Document parseDocument(String filename) throws Exception {
FileReader reader = new FileReader(filename);
String firstLine = new BufferedReader(reader).readLine();
reader.close();
Document document = null;
if (firstLine.startsWith("<?xml")) {
System.err.println("XML detected; using default XML parser.");
} else {
try {
Class nekoParserClass = Class.forName("org.cyberneko.html.parsers.DOMParser");
Object parser = nekoParserClass.newInstance();
Method parse = nekoParserClass.getMethod("parse", new Class[] {String.class});
Method getDocument = nekoParserClass.getMethod("getDocument", new Class[0]);
parse.invoke(parser, filename);
document = (Document) getDocument.invoke(parser);
} catch (Exception e) {
System.err.println("NekoHTML HTML parser not found; HTML4 support disabled.");
}
}
if (document == null) {
DocumentBuilderFactory factory = DocumentBuilderFactory.newInstance();
try { // http://www.w3.org/blog/systeam/2008/02/08/w3c_s_excessive_dtd_traffic
factory.setFeature("http://apache.org/xml/features/nonvalidating/load-external-dtd", false);
} catch (ParserConfigurationException e) {
System.err.println("Warning: Could not disable external DTD loading");
}
DocumentBuilder builder = factory.newDocumentBuilder();
document = builder.parse(filename);
}
return document;
}
// -----------------------------------------------------
// Description: Determine if a template exists for the
// given file and return true on success.
// -----------------------------------------------------
boolean templateExistFor(String fileName, String destinationDirectory) {
templateIdentified = false;
// from template
int intOriginX = 0;
int intOriginY = 0;
// from template
int intX = 0;
int intY = 0;
int intWidth = 0;
int intHeight = 0;
// calculated
int viaTemplateX = 0;
int viaTemplateY = 0;
int viaTemplatePlusWidth = 0;
int viaTemplatePlusHeight = 0;
String fileName_woext = fileName.substring(0, fileName.lastIndexOf('.'));
String pathPlusFileName_woext =
destinationDirectory + File.separator + fileName_woext + File.separator + fileName_woext;
try {
File folder = new File(templateLocation);
File[] listOfFiles = folder.listFiles();
long startTime = System.currentTimeMillis();
for (File file : listOfFiles) {
if (file.isFile()) {
templateName = "";
DocumentBuilderFactory dbFactory = DocumentBuilderFactory.newInstance();
DocumentBuilder dBuilder = dbFactory.newDocumentBuilder();
Document doc = dBuilder.parse(file);
doc.getDocumentElement().normalize();
System.out.println("Root element :" + doc.getDocumentElement().getNodeName());
NodeList nList = doc.getElementsByTagName("condition");
for (int temp = 0; temp < nList.getLength(); temp++) {
Node nNode = nList.item(temp);
if (nNode.getNodeType() == Node.ELEMENT_NODE) {
Element eElement = (Element) nNode;
String myX = getTagValue("x", eElement);
String myY = getTagValue("y", eElement);
String myWidth = getTagValue("width", eElement);
String myHeight = getTagValue("height", eElement);
String mySuffix = getTagValue("suffix", eElement);
String myRequirements = getTagValue("required", eElement);
intX = Integer.parseInt(myX);
intY = Integer.parseInt(myY);
intWidth = Integer.parseInt(myWidth);
intHeight = Integer.parseInt(myHeight);
viaTemplateX = intX - (intOriginX - pageX);
viaTemplateY = intY - (intOriginY - pageY);
viaTemplatePlusWidth = viaTemplateX + intWidth;
viaTemplatePlusHeight = viaTemplateY + intHeight;
Spawn.execute(
Settings.Programs.CONVERT.path(),
pathPlusFileName_woext + "_trim.png",
"-crop",
String.format("%dx%d+%d+%d", intWidth, intHeight, viaTemplateX, viaTemplateY),
"+repage",
pathPlusFileName_woext + "_" + mySuffix + ".png");
Spawn.execute(
Settings.Programs.CONVERT.path(),
pathPlusFileName_woext + "_trim.png",
"-fill",
"none",
"-stroke",
"red",
"-strokewidth",
"3",
"-draw",
String.format(
"rectangle %d,%d %d,%d",
viaTemplateX, viaTemplateY, viaTemplatePlusWidth, viaTemplatePlusHeight),
"+repage",
pathPlusFileName_woext + "_draw.png");
Spawn.execute(
Settings.Programs.TESSERACT.path(),
pathPlusFileName_woext + "_" + mySuffix + ".png",
pathPlusFileName_woext + "_" + mySuffix);
String line = ""; // String that holds current file line
String accumulate = "";
try {
FileReader input = new FileReader(pathPlusFileName_woext + "_" + mySuffix + ".txt");
BufferedReader bufRead = new BufferedReader(input);
line = bufRead.readLine();
while (line != null) {
accumulate += line;
line = bufRead.readLine();
}
bufRead.close();
input.close();
} catch (IOException e) {
e.printStackTrace();
}
String[] requirements = myRequirements.split("#");
for (String requirement : requirements) {
if (requirement.equals(accumulate.trim())) {
templateName = file.getName();
templateIdentified = true;
return templateIdentified;
}
}
}
}
}
}
// record end time
long endTime = System.currentTimeMillis();
System.out.println(
"Template Search [" + fileName + "] " + (endTime - startTime) / 1000 + " seconds");
} catch (Exception e) {
e.printStackTrace();
}
return templateIdentified;
}
public static void main(String args[]) {
if (args.length < 3) {
System.err.println("Usage: Generate model-file ( -s length | -c count length )");
System.exit(1);
}
try {
SAXParserFactory saxFactory = SAXParserFactory.newInstance();
saxFactory.setNamespaceAware(true);
SAXParser saxParser = saxFactory.newSAXParser();
XMLReader xmlReader = saxParser.getXMLReader();
XMLModelReader modelReader = new XMLModelReader(xmlReader);
FileReader modelInput = new FileReader(args[0]);
InputSource source = new InputSource(modelInput);
Model model = modelReader.load(source);
modelInput.close();
model.check();
if (args[1].equals("-s")) {
int[][] result = model.generateSequence(Integer.parseInt(args[2]));
List lexicon = model.getLexicon();
for (int i = 0; i < result[0].length; i++) {
System.out.print((Character) lexicon.get(result[0][i]));
}
System.out.println();
int numberOfStates = model.getNumberOfStates();
if (numberOfStates > 9) {
for (int i = 0; i < result[1].length; i++) {
System.out.print(result[1][i]);
System.out.print(',');
}
} else {
for (int i = 0; i < result[1].length; i++) {
System.out.print(result[1][i]);
}
}
System.out.println();
for (int i = 1; i < numberOfStates; i++) {
System.out.println(i + " = " + model.getStateName(i));
}
} else if (args[1].equals("-c")) {
List lexicon = model.getLexicon();
int count = Integer.parseInt(args[2]);
int length = Integer.parseInt(args[3]);
ListOfSequences uniqueSeqs = new ListOfSequences();
for (int i = 0; i < count; i++) {
int[][] result = model.generateSequence(length);
uniqueSeqs.addSequence(result[0]);
}
Iterator seqs = uniqueSeqs.iterator();
while (seqs.hasNext()) {
SequenceCount seq = (SequenceCount) seqs.next();
int[] key = seq.getSequence();
for (int i = 0; i < key.length; i++) {
System.out.print((Character) lexicon.get(key[i]));
}
System.out.print(',');
System.out.println(seq.getCount());
}
}
} catch (java.io.IOException ex) {
ex.printStackTrace();
} catch (org.xml.sax.SAXException ex) {
System.err.println(ex.getMessage());
} catch (javax.xml.parsers.ParserConfigurationException ex) {
ex.printStackTrace();
}
}