/**
* Extracts all words of tagged data file in a directory.
*
* @param directoryName
* @return a set of words
*/
public static Set<String> extract(String directoryName) {
Set<String> words = new TreeSet<String>();
// create a file filter
TextFileFilter filter = new TextFileFilter(POS_FILE_EXTENSION);
File directory = new File(directoryName);
File[] files = FileIterator.listFiles(directory, filter);
System.err.println("# of files = " + files.length);
for (File file : files) {
// get sentences
String[] sentences = UTF8FileUtility.getLines(file.getAbsolutePath());
for (String s : sentences) {
words.addAll(getWords(s));
}
}
return words;
}
/**
* Tokenizes all files in a directory.
*
* @param inputDir an input dir
* @param outputDir an output dir
*/
public void tokenizeDirectory(String inputDir, String outputDir) {
TextFileFilter fileFilter = new TextFileFilter(TokenizerOptions.TEXT_FILE_EXTENSION);
File inputDirFile = new File(inputDir);
// get the current dir
String currentDir = new File(".").getAbsolutePath();
String inputDirPath = currentDir + File.separator + inputDir;
String outputDirPath = currentDir + File.separator + outputDir;
if (DEBUG) {
System.out.println("currentDir = " + currentDir);
System.out.println("inputDirPath = " + inputDirPath);
System.out.println("outputDirPath = " + outputDirPath);
}
// get all input files
File[] inputFiles = FileIterator.listFiles(inputDirFile, fileFilter);
System.out.println("Tokenizing all files in the directory, please wait...");
long startTime = System.currentTimeMillis();
for (File aFile : inputFiles) {
// get the simple name of the file
String input = aFile.getName();
// the output file have the same name with the automatic file
String output = outputDirPath + File.separator + input;
// tokenize the file
tokenize(aFile.getAbsolutePath(), output);
}
long endTime = System.currentTimeMillis();
float duration = (float) (endTime - startTime) / 1000;
System.out.println(
"Tokenized "
+ nTokens
+ " words of "
+ inputFiles.length
+ " files in "
+ duration
+ " (s).\n");
}
