Now comes the most mysterious part of all: XML. XML is a term which many have heard, but which no one seems to understand. XML is basically a method for encapsulating data with descriptive tags. So, unlike HTML which mostly describes how to format the data, XML’s tags describe what the data is (a price, a link, etc). Why use XML in a preferences file? So the data does not have to reside in any particular order; its tags make it clear what type of data it is.
Well-formed, validated XML files are regulated by a descriptor file (DTD). Here’s the DTD file for this project. I won’t try to explain all the syntax here. For a tutorial on XML, a place to start is IBM’s Alphaworks site (http://www-4.ibm.com/software/developer/education/tutorial-prog/writing.html).
| <?xml encoding="US-ASCII"?> <!ELEMENT myFetchAppPrefs ( ProxyInfo*, SiteInfo+ ) > <!ELEMENT ProxyInfo EMPTY > <!ATTLIST ProxyInfo IP CDATA #REQUIRED Port CDATA #REQUIRED username CDATA #IMPLIED pass CDATA #IMPLIED> <!ELEMENT SiteInfo EMPTY > <!ATTLIST SiteInfo url CDATA #REQUIRED username CDATA #IMPLIED password CDATA #IMPLIED> |
Now that we have the descriptor file, it becomes a matter of entering the data into an XML file. I used IBM’s free Xeena to create the XML file, although it requires some work to get installed and running properly. The file is just plain text, so it can also be edited with any text editor, if you elect to go that route.
| <?xml version="1.0" encoding="UTF-8"?> <!DOCTYPE EnergyFetchAppPrefs SYSTEM "myFetchAppPrefs.dtd" > <myFetchAppPrefs> <ProxyInfo Port="8080" IP="192.168.34.44"/> <SiteInfo url="https://store.sun.com"/> <SiteInfo url="http://www.apple.com"/> <SiteInfo url="ftp://nerctagftp.nerc.com/active/CA%20sheet.csv"/> </myFetchAppPrefs> |
Note that the last entry is an FTP site and that the filename we want to retrieve has a space in it. While this is considered poor form, you can get around it by using the %20 symbol instead of the space. If you wanted to include a username and password the entry would look like:
| <SiteInfo url="http://www.somesite.com" username=”username” password=”pass”/> |
Likewise with the proxy info, should you require a username and password there:
| <ProxyInfo Port="8080" IP="192.168.34.44" username=”username” pass=”pass”/> |
To parse the XML file from within our Java code, we open the document and parse it using both W3- and IBM-provided classes:
| DOMParser parser = new DOMParser(); parser.parse(filename); Document document = parser.getDocument(); Element root = (Element)document.getDocumentElement(); |
Then we take the Element root (which is actually a node) and traverse the branch, one node at a time and take the data as it comes. In this sample application, I have chosen to allocate an array of SiteObjects (part three); not the most elegant approach, but hey, this is just a sample app, right? We only need a single ProxyObject since that will not change for a given machine. Note that when you declare (instantiate) an array of a custom object type, you have to then allocate memory (instantiate). I chose to do this in the init() function in the final code.
| NamedNodeMap theMap=null; Node pointer=null; int counter=0; if (node.hasChildNodes()) { NodeList nl = node.getChildNodes(); int size = nl.getLength(); for (int i = 0; i < size; i++) { theMap=nl.item(i).getAttributes(); // does this node have any attributes (data, in our case)? if (theMap!=null) { // Is there any information to get from this node? if (nl.item(i).getNodeName().equalsIgnoreCase("proxyinfo")) { // we've got proxy info! for (int x=0; x<theMap.getLength(); x++) { // get the proxy values and stuff them into the object if (theMap.item(x).getNodeName().equalsIgnoreCase("port")) proxyInfo.ProxyPort=theMap.item(x).getNodeValue(); if (theMap.item(x).getNodeName().equalsIgnoreCase("ip")) proxyInfo.ProxyServer=theMap.item(x).getNodeValue(); if (theMap.item(x).getNodeName().equalsIgnoreCase("username")) proxyInfo.username=theMap.item(x).getNodeValue(); if (theMap.item(x).getNodeName().equalsIgnoreCase("pass")) proxyInfo.password=theMap.item(x).getNodeValue(); } // for } // if=proxyinfo else if (nl.item(i).getNodeName().equalsIgnoreCase("siteinfo")) { // Site Information. Note that we ignore any other Node Names (types) for (int x=0; x<theMap.getLength(); x++) { // get the site values and stuff them into the object if (theMap.item(x).getNodeName().equalsIgnoreCase("url")) siteInfo[counter].URL=theMap.item(x).getNodeValue(); if (theMap.item(x).getNodeName().equalsIgnoreCase("username")) siteInfo[counter].username=theMap.item(x).getNodeValue(); if (theMap.item(x).getNodeName().equalsIgnoreCase("password")) siteInfo[counter].password=theMap.item(x).getNodeValue(); } // for counter++; // keep track! if (counter>gArrayLength) { System.out.println("Error: More sites than space available.”+ “ Please increase the size of the gArrayLength variable."); return; } } // else if=siteinfo } // if theMap!=null } // for i<size } // if has child nodes } |
Although this might seem like a bit of work, the one real advantage is that we can take an XML file generated by anyone, using any XML software, as long as they used our DTD file, and the information can be entered in any order.
