public void map(LongWritable key, Text value, Context context)
throws IOException, InterruptedException {
String strId = "";
String strBody = "";
// Parse the xml and read data (page id and article body)
// Using XOM library
Builder builder = new Builder();
try {
Document doc = builder.build(value.toString(), null);
Nodes nodeId = doc.query("//eecs485_article_id");
strId = nodeId.get(0).getChild(0).getValue();
Nodes nodeBody = doc.query("//eecs485_article_body");
strBody = nodeBody.get(0).getChild(0).getValue();
} catch (ParsingException ex) {
System.out.println("Not well-formed.");
System.out.println(ex.getMessage());
} catch (IOException ex) {
System.out.println("io exception");
}
// Tokenize document body
Pattern pattern = Pattern.compile("\\w+");
Matcher matcher = pattern.matcher(strBody);
while (matcher.find()) {
// Write the parsed token
// key = term, docid value = 1
context.write(new Text(matcher.group() + "," + strId), one);
}
}
private Collection<ConnectableDefinition> loadConnectableDefs(File connectableDefs) {
Collection<ConnectableDefinition> defs = new HashSet<ConnectableDefinition>();
try {
Builder parser = new Builder();
Document doc = parser.build(connectableDefs);
Element root = doc.getRootElement();
Elements definitions = root.getChildElements();
for (int i = 0; i < definitions.size(); i++) {
Element definition = definitions.get(i);
String id = definition.getAttributeValue("id");
String pinCount = definition.getAttributeValue("pins");
String type = definition.getAttributeValue("type");
ConnectableDefinition d = new ConnectableDefinition(id, type, Integer.parseInt(pinCount));
Elements pins = definition.getChildElements();
for (int j = 0; j < pins.size(); j++) {
Element pinElement = pins.get(j);
String pinNumber = pinElement.getAttributeValue("number");
String pinName = pinElement.getAttributeValue("name");
d.addPin(Integer.parseInt(pinNumber), pinName);
}
defs.add(d);
}
} catch (ParsingException ex) {
System.err.println("malformed XML file : " + ex.getMessage());
} catch (IOException ex) {
System.err.println("io error : " + ex.getMessage());
}
return defs;
}
protected ComponentSet loadSet(File componentsDef, File connectionsDef) {
ComponentSet set = null;
// for collecting all connectables
Map<String, Connectable> connectables = new HashMap<String, Connectable>();
try {
// open the components.xml file
Builder builder = new Builder();
Document doc = builder.build(componentsDef);
Element componentsRoot = doc.getRootElement();
int rows = Integer.parseInt(componentsRoot.getAttributeValue("rows"));
int cols = Integer.parseInt(componentsRoot.getAttributeValue("cols"));
boolean autoLocate = Boolean.parseBoolean(componentsRoot.getAttributeValue("autoLocate"));
// add connectables (components need to be located in the set,
// endpoints are not so they are just added to the connectables map
Elements componentElements = componentsRoot.getChildElements();
if (autoLocate) {
set = addComponentsWithAutoLocate(connectables, rows, cols, componentElements);
} else {
set = addComponents(connectables, rows, cols, componentElements);
}
doc = builder.build(connectionsDef);
Element connectionsRoot = doc.getRootElement();
// add connections
Elements connections = connectionsRoot.getChildElements();
for (int i = 0; i < connections.size(); i++) {
Element c = connections.get(i);
String from = c.getAttributeValue("from");
String to = c.getAttributeValue("to");
String fromPinLabel = c.getAttributeValue("fromPin");
String toPinLabel = c.getAttributeValue("toPin");
Connectable fromComp = connectables.get(from);
Connectable toComp = connectables.get(to);
Collection<Pin> fromPins = fromComp.getPins(fromPinLabel);
Collection<Pin> toPins = toComp.getPins(toPinLabel);
Connection con = new Connection(fromComp, fromPins, toComp, toPins);
set.addConnection(con);
}
} catch (ParsingException ex) {
System.err.println("malformed XML file : " + ex.getMessage());
} catch (IOException ex) {
System.err.println("io error : " + ex.getMessage());
}
return set;
}
public static void main(String[] args) {
if (args.length <= 0) {
System.out.println("Usage: java nu.xom.samples.PureValidator URL");
return;
}
try {
Builder parser = new Builder(true, new MinimalNodeFactory());
parser.build(args[0]);
System.out.println(args[0] + " is valid.");
} catch (ValidityException ex) {
System.out.println(args[0] + " is not valid.");
System.out.println(ex.getMessage());
System.out.println(" at line " + ex.getLineNumber() + ", column " + ex.getColumnNumber());
} catch (ParsingException ex) {
System.out.println(args[0] + " is not well-formed.");
System.out.println(ex.getMessage());
System.out.println(" at line " + ex.getLineNumber() + ", column " + ex.getColumnNumber());
} catch (IOException ex) {
System.out.println("Due to an IOException, the parser could not check " + args[0]);
}
}
