The various user interface elements, as shown so far, all rely on a single Search class for the actual web service data retrieval; the JSP pages don't contain any code that actually communicates with web services. Instead, a single Search object relies on supporting classes, as shown in Figure 4-7, to retrieve the data.
The four static methods, addSearch( ), getAllSearches( ), getSearch( ), and removeSearch( ), are the main points of interest to the JSP pages. All data about the web services results is contained in a set of simple name/value pairs, contained in the java.util.Hashtable variable attributes. When a search is run, the Search object is passed to the three supporting web service connection classes, and the appropriate data is set in attributes.
The code shown in Example 4-6 shows the code for the Search class. Pay particular interest to the update( ) method.
package com.cascadetg.ch04; import java.util.Hashtable; public class Search { // Static search Hashtable used for persistence. // This means that the data stored by this application is lost // every time the server is restarted. Adding support for // persistence is left as an exercise for the reader - the most // likely, of course, being saving the data to a database in some // fashion. // If you are interested in adding persistence, you'll want to // intercept the addSearch( ), removeSearch( ), and update( ) methods // in this class. private static Hashtable searches = new Hashtable( ); public static boolean addSearch( String productID, String productTitle) { if (productID == null) return false; if (productID.length( ) < 1) return false; if (productTitle == null) return false; if (productID.length( ) < 1) return false; if (searches.get(productID) != null) searches.remove(productID); Search mySearch = new Search( ); mySearch.setProductID(productID); mySearch.setProductTitle(productTitle); searches.put(productID, mySearch); mySearch.update( ); return true; } public static boolean removeSearch(String productID) { if (searches == null) { return true; } if (searches.get(productID) == null) return false; searches.remove(productID); return true; } public static Hashtable getAllSearches( ) { return searches; } public static Search getSearch(String inProductID) { return (Search)searches.get(inProductID); } // Simple object data. Note the use of a Hashtable to store // a variable set of attributes private String productID; private String productTitle; private Hashtable attributes = new Hashtable( ); public void setAttribute(String name, String value) { attributes.put(name, value); } public String getAttribute(String name) { return (String)attributes.get(name); } public java.util.Map getAttributes( ) { java.util.TreeMap map = new java.util.TreeMap(java.text.Collator.getInstance( )); map.putAll(attributes); return map; } /** Returns the product ID (specifically, a book ISBN). */ public String getProductID( ) { return productID; } /** Set the product ID (specificially, a book ISBN). */ public void setProductID(String productID) { this.productID = productID; } public String getProductTitle( ) { return productTitle; } public void setProductTitle(String productTitle) { this.productTitle = productTitle; } // Updating data logic // Note that this updating logic isn't as useful as it should be, // as the update( ) ought to kick off a new thread. private boolean is_updating = false; public boolean isUpdating( ) { return is_updating; } public void update( ) { // Don't allow the user to set up dozens of hits on the site at // once. Ideally, you should be more conservative about kicking // off requests, perhaps using an aggregation approach (as // described in Chapter 8, News Aggregator if (is_updating) return; is_updating = true; // Create a new Search and set the values to the existing // search Search mySearch = new Search( ); mySearch.setProductID(this.productID); mySearch.setProductTitle(this.productTitle); // Kick off the requests to the various web services new AmazonConnection( ).addAttributes(mySearch); new EbayConnection( ).addAttributes(mySearch); new GoogleConnection( ).addAttributes(mySearch); // Replace the old attributes with the newly discovered values this.attributes = mySearch.attributes; is_updating = false; } }
It's easy to imagine enhancements: the underlying system might implement a Connection interface, with the potential to add new web services dynamically.
One other thing you may have noticed: the getAttributes() method returns a java.util.Map, instead of the underlying Hashtable. As shown in Example 4-7, the code creates a TreeMap that keeps the attributes sorted by the system's default locale interpretation of an "alphabetical" sorting routine.
java.util.TreeMap map = new java.util.TreeMap(java.text.Collator.getInstance( )); map.putAll(attributes);
Now, on to the real meat of the application?the web services connectivity.
The Amazon web services package is implemented according to the commonly regarded interpretation of web services?a series of SOAP methods, complete with a WSDL file describing their offerings. We will use Axis, as described in Chapter 3, to work with Amazon's web services.
Before working with Amazon's web services, you must register with Amazon's developer program at http://www.amazon.com/webservices/, as shown in Chapter 4.
The instructions are straightforward: you must download the developer kit (a single file, kit.zip) and get a web services token. This token, a simple string, is used by Amazon to uniquely identify you when you are accessing web services. You should treat this token as you would any password.
As shown in Figure 4-9, you'll find a number of examples for a variety of programming languages as well as the documentation.
Open the file kit/API Guide/index.html in your web browser to see the documentation, as shown in Figure 4-10.
In our case, we'll use Axis to generate access code from the Amazon-supplied WSDL file. The Amazon WSDL file, as described in the READMEFIRST.txt file, can be found at http://soap.amazon.com/schemas3/AmazonWebServices.wsdl. This WSDL file is used by the Axis WSDL2Java tool to generate a set of Java bindings. The easiest way to do this is to open a command prompt and navigate to the Axis library installation directory. Then, execute the following command, all as one line, as shown in Example 4-8. Note that you need an Internet connection for this to work.
C:\devenv\axis-1_1\lib>java -classpath commons-logging.jar; log4j-1.2.8.jar;wsdl4j.jar;axis.jar;commons-discovery.jar; jaxrpc.jar;saaj.jar org.apache.axis.wsdl.WSDL2Java http://soap.amazon.com/schemas3/AmazonWebServices.wsdl
If successful, there will be no visible output on the command line, but you will now have a set of Java classes corresponding to the Amazon SDK in the Axis directory. As shown in Figure 4-11, these classes are neatly sorted into a com.amazon.soap package. Although you are given the source to these classes by this tool, you won't want to edit them manually. You will likely want to regenerate these files any time Amazon adds new functionality.
Copy these files out of this directory and into your Java source tree.
The bulk of the generated classes essentially serve as data holders for Amazon web service operations. To access the web services, use the AmazonSearchServiceLocator class to retrieve a service and an AmazonSearchPort to handle web service requests. The AmazonSearchServiceLocator class hierarchy is shown in Figure 4-12.
After retrieving a service, you use a search port object to perform requests. The generated class hierarchy is shown in Figure 4-13. Note that the methods of the AmazonSearchPort class take instances of other classes generated by the Axis toolkit as parameters.
The code shown in Example 4-9 shows how to actually make a connection to the Amazon web services. In this case, an instance of the AsinRequest class is used. The call to myAmazonSearchPort.asinSearchRequest(myAsinRequest) actually performs the request. It transforms the underlying data to SOAP, sends the data over the network, retrieves the resulting data, and transforms that data back into Java object(s), in this case a ProductInfo object.
package com.cascadetg.ch04; import java.util.Date; import com.amazon.soap.*; public class AmazonConnection { public void addAttributes(Search inSearch) { inSearch.setAttribute( "Amazon Last Check", new Date( ).toLocaleString( )); // Mac OS X for Java Geeks ISBN //String isbn = "0596004001"; try { AmazonSearchService myAmazonSearchService = new AmazonSearchServiceLocator( ); AmazonSearchPort myAmazonSearchPort = myAmazonSearchService.getAmazonSearchPort( ); AsinRequest myAsinRequest = new AsinRequest( ); // Use this to set your Amazon Associates ID // For more info on Amazon Associates, see... // http://www.amazon.com/associates myAsinRequest.setTag(DeveloperTokens.amazon_associates); myAsinRequest.setDevtag(DeveloperTokens.amazon_token); myAsinRequest.setAsin(inSearch.getProductID( )); myAsinRequest.setLocale("us"); myAsinRequest.setType("heavy"); ProductInfo myProductInfo = myAmazonSearchPort.asinSearchRequest(myAsinRequest); Details[] myDetailsArray = myProductInfo.getDetails( ); Details myDetail = null; if (myDetailsArray != null) { myDetail = myDetailsArray[0]; inSearch.setAttribute( "Amazon Product Name", myDetail.getProductName( )); inSearch.setAttribute( "Amazon Release Date", myDetail.getReleaseDate( )); inSearch.setAttribute( "Amazon Actual Prize", myDetail.getOurPrice( )); inSearch.setAttribute( "Amazon List Price", myDetail.getListPrice( )); inSearch.setAttribute( "Amazon Used Price", myDetail.getUsedPrice( )); inSearch.setAttribute( "Amazon Sales Rank", myDetail.getSalesRank( )); inSearch.setAttribute( "Amazon Availability", myDetail.getAvailability( )); //myDetail.getImageUrlSmall( ); //myDetail.getImageUrlMedium( ); //myDetail.getImageUrlLarge( ); } } catch (Exception e) { e.printStackTrace( ); } } }
The code shown in Example 4-9 is as close to a standard RPC system as a web service is likely to be. None of the actual connectivity details are in the code: all of the URLs, ports, and other connectivity information are in the WSDL and then embedded in the code generated by the WSDL2Java tool.
eBay's web service implementation provides access to their systems via XML over HTTPS. A simple HTTPS connection provides in-transit security, and XML data is sent and received to perform operations. While establishing an HTTPS connection to the eBay servers is a relatively trivial matter, working with the XML documents is more time-consuming.
eBay in TransitionIn mid-2004, eBay announced support for SOAP, initially limited to few core APIs. Depending on your needs, you may be able to use the SOAP interfaces provided by eBay in lieu of the XML-via-HTTPS mechanism described here. As of this writing, both the eBay and PayPal platforms are undergoing significant changes, including the transition to SOAP, more complex user authorization systems, and integration of XML descriptions of complex user interface elements for working with more robust item metadata. Visit http://developer.ebay.com/ frequently for the latest information on the eBay platform. |
As of mid-2004, there are several tiers of access offered by eBay, based more or less on the quantity of web service requests you require. The free, individual developer version allows for only 50 calls a day?not very many, but likely enough to build and do some basic testing of your application. The next step up, called Basic, offers 30,000 calls per month for $500 per year.
By default, you have access to the eBay sandbox server, a test environment set up for developers to test their application without incurring service fees. To access the production server, you must be certified by eBay (as of this writing, eBay charges $100 for certification). If you're looking at building a production application that accesses eBay services, you need to include these access and certification fees in your budget. Don't forget that it can take several days to get your application certified.
The eBay sandbox, http://sandbox.ebay.com/, shown in Figure 4-14, is a good development resource. It's a recreation of the eBay web site (http://www.ebay.com/) and is intended for developers to test their application. There are several portions of the web site that aren't implemented or work slightly differently; for example, you can't actually enter a credit card and authenticate a user to sell items on the test server. Instead, you use eBay web services to do this programmatically using the web services API.
The eBay developer registration system is a bit more complex than other offerings in order to support more complex deployment and usage models. First, you need to register for eBay itself and then go through the six-step registration process (http://developer.ebay.com/ Membership Join). When you're done registering, you can generate a set of three security tokens, a Developer ID (devID), Application ID (appID), and Certification ID (certID). Make sure that your pop-up blocking software is disabled when you generate your security tokens.
|
In order to make it easier to connect to eBay in this chapter (and others), the sample applications use a standard eBay connectivity class, as shown in Figure 4-15, as a reusable component for connecting to eBay's XML-via-HTTPS web services.
The code for the eBay class is shown in Example 4-10. All the details of opening a connection to eBay, sending the XML, and then retrieving the results are wrapped by this class. The code is designed to make a call as simple as possible: create a new EbayAPISimpleCall object, set the eBay "verb" you wish to call, and add the arguments you need with EbayAPISimpleCall.setArgument().
package com.cascadetg.ch04; import java.net.*; import java.io.*; import java.util.Hashtable; /** * Based on the eBay sample code, APISimpleCallJava.java */ public class EbayAPISimpleCall { // The eBay API function private String apiVerb = "GeteBayOfficialTime"; /** API Verb used to get the official eBay time */ public static final String GetOfficialTime = "GeteBayOfficialTime"; /** API Verb used to search on eBay */ public static final String GetSearchResults = "GetSearchResults"; // The version threshold of compatibility for the eBay API call private String compatibilityLevel = "331"; // The error level for the API call private String errorLevel = "1"; // Set siteId to correspond to the US Site private String siteId = "0"; // "0" is the only supported detail level for GetebayOfficialTime private String detailLevel = "0"; public EbayAPISimpleCall( ) { } boolean isProduction = false; public org.jdom.Document executeCall( ) { // Supply a URL to the desired API server, for example: // Sandbox: https://api.sandbox.ebay.com/ws/api.dll // Production: https://api.ebay.com/ws/api.dll String targetURL; if (!isProduction) { targetURL = "https://api.sandbox.ebay.com/ws/api.dll"; } else { targetURL = "https://api.ebay.com/ws/api.dll"; } // This string will contain the assembled XML for the API call String eBayXML; try { // Create a new URL object using the specified targetURL URL website = new URL(targetURL); // Create a connection to the URL Object foo = website.openConnection( ); // Note that this is actually an SSL connection, even // though cast to an HttpURLConnection. Different JDK // versions use a different SSL connection class, but they // all use HttpURLConnection as the base class (which has // the methods we need). HttpURLConnection connection = (HttpURLConnection)website.openConnection( ); // Specify that the connection will be used for input and // output connection.setDoInput(true); connection.setDoOutput(true); // Specify the method for the URL request connection.setRequestMethod("POST"); // Add the eBay specific headers needed for an eBay API // call // see: "Making the HTTP request" in the eBay API // documentation addeBayAPIHeaders(connection); // Build the XML call String for eBayGetOfficialTime eBayXML = buildBaseXML( DeveloperTokens.eBay_userid, DeveloperTokens.ebay_userPassword) + buildVerbCall( ); // Create the Output and Print Streams OutputStream output = connection.getOutputStream( ); PrintStream pout = new PrintStream(output); // 'Upload' the eBayXML String pout.print(eBayXML); pout.close( ); // Create the Input Streams InputStream input = connection.getInputStream( ); BufferedInputStream bufIn = new BufferedInputStream(input); org.jdom.Document result = new org.jdom.input.SAXBuilder( ).build(bufIn); connection.disconnect( ); return result; } catch (MalformedURLException ex) { ex.printStackTrace( ); } catch (IOException ioException) { ioException.printStackTrace( ); } catch (org.jdom.JDOMException JDOMprob) { JDOMprob.printStackTrace( ); } return null; } // A test routine to verify that everything is working. public static void main(String[] args) throws IOException { EbayAPISimpleCall myCall = new EbayAPISimpleCall( ); myCall.setApiVerb(GetOfficialTime); org.jdom.Document myDocument = myCall.executeCall( ); new org.jdom.output.XMLOutputter( ).output( myDocument, System.out); myCall.setApiVerb(GetSearchResults); myCall.setArgument("Query", "test"); myDocument = myCall.executeCall( ); new org.jdom.output.XMLOutputter( ).output( myDocument, System.out); String result = myDocument.getRootElement( ).getChild("eBayTime").getText( ); System.out.println(result); result = myDocument .getRootElement( ) .getChild("Search") .getChild("GrandTotal") .getText( ); System.out.println(result); } private void addeBayAPIHeaders(HttpURLConnection connection) { // Generate and add the Session Certificate Header connection.addRequestProperty( "X-EBAY-API-SESSION-CERTIFICATE", DeveloperTokens.ebay_devId + ";" + DeveloperTokens.appId + ";" + DeveloperTokens.certId); // Add the Compatibility Level Header connection.addRequestProperty( "X-EBAY-API-COMPATIBILITY-LEVEL", compatibilityLevel); // Add the Developer Name, Application Name, and Certification // Name headers connection.addRequestProperty( "X-EBAY-API-DEV-NAME", DeveloperTokens.ebay_devId); connection.addRequestProperty( "X-EBAY-API-APP-NAME", DeveloperTokens.appId); connection.addRequestProperty( "X-EBAY-API-CERT-NAME", DeveloperTokens.certId); // Add the API verb Header connection.addRequestProperty("X-EBAY-API-CALL-NAME", apiVerb); // Add the Site Id Header connection.addRequestProperty("X-EBAY-API-SITEID", siteId); // Add the Detail Level Header connection.addRequestProperty( "X-EBAY-API-DETAIL-LEVEL", detailLevel); // Add the Content-Type Header connection.addRequestProperty("Content-Type", "text/xml"); // NOTE: eBay recommends setting the Content-Length header // see: "Making the HTTP request" in the eBay API documentation } private String buildBaseXML(String userId, String userPassword) { return "<?xml version='1.0' encoding='utf-8'?>\r\n" + "<request>\r\n" + "<RequestUserId>" + userId + "</RequestUserId>\r\n" + "<RequestPassword>" + userPassword + "</RequestPassword>\r\n"; } Hashtable arguments = new Hashtable( ); public void clearArguments( ) { arguments = new Hashtable( ); } public Hashtable getArguments( ) { return arguments; } public void setArgument(String argument, String value) { arguments.put(argument, value); } public String getArgument(String argument) { return (String)arguments.get(argument); } public void setArguments(Hashtable arguments) { this.arguments = arguments; } private String buildVerbCall( ) { StringBuffer verbCall = new StringBuffer( ); verbCall.append("<ErrorLevel>"); verbCall.append(errorLevel); verbCall.append("</ErrorLevel>\r\n"); verbCall.append("<Verb>"); verbCall.append(apiVerb); verbCall.append("</Verb>\r\n"); verbCall.append("<DetailLevel>"); verbCall.append(detailLevel); verbCall.append("</DetailLevel>\r\n"); verbCall.append("<SiteId>"); verbCall.append(siteId); verbCall.append("</SiteId>\r\n"); if (arguments.size( ) > 0) { java.util.Enumeration keys = arguments.keys( ); while (keys.hasMoreElements( )) { String current = (String)keys.nextElement( ); verbCall.append("<"); verbCall.append(current); verbCall.append(">"); verbCall.append(arguments.get(current)); verbCall.append("</"); verbCall.append(current); verbCall.append(">"); } } verbCall.append("</request>"); return verbCall.toString( ); } /** * @return Returns the detailLevel. */ public String getDetailLevel( ) { return detailLevel; } /** * @param detailLevel * The detailLevel to set. */ public void setDetailLevel(String detailLevel) { this.detailLevel = detailLevel; } /** * @return Returns the errorLevel. */ public String getErrorLevel( ) { return errorLevel; } /** * @param errorLevel * The errorLevel to set. */ public void setErrorLevel(String errorLevel) { this.errorLevel = errorLevel; } /** * @return Returns the siteId. */ public String getSiteId( ) { return siteId; } /** * @param siteId * The siteId to set. */ public void setSiteId(String siteId) { this.siteId = siteId; } /** * @return Returns the compatibilityLevel. */ public String getCompatibilityLevel( ) { return compatibilityLevel; } /** * @param compatibilityLevel * The compatibilityLevel to set. This controls the * versioning of the XML interface for the API. This is * how eBay allows multiple versions of their API to be * supported simultaneously. */ public void setCompatibilityLevel(String compatibilityLevel) { this.compatibilityLevel = compatibilityLevel; } /** * @return Returns the isProduction. */ public boolean isProduction( ) { return isProduction; } /** * @param isProduction * By default (isProduction = false), the server will * contact the eBay test environment, called the * "sandbox." If you wish to use the production * environment, you'll need to set this to true. */ public void setProduction(boolean isProduction) { this.isProduction = isProduction; } public String getApiVerb( ) { return apiVerb; } /** * Two standard verbs are provided with this class, * GeteBayOfficialTime and GetSearchResults. Others can be added * from the eBay documentation. */ public void setApiVerb(String apiVerb) { this.apiVerb = apiVerb; } }
For eBay, a verb is the term that describes the name of a web service method. Each verb describes a particular action, with a set of one or more possible parameters. These verbs generally have self-describing names, such as GeteBayOfficialTime or GetSearchResults. eBay's documentation lists many verbs; using our EbayAPISimpleCall. setApiVerb() method, it's easy to call any of them. Two static String objects are defined in this class; they correspond to the official eBay verbs. An enterprising developer might imagine developing a custom class hierarchy corresponding to the eBay framework. However, keep in mind that this class hierarchy quickly looks like the bindings generated by a SOAP framework!
When you're ready, call EbayAPISimpleCall.executeCall( ), and you'll receive a result via a JDOM object (which holds the XML returned by eBay). If the result is an eBay error message, that is the contents of the XML; otherwise, it should be the data you need.
The code to make the call into eBay is only support infrastructure; it doesn't actually help you write your application. It's analogous to the code generated by the WSDL2Java code in the previous example (talking to Amazon), but in this case, you have to maintain it yourself.
What Is JDOM?JDOM is a library designed to make working with XML feel more "natural" to a Java developer. It can be downloaded from http://www.jdom.org/. There are a few JAR files included with the distribution; you need these to build this project. As of this writing, I'm using JDOM Beta 9. Downloading the ZIP binary archive, find several JAR files in the lib directory, including ant.jar, jaxen-core.jar, jaxen-jdom.jar, saxpath.jar, xalan.jar, xerces.jar, and xml-apis.jar. Add these to your class path and you're ready to start working with JDOM. |
Now that you have a reusable wrapper for accessing eBay web services, it's time to look at the code to actually get useful data back from the eBay service.
As mentioned earlier, you must create and validate a user on the sandbox in order to perform certain activities, such as searches. The main( ) method of the code shown in Example 4-11 can be used to authenticate a user as a command-line utility if needed. The addAttributes() method is called by the Search object to retrieve data from eBay?a straightforward operation, given the wrapper shown in Example 4-10.
package com.cascadetg.ch04; import java.util.Date; import org.jdom.output.XMLOutputter; public class EbayConnection { public org.jdom.Document validateSandboxUser( String username, String password) { EbayAPISimpleCall myCall = new EbayAPISimpleCall( ); // Validate this user on the sandbox server only myCall.setProduction(false); myCall.setApiVerb("ValidateTestUserRegistration"); myCall.setArgument("RequestUserId", username); myCall.setArgument("RequestPassword", password); return myCall.executeCall( ); } /** * Call this class, passing in the username and password you would * like to register on the eBay test server. * * @param args */ static public void main(String[] args) { if (args == null) { System.out.println("Need two arguments (name & password)."); return; } if (args.length < 2) { System.out.println("Need two arguments (name & password)."); return; } try { new XMLOutputter( ).output( new EbayConnection( ).validateSandboxUser( args[0], args[1]), System.out); } catch (Exception e) { System.out.println("Unable to register user."); e.printStackTrace( ); } } public void addAttributes(Search inSearch) { inSearch.setAttribute( "eBay Last Check", new Date( ).toLocaleString( )); EbayAPISimpleCall myCall = new EbayAPISimpleCall( ); myCall.setApiVerb(EbayAPISimpleCall.GetSearchResults); myCall.setArgument("Query", inSearch.getProductTitle( )); myCall.setArgument("Order", "MetaHighestPriceSort"); org.jdom.Document myResults = myCall.executeCall( ); org.jdom.Element root = myResults.getRootElement( ); long count = Long.parseLong( root .getChild("Search") .getChild("GrandTotal") .getText( )); inSearch.setAttribute( "eBay Total Matching Listings", Long.toString(count)); if (count > 0) { org.jdom.Element item = root.getChild("Search").getChild("Items").getChild( "Item"); inSearch.setAttribute( "eBay Highest Bid Item Price", item.getChildText("LocalizedCurrentPrice")); inSearch.setAttribute( "eBay Highest Bid Item Bids", item.getChildText("BidCount")); inSearch.setAttribute( "eBay Highest Bid Item Link", item.getChildText("Link")); } } }
Authorizing a Sandbox User AccountLet's say that you want to test the system's search capability by adding a new item (for example, you'd like to add my book, Mac OS X for Java Geeks, so it will show up in a search). Or, you'd like to create a seller and a couple of buyers on the sandbox to test the full flow of an auction. To do either activity, you must mark a user account on the sandbox server as an authorized seller. Start by using the web interface on http://sandbox.ebay.com/ to create a user. Note that you can't go through the seller authentication mechanism via your web browser: it will fail at the end with a warning. Instead, after you've created a user account, use the code in EbayConnection.java to authorize the account as a seller. Assuming everything is compiled and working properly, just execute this command: java -classpath .;..\lib\jdom.jar com.cascadetg.ch04 .EbayConnection username password This registers the user as having been validated on the eBay server. |
Lets take a closer look at the connectivity code, as shown in Example 4-12. It starts by creating a object to represent the call to eBay, EbayAPISimpleCall. It then sets the eBay verb to use. In the sample, a search is being performed, but you can find documentation on other supported verbs at http://developer.ebay.com/DevZone/docs/API_Doc/index.asp.
Next, the arguments are set. For a search query, the search terms must be specified?in this case, the title of the book. The application also needs the highest value auction result, so a second argument, Order, is added with the string MetaHighestPriceSort (as specified by the eBay documentation for the search verb). The call is then issued, and an XML document containing the results of the query is returned. From there, it's just a question of using the JDOM interfaces to walk through the returned XML.
EbayAPISimpleCall myCall = new EbayAPISimpleCall( ); myCall.setApiVerb(EbayAPISimpleCall.GetSearchResults); myCall.setArgument("Query", inSearch.getProductTitle( )); myCall.setArgument("Order", "MetaHighestPriceSort"); org.jdom.Document myResults = myCall.executeCall( );
To more fully support this, you may wish to perform more robust error handling (indeed, you'll be required to do this to pass eBay's certification), but it shows how even with a nonstandard API, a bit of work can decouple the details of establishing the network connection from the rest of your application logic.
After the previous examples, it may surprise you to see how easy it is to connect to Google's web service offerings. Visiting http://www.google.com/apis/, shown in Figure 4-16, it's pretty straightforward to start building applications using the Google web service interface. Make sure you download the developer kit as shown on the front page; you will need the provided libraries.
Google does offer APIs via SOAP and WSDL, but for Java developers, there's a pre-built library that does all the work for you. It's as if they took the time to run Axis WSDL2Java for you, and then just provided you with a library to use (in fact, that's very close to what they actually did). It's entirely possible to use Axis to generate the Java code from their WSDL file (posted at http://api.google.com/GoogleSearch.wsdl) just as you did for Amazon. Instead, you should rely on the Java library they provided, called googleapi.jar, as shown in Figure 4-17. You must add this library to your class path. For more information on the Google API, open the APIs_Reference.html file.
The code as shown in Example 4-13 could hardly be simpler. More lines are spent looping through the results and adding them to the Search object's attributes than it takes to get data back from the Google server itself. The code converts the GoogleSearchResult objects to Strings and add them to the results given, but it's easy to imagine using the various methods on the GoogleSearchResult object to get more detail and provide even better formatting.
package com.cascadetg.ch04; import java.util.Date; import com.google.soap.search.*; public class GoogleConnection { public void addAttributes(Search inSearch) { inSearch.setAttribute( "Google Last Check", new Date( ).toLocaleString( )); GoogleSearch search = new GoogleSearch( ); // Set mandatory attributes search.setKey(DeveloperTokens.googleKey); search.setQueryString(inSearch.getProductTitle( )); // Set optional attributes search.setSafeSearch(true); // Invoke the actual search GoogleSearchResult result = null; try { result = search.doSearch( ); } catch (GoogleSearchFault e) { e.printStackTrace( ); } // process the result if (result != null) { inSearch.setAttribute( "Google Number of Hits", Integer.toString( result.getEstimatedTotalResultsCount( ))); GoogleSearchResultElement[] mySearchElements = result.getResultElements( ); for (int i = 0; i < mySearchElements.length; i++) { inSearch.setAttribute( "Google Result " + i, mySearchElements[i].toString( )); if (i > 4) { i = mySearchElements.length; } } } } }
You may have noticed references to a DeveloperTokens class (e.g. DeveloperTokens.amazon_associates, or DeveloperTokens.amazon_token). This class is a single static class with a set of Strings containing the various developer tokens used throughout this chapter. The class shown in Example 4-14 contains only placeholder strings of the approximate correct length.
package com.cascadetg.ch04; public class DeveloperTokens { // Amazon userID token public static String amazon_token = "12345678901234"; // Amazon assoicates ID public static String amazon_associates = "123456789-20"; // eBay userID (the "visible" user identification, i.e. the buyer & // seller user. public static String eBay_userid = "1234567890"; public static String ebay_userPassword = "1234567890"; // devId, appId, certId as specified by the eBay Developer's // Program public static String ebay_devId = "123456789012345678901234567890"; public static String appId = "123456789012345678901234567890"; public static String certId = "123456789012345678901234567890"; // Google web services ID key public static String googleKey = "12345678901234567890123456789012"; }
As you've seen, actually making the connection to a provider isn't that hard (especially after creating bindings or utility classes to make it easier to access). Different vendors have different registration systems, pricing models, and different expectations in terms of security, but at least all offer free access to their system for developers to start building and testing applications, and all have published documentation and interfaces.