public static void main(String[] args) throws Exception {
Reader trainingFile = null;
// Process arguments
int restArgs = commandOptions.processOptions(args);
// Check arguments
if (restArgs != args.length) {
commandOptions.printUsage(true);
throw new IllegalArgumentException("Unexpected arg " + args[restArgs]);
}
if (trainFileOption.value == null) {
commandOptions.printUsage(true);
throw new IllegalArgumentException("Expected --train-file FILE");
}
if (modelFileOption.value == null) {
commandOptions.printUsage(true);
throw new IllegalArgumentException("Expected --model-file FILE");
}
// Get the CRF structure specification.
ZipFile zipFile = new ZipFile(modelFileOption.value);
ZipEntry zipEntry = zipFile.getEntry("crf-info.xml");
CRFInfo crfInfo = new CRFInfo(zipFile.getInputStream(zipEntry));
StringBuffer crfInfoBuffer = new StringBuffer();
BufferedReader reader =
new BufferedReader(new InputStreamReader(zipFile.getInputStream(zipEntry)));
String line;
while ((line = reader.readLine()) != null) {
crfInfoBuffer.append(line).append('\n');
}
reader.close();
// Create the CRF, and train it.
CRF4 crf = createCRF(trainFileOption.value, crfInfo);
// Create a new zip file for our output. This will overwrite
// the file we used for input.
ZipOutputStream zos = new ZipOutputStream(new FileOutputStream(modelFileOption.value));
// Copy the CRF info xml to the output zip file.
zos.putNextEntry(new ZipEntry("crf-info.xml"));
BufferedWriter writer = new BufferedWriter(new OutputStreamWriter(zos));
writer.write(crfInfoBuffer.toString());
writer.flush();
zos.closeEntry();
// Save the CRF classifier model to the output zip file.
zos.putNextEntry(new ZipEntry("crf-model.ser"));
ObjectOutputStream oos = new ObjectOutputStream(zos);
oos.writeObject(crf);
oos.flush();
zos.closeEntry();
zos.close();
}
private void writeObject(ObjectOutputStream out) throws IOException {
out.writeInt(CURRENT_SERIAL_VERSION);
out.writeInt(string1Length);
out.writeInt(string2Length);
out.writeObject(string1);
out.writeObject(string2);
if (string1Blocks == null) {
out.writeInt(NULL_INTEGER);
} else {
int size = string1Blocks.length;
out.writeInt(size);
for (int i = 0; i < size; i++) {
out.writeObject(string1Blocks[i]);
}
}
if (string2Blocks == null) {
out.writeInt(NULL_INTEGER);
} else {
int size = string2Blocks.length;
out.writeInt(size);
for (int i = 0; i < size; i++) {
out.writeObject(string2Blocks[i]);
}
}
out.writeObject(string1Present);
out.writeObject(string2Present);
out.writeObject(lexicon);
if (block1Indices == null) {
out.writeInt(NULL_INTEGER);
} else {
int size = block1Indices.length;
out.writeInt(size);
for (int i = 0; i < size; i++) {
out.writeInt(block1Indices[i]);
}
}
if (block2Indices == null) {
out.writeInt(NULL_INTEGER);
} else {
int size = block2Indices.length;
out.writeInt(size);
for (int i = 0; i < size; i++) {
out.writeInt(block2Indices[i]);
}
}
out.writeChar(delim);
}