Security Resources
  167. SAML
  168. Informit Articles and Sample Chapters
  169. Books and e-Books
  170. Specification and Implementation
  171. XML Digital Signatures
  172. XML Digital Signatures Resources
  173. XML Key Management Services
  174. Resources for XML Key Management Services
  175. Internationalization
  176. Resources
  177. Grid Computing
  178. Grid Resources
  179. Web Services Resource Framework
  180. Web Services Resource Framework Resources
  181. WS-Addressing
  182. WS-Addressing Resources
  183. WS-Notifications
  184. New Languages: XML in Use
  185. Informit Articles and Sample Chapters
  186. Books and e-Books
  187. Official Documentation
  188. Google Web Toolkit
  189. GWT Basic Interactivity
  190. Google Sitemaps
  191. Google Sitemaps Resources
  192. Accessibility
  193. Web Accessibility
  194. XML Accessibility
  195. Accessibility Resources
  196. The Semantic Web
  197. Defining a New Ontology
  198. OWL: Web Ontology Language
  199. Semantic Web Resources
  200. Google Base
  201. Microformats
  202. StructuredBlogging
  203. Live Clipboard
  204. WML
  205. XHTML-MP
  206. WML Resources
  207. Google Web Services
  208. Google Web Services API
  209. Google Web Services Resources
  210. The Yahoo! Web Services Interface
  211. Yahoo! Web Services and PHP
  212. Yahoo! Web Services Resources
  213. eBay REST API
  214. WordML
  215. WordML Part 2: Lists
  216. WordML Part 3: Tables
  217. WordML Resources
  218. DocBook
  219. Articles
  220. Books and e-Books
  221. Official Documentation and Implementations
  222. XML Query
  223. Informit Articles and Sample Chapters
  224. Books and e-Books
  225. Official Documentation
  226. XForms
  227. Informit Articles and Sample Chapters
  228. Books and e-Books
  229. Official Documentation
  230. Resource Description Framework (RDF)
  231. Informit Articles and Sample Chapters
  232. Books and e-Books
  233. Official Documentation
  234. Topic Maps
  235. Informit Articles and Sample Chapters
  236. Books and e-Books
  237. Official Documentation, Implementations, and Other Resources
  238. Rich Site Summary (RSS)
  239. Informit Articles and Sample Chapters
  240. Books and e-Books
  241. Official Documentation
  242. Simple Sharing Extensions (SSE)
  243. Atom
  244. Podcasting
  245. Podcasting Resources
  246. Scalable Vector Graphics (SVG)
  247. Informit Articles and Sample Chapters
  248. Books and e-Books
  249. Official Documentation
  250. OPML
  251. OPML Resources
  252. Summary
  253. Projects
  254. JavaScript TimeTracker: JSON and PHP
  255. The Javascript Timetracker
  256. Refactoring to Javascript Objects
  257. Creating the Yahoo! Widget
  258. Web Mashup
  259. Google Maps
  260. Indeed Mashup
  261. Mashup Part 3: Putting It All Together
  262. Additional Resources
  263. Frequently Asked Questions About XML
  264. What's XML, and why should I use it?
  265. What's a well-formed document?
  266. What's the difference between XML and HTML?
  267. What's the difference between HTML and XHTML?
  268. Can I use XML in a browser?
  269. Should I use elements or attributes for my document?
  270. What's a namespace?
  271. Where can I get an XML parser?
  272. What's the difference between a well-formed document and a valid document?
  273. What's a validating parser?
  274. Should I use DOM or SAX for my application?
  275. How can I stop a SAX parser before it has parsed the entire document?
  276. 2005 Predictions
  277. 2006 Predictions
  278. Nick's Book Picks

Normally, when we think of Web services, we think about XML going back and forth. That's only natural, given the preponderance of SOAP and REST of applications, but it isn't actually the only way to do things. Today we're going to talk about JavaScript Object Notation (JSON). JSON is a way to represent the data and methods of an object in a text string, which enables you to reconstitute it as an object on demand.

There are several advantages to using JSON over XML. For one thing, it's much less verbose, which is good if you are passing large amounts of data. For another thing, your logic and structure or already in place when he received the object. You don't need to re-create them. Also, as we'll see in this section, there are certain advantages in terms of Web applications and the avoidance of cross domain restrictions.

On the other hand, JSON can be very difficult to read if you're a human instead of a computer. For example, consider this representation of time data for a project:

{"project": {"ProjectName":"InformIT Column",
                 "TotalTime":"3289",
                 "LastWorkedOn": {
                     "Day":24,
                     "Month":1,
                     "Year":2006 },
                 "TimeThisWeek":"135",
                 "TimeToday":"20",
                 "CurrentProject":"JSON Guide Entry"}}

Now, I'm sure that there are people who have no trouble separating the labels (such as ProjectName) from the data (such as InformIT Column), but I'm not one of them. But let's look at the structure so we can see how this works.

We start with the fact that the entire object is placed within curly braces ({}). Then we have the actual object name, project, followed by the contents of the object, also in braces. Each member is signified by a label, the colon, and its value. In some cases, we have a complex object. For example, LastWorkedOn is actually an object itself, as you can tell because its value is enclosed in curly braces. That object has three members, Day, Month, and Year. And

Let's see how this would actually work in JavaScript. Start by creating a basic page with a script:

<html>
<head>
  <title>JSON testing</title>

  <script type="text/javascript">

  function showObject(){

     var str = '{"project": {'+
                   '"ProjectName":"InformIT Column",'+
                   '"TotalTime":"3289",'+
                   '"LastWorkedOn": {'+
                   '      "Day":24,'+
                   '      "Month":1,'+
                   '      "Year":2006 },'+
                   '"TimeThisWeek":"135",'+
                   '"TimeToday":"20",'+
                   '"CurrentProject":"JSON Guide Entry"}}';

     var projectObj = eval("("+str+")");
     document.write("Project is: "+projectObj.project.ProjectName);
     document.write("<br />");
     document.write("Last worked on: "+
                          projectObj.project.LastWorkedOn.Month + "/" +
                          projectObj.project.LastWorkedOn.Day + "/" +
                          projectObj.project.LastWorkedOn.Year);

  }

  </script>


</head>
<body>

  <h1>JavaScript Object Notation</h1>

  <script type="text/javascript">
    showObject();
  </script>

</body>
</html>

Year we've created a simple script that includes the text that represents the object. We then use the eval () function to convert the text (enclosed in parentheses) back into an object. (If you're concerned about the security implications of using the eval() function, fear not. We'll get to that shortly.) Once we have the object, we can access its data just as we would access the data of any other JavaScript object. Here you see the project name, and the contents of the LastWorkedOn object. Finally, we call the script from within the body of the page. The results look like this:

Basic page

We can also use JSON to specify arrays. For example, we might have a projects object with more than one project:

...
  function showArrayOfObjects(){

     var str = '{"project": [{'+
                   '"ProjectName":"InformIT Column",'+
                   '"TotalTime":"3289",'+
                   '"LastWorkedOn": {'+
                   '      "Day":24,'+
                   '      "Month":1,'+
                   '      "Year":2006 },'+
                   '"TimeThisWeek":"135",'+
                   '"TimeToday":"20",'+
                   '"CurrentProject":"JSON Guide Entry"},'+
                   '{'+
                   '"ProjectName":"Vitamin Store Site",'+
                   '"TotalTime":"42",'+
                   '"LastWorkedOn": {'+
                   '      "Day":20,'+
                   '      "Month":1,'+
                   '      "Year":2006 },'+
                   '"TimeThisWeek":"2",'+
                   '"TimeToday":"0",'+
                   '"CurrentProject":"Client Review"}]}';


     var projectObj = eval("("+str+")");
     document.write("First project is: "+projectObj.project[0].ProjectName);
     document.write("<br />");
     document.write("Last worked on: "+
                          projectObj.project[0].LastWorkedOn.Month + "/" +
                          projectObj.project[0].LastWorkedOn.Day + "/" +
                          projectObj.project[0].LastWorkedOn.Year);


     document.write("<br /><br />");

     document.write("Second project is: "+projectObj.project[1].ProjectName);
     document.write("<br />");
     document.write("Last worked on: "+
                          projectObj.project[1].LastWorkedOn.Month + "/" +
                          projectObj.project[1].LastWorkedOn.Day + "/" +
                          projectObj.project[1].LastWorkedOn.Year);


  }
  </script>

</head>
<body>

  <h1>JavaScript Object Notation</h1>

  <script type="text/javascript">
    showArrayOfObjects();
  </script>

</body>
</html>

Notice first of all that I have enclosed the individual elements of the array in brackets ([]). In this case, we're dealing with an array of objects, but it could just as easily have been an array of strings or numbers, as long as everything is enclosed in brackets and separated by commas.

Once we have the object, again we work with it like any other object; we are specifying a specific element of the array, which gives us the object in that position, which we treat like any other object. The results look like this:

Basic arrays

That's the basic idea, but now we need to do with this eval() thing. You see, this function will execute any code you give it, even if it comes in as a string. This is terrific for executing dynamically generated code, but terrifying if you are getting the text from an outside source. So in general, it's not a good idea to create your objects with a raw call to the eval() function. Instead, you'll want to use a JSON parser. You can download one from here. In fact, you can download JSON parsers for virtually any language from JSON.org.

Once you have the parser, you can use it to create your objects:

<html>
<head>
  <title>JSON testing</title>

  <script type="text/javascript" src="./json.js" ></script>

  <script type="text/javascript">

  function showArrayOfObjectsWithParser(){

     var str = '{"project": [{'+
...
                   '"CurrentProject":"Client Review"}]}';


     var projectObj = JSON.parse(str);
     document.write("First project is: "+projectObj.project[0].ProjectName);
     document.write("<br />");
...
  }

  </script>

</head>
<body>

  <h1>JavaScript Object Notation</h1>

  <script type="text/javascript">
    showArrayOfObjectsWithParser();
  </script>

</body>
</html>

Once we include the parser with its own script tag, the only change is the way in which we create the object. Notice that we are no longer enclosing a string in parentheses. Other than that, everything is exactly the same.

Okay, now that we know how to use these objects, let's put them to work in a more practical situation.

Yahoo Web Services now gives you the option to output JSON text rather than XML. For example, the query:

http://api.search.yahoo.com/ImageSearchService/V1/imageSearch?appid=YahooDemo&query=Madonna&results=2&output=json

(scroll to the right to see the new parameter)

will return the string:

{"ResultSet":{
            "totalResultsAvailable":"305090",
            "totalResultsReturned":2,
            "firstResultPosition":1,
            "Result":[{
                 "Title":"madonna 118",
                 "Summary":"Picture 118 of 184",
                 "Url":"http:\/\/www.celebritypicturesarchive.com\/pictures\/m\/madonna\/madonna-118.jpg",
                 "ClickUrl":"http:\/\/www.celebritypicturesarchive.com\/pictures\/m\/madonna\/madonna-118.jpg",
                 "RefererUrl":"http:\/\/www.celebritypicturesarchive.com\/pgs\/m\/Madonna\/Madonna%20picture_118.htm",
                 "FileSize":"40209",
                 "FileFormat":"jpeg",
                 "Height":"700",
                 "Width":"473",
                 "Thumbnail":{
                        "Url":"http:\/\/re2.mm-b1.yimg.com\/image\/500892420",
                        "Height":"130",
                         "Width":"87"}},
                {...}]}}

instead of the XML:

<?xml version="1.0" encoding="UTF-8"?>
<ResultSet xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="urn:yahoo:srchmi"  
  xsi:schemaLocation="urn:yahoo:srchmi http://api.search.yahoo.com/ImageSearchService/V1/ImageSearchResponse.xsd" 
  totalResultsAvailable="305090" 
  totalResultsReturned="2" 
  firstResultPosition="1">
  <Result>
    <Title>madonna 118</Title>
    <Summary>Picture 118 of 184</Summary>

    <Url>http://www.celebritypicturesarchive.com/pictures/m/madonna/madonna-118.jpg</Url>
    <ClickUrl>http://www.celebritypicturesarchive.com/pictures/m/madonna/madonna-118.jpg</ClickUrl>
    <RefererUrl>http://www.celebritypicturesarchive.com/pgs/m/Madonna/Madonna%20picture_118.htm</RefererUrl>
    <FileSize>40209</FileSize>

    <FileFormat>jpeg</FileFormat>
    <Height>700</Height>
    <Width>473</Width>
    <Thumbnail>

        <Url>http://re2.mm-b1.yimg.com/image/500892420</Url>
        <Height>130</Height>
        <Width>87</Width>
    </Thumbnail>

  </Result>
  <Result>...</Result>
</ResultSet>

The data is the same, but it's returned in JSON format.

Okay, so how do we use it? Well, we use it by feeding it to a script, just as we would use any other data. For example, we can create a script that extract the number of results and adds it to an empty div:

<html>
<head>
  <title>JSON testing</title>

  <script type="text/javascript" src="./json.js" ></script>

  <script type="text/javascript">

    function getNumberOfResults(){

      var yahooStr = '{"ResultSet":{"totalResultsAvailable":"305090","totalResultsReturned":2,"firstResultPosition":1,"Result":[{"Title":"madonna 118","Summary":"Picture 118 of 184","Url":"http:\/\/www.celebritypicturesarchive.com\/pictures\/m\/madonna\/madonna-118.jpg","ClickUrl":"http:\/\/www.celebritypicturesarchive.com\/pictures\/m\/madonna\/madonna-118.jpg","RefererUrl":"http:\/\/www.celebritypicturesarchive.com\/pgs\/m\/Madonna\/Madonna%20picture_118.htm","FileSize":"40209","FileFormat":"jpeg","Height":"700","Width":"473","Thumbnail":{"Url":"http:\/\/re2.mm-b1.yimg.com\/image\/500892420","Height":"130","Width":"87"}},{"Title":"madonna britney spe 111697a","Summary":"Britney og Madonna kom sv\u00e6rt n\u00e6r hverandre under MTVutdelingen","Url":"http:\/\/pub.tv2.no\/multimedia\/TV2\/archive\/00111\/madonna_britney_spe_111697a.jpg","ClickUrl":"http:\/\/pub.tv2.no\/multimedia\/TV2\/archive\/00111\/madonna_britney_spe_111697a.jpg","RefererUrl":"http:\/\/pub.tv2.no\/TV2\/underholdning\/article140732.ece","FileSize":"21801","FileFormat":"jpeg","Height":"200","Width":"250","Thumbnail":{"Url":"http:\/\/re2.mm-c1.yimg.com\/image\/979408253","Height":"88","Width":"110"}}]}}';
      var yahooObj = JSON.parse(yahooStr);
      document.getElementById("resultDiv").innerHTML = 
          "<b>Results: </b>"+yahooObj.ResultSet.totalResultsAvailable;

    }
  </script>


</head>
<body>

  <h1>JavaScript Object Notation</h1>

  <div id="resultDiv"></div>

  <a href="javascript:getNumberOfResults()">Get Results</a>

</body>
</html>

Starting at the bottom, we created the div, and a link to the getNumberOfResults() function. That function has the hardcoded string that represents the object (we'll change that in a minute), parses it, and adds the results to the div. When you first load the page, all you will see is the link. Once you click the link, you'll see the results:

The results

That's all well and good, but how do we get the information loaded with the page? Well, consider this page:

<html>

<head>
  <title>JSON testing</title>

  <script type="text/javascript" src="./json.js" ></script>

  <script type="text/javascript">

    function showNumberOfResults(yahooStr){

      var yahooObj = JSON.parse(yahooStr);
      document.getElementById("resultDiv").innerHTML = 
          "<b>Results: </b>"+yahooObj.ResultSet.totalResultsAvailable;

    }

  </script>


</head>
<body>

  <h1>JavaScript Object Notation</h1>

  <div id="resultDiv"></div>

  <script type="text/javascript">
     var yahooStr = '{"ResultSet":{"totalResultsAvailable":"305090","totalResultsReturned":2,"firstResultPosition":1,"Result":[{"Title":"madonna 118","Summary":"Picture 118 of 184","Url":"http:\/\/www.celebritypicturesarchive.com\/pictures\/m\/madonna\/madonna-118.jpg","ClickUrl":"http:\/\/www.celebritypicturesarchive.com\/pictures\/m\/madonna\/madonna-118.jpg","RefererUrl":"http:\/\/www.celebritypicturesarchive.com\/pgs\/m\/Madonna\/Madonna%20picture_118.htm","FileSize":"40209","FileFormat":"jpeg","Height":"700","Width":"473","Thumbnail":{"Url":"http:\/\/re2.mm-b1.yimg.com\/image\/500892420","Height":"130","Width":"87"}},{"Title":"madonna britney spe 111697a","Summary":"Britney og Madonna kom sv\u00e6rt n\u00e6r hverandre under MTVutdelingen","Url":"http:\/\/pub.tv2.no\/multimedia\/TV2\/archive\/00111\/madonna_britney_spe_111697a.jpg","ClickUrl":"http:\/\/pub.tv2.no\/multimedia\/TV2\/archive\/00111\/madonna_britney_spe_111697a.jpg","RefererUrl":"http:\/\/pub.tv2.no\/TV2\/underholdning\/article140732.ece","FileSize":"21801","FileFormat":"jpeg","Height":"200","Width":"250","Thumbnail":{"Url":"http:\/\/re2.mm-c1.yimg.com\/image\/979408253","Height":"88","Width":"110"}}]}}';
  </script>

  <a href="javascript:showNumberOfResults(yahooStr)">Get Results</a>


</body>
</html>

Because of the number of nested quotation marks, I've created a variable and called it from the function, but you can see that all we're doing here is passing the JSON object to a function. Now, what if we could have the request for data call the function itself? It turns out that if we add a callback parameter to the request, we can actually have a call a function. For example:

<html>
<head>
  <title>JSON testing</title>

  <script type="text/javascript" src="./json.js" ></script>

  <script type="text/javascript">

    function showNumberOfResults(yahooObj){

      document.getElementById("resultDiv").innerHTML = 
           "<b>Results: </b>"+yahooObj.ResultSet.totalResultsAvailable;

    }

  </script>


</head>
<body>

  <h1>JavaScript Object Notation</h1>

  <div id="resultDiv"></div>

  <script type="text/javascript" src="https://api.search.yahoo.com/ImageSearchService/V1/imageSearch?appid=YahooDemo&query=Madonna&results=2&output=json&callback=showNumberOfResults" ></script>

</body>
</html>

Again starting at the bottom, we are creating a script tag that has as its source the actual YWS request. If you scroll all the way to the right, you'll see that we've added a new parameter:

&callback=showNumberOfResults

What this does is tell Yahoo! To return a call to the showNumberOfResults script, passing the object as a parameter. If you then look at the top of the page, we slightly altered that function to take the object rather than the string. When the page loads, the browser comes to this tag, loads the source, and calls the function. If you load this page, you will see the number of results after a slight delay:

The results

This still leaves us with one problem: how do we create a page that takes data from the user -- such as the keyword they want to search on -- and returns the results to the browser? The answer lies in the script tag. In this case, we hardcoded it onto the page. But what if we created it dynamically, in response to a user action? Try this:

<html>
<head>
  <title>JSON testing</title>

  <script type="text/javascript" src="./json.js" ></script>
 
  <script type="text/javascript">
  
    function showLiveResults(yahooObj){

      document.getElementById("resultDiv").innerHTML = 
               "<b>Results: </b>"+yahooObj.ResultSet.totalResultsAvailable;

    }

    function makeResultScript(){
      var form = document.getElementById("keywordForm");
      var keyword = form.elements["keyword"].value;

      var url = "http://api.search.yahoo.com/ImageSearchService/V1/imageSearch?appid=YahooDemo"+
            "&query="+keyword+
            "&results=2&output=json&callback=showLiveResults";
      
    headElement = document.getElementsByTagName("head").item(0);
    var scriptTag = document.createElement("script");
    scriptTag.setAttribute("type", "text/javascript");
    scriptTag.setAttribute("src", url);

    headElement.appendChild(scriptTag);

    }


  </script>


</head>
<body>

   <h1>JavaScript Object Notation</h1>

   <div id="resultDiv"></div>

   <form id="keywordForm">
      Keyword: <input type="text" name="keyword" />
      <input type="button" onclick="makeResultScript()" value="Get Number Of Results" />
    </form>

</body>

</html>

Starting once again at the bottom, we've created a form. That form takes a keyword, and when you click the button, it calls the makeResultScript() function. That function gets a reference to the form, from which it extracts the keyword the user has chosen. It then uses that keyword to create the appropriate URL for the request. (Notice that we've changed the name of the function to showLiveResults.) It then gets a reference to the head element, and then creates a new script element with the appropriate type and src element.

The moment it adds that element to the head element, the call goes out to the Web service, and when it comes back, the object is fed to the showLiveResults function. That function simply extracts the data and adds it to the page, just as we've been doing all along. The result is a page in which you can at a keyword and get the number of results almost instantly:

The results

There's just one more thing left to do: remove the clutter. We've now added a new tag to the DOM, so let's go about removing it in case we want to make another request:

<html>

<head>
  <title>JSON testing</title>

  <script type="text/javascript" src="./json.js" ></script>
 
  <script type="text/javascript">
  

    var scriptCounter = 1;

    function showLiveResults(yahooObj){

      document.getElementById("resultDiv").innerHTML = "<b>Results: </b>"+yahooObj.ResultSet.totalResultsAvailable;

      var scriptTag = document.getElementById(scriptCounter);
      headElement = document.getElementsByTagName("head").item(0);
      headElement.removeChild(scriptTag);
      scriptCounter++;

    }

    function makeResultScript(){
      var form = document.getElementById("keywordForm");
      var keyword = form.elements["keyword"].value;

      var url = "http://api.search.yahoo.com/ImageSearchService/V1/imageSearch?appid=YahooDemo"+
            "&query="+keyword+
            "&results=2&output=json&callback=showLiveResults";
      
    headElement = document.getElementsByTagName("head").item(0);
    var scriptTag = document.createElement("script");

    scriptTag.setAttribute("id", scriptCounter);
    scriptTag.setAttribute("type", "text/javascript");
    scriptTag.setAttribute("src", url);

    headElement.appendChild(scriptTag);

    }

  </script>
...

When we create the script tag, we give it a unique id attribute, based on the scriptCounter variable. When the script has been executed, we then use standard DOM techniques to remove it from the document, incrementing the counter as we go. You shouldn't see any change in the document, but you should be able to make multiple requests without reloading the page.

And there you have it. Accessing a Web service from the browser without reloading and without having to use a proxy to avoid cross domain problems. And it's fast, too. Yahoo! Is not the only web service that provides JSON data, either. You can also use this technique with del.icio.us, and I wouldn't be surprised if other services jumped on the bandwagon in the near future.

InformIT Promotional Mailings & Special Offers

I would like to receive exclusive offers and hear about products from InformIT and its family of brands. I can unsubscribe at any time.