/**
* Read data from input stream <code>is</code>, in the appropriate way as determined by our <code>
* type</code>.
*
* @param is the InputStream from which to read.
* @param endMarker a string marking end of file. If this is null, read until end-of-file; if it
* is non-null, read up to (and including) the first line containing the end marker string.
* This will be ignored for audio data.
* @throws ParserConfigurationException ParserConfigurationException
* @throws SAXException SAXException
* @throws IOException IOException
* @throws TransformerConfigurationException TransformerConfigurationException
* @throws TransformerException TransformerException
*/
public void readFrom(InputStream is, String endMarker)
throws ParserConfigurationException, SAXException, IOException,
TransformerConfigurationException, TransformerException {
if (type.isXMLType() || type.isTextType())
readFrom(new InputStreamReader(is, "UTF-8"), endMarker);
else { // audio
// ignore endMarker
setAudio((AudioInputStream) is);
}
}
/**
* Read data from reader <code>r</code> in the appropriate way as determined by our <code>type
* </code>. Only XML and Text data can be read from a reader, audio data cannot. "Helpers" needed
* to read the data, such as XML parser objects, are created when they are needed. If doWarnClient
* is set to true, warning and error messages related to XML parsing are logged to the log
* category connected to the client from which this request originated.
*
* @param from the Reader from which to read.
* @param endMarker a string marking end of file. If this is null, read until end-of-file; if it
* is non-null, read up to (and including) the first line containing the end marker string.
* @throws ParserConfigurationException ParserConfigurationException
* @throws SAXException SAXException
* @throws IOException IOException
*/
public void readFrom(Reader from, String endMarker)
throws ParserConfigurationException, SAXException, IOException {
// For the case that the data to be read it is not
// followed by end-of-file, we use a ReaderSplitter which
// provides a reader artificially "inserting" an end-of-file
// after a line containing the pattern given in endMarker.
Reader r = from;
if (endMarker != null) {
ReaderSplitter fromSplitter = new ReaderSplitter(from, endMarker);
r = fromSplitter.nextReader();
}
readFrom(r);
}
/**
* Read data from input stream <code>is</code>, in the appropriate way as determined by our <code>
* type</code>.
*
* @param is is
* @throws ParserConfigurationException ParserConfigurationException
* @throws SAXException SAXException
* @throws IOException IOException
* @throws TransformerConfigurationException TransformerConfigurationException
* @throws TransformerException TransformerException
*/
public void readFrom(InputStream is)
throws ParserConfigurationException, SAXException, IOException,
TransformerConfigurationException, TransformerException {
readFrom(is, null);
}
public void computeFeaturesFor(String basename) throws IOException, Exception {
String text;
Locale localVoice;
localVoice = MaryUtils.string2locale(locale);
// First, test if there is a corresponding .rawmaryxml file in textdir:
File rawmaryxmlFile =
new File(db.getProp(db.MARYXMLDIR) + basename + db.getProp(db.MARYXMLEXT));
if (rawmaryxmlFile.exists()) {
text = FileUtils.getFileAsString(rawmaryxmlFile, "UTF-8");
} else {
text =
getMaryXMLHeaderWithInitialBoundary(locale)
+ FileUtils.getFileAsString(
new File(db.getProp(db.TEXTDIR) + basename + db.getProp(db.TEXTEXT)), "UTF-8")
+ "</maryxml>";
}
File pfeatFile = new File(unitfeatureDir, basename + featsExt);
OutputStream os = new BufferedOutputStream(new FileOutputStream(pfeatFile));
MaryClient maryClient = getMaryClient();
/*Vector voices = maryClient.getVoices(localVoice);
MaryClient.Voice defaultVoice = (MaryClient.Voice) voices.firstElement();
String voiceName = defaultVoice.name();*/
// maryClient.process(text, maryInputType, maryOutputType, null, null, os);
maryClient.process(text, maryInputType, maryOutputType, locale, null, "slt-arctic", os);
// maryClient.process(text, maryInputType, maryOutputType, null, "slt-arctic", os, timeout);
// maryClient.getOutputDataTypes().size()
// MaryData result = new MaryData(os);
os.flush();
os.close();
// System.out.println(" TO STRING: "+new FileReader(pfeatFile).toString());
// BufferedReader bfr = new BufferedReader(new FileReader(pfeatFile));
String line;
MaryData d = new MaryData(MaryDataType.get("PHONEMISED_EN"), Locale.US);
// d.readFrom(new ByteArrayInputStream(os.toByteArray()));
d.readFrom(new FileReader(pfeatFile));
// MaryData d = new MaryData(pfeatFile);
Document doc = d.getDocument();
// Document acoustparams = d.getDocument();
// NodeIterator it = ((DocumentTraversal)acoustparams).createNodeIterator(acoustparams,
// NodeFilter.SHOW_ELEMENT,new NameNodeFilter(new String[]{MaryXML.TOKEN,
// MaryXML.BOUNDARY}),false);
NodeIterator it =
((DocumentTraversal) doc)
.createNodeIterator(
doc, NodeFilter.SHOW_ELEMENT, new NameNodeFilter(MaryXML.TOKEN), false);
Element t = null;
while ((t = (Element) it.nextNode()) != null) {
if (t.hasAttribute("g2p_method")) {
String g2p = t.getAttribute("g2p_method");
String nodeText = t.getTextContent().trim();
if (g2p.equals("rules")) { // && nodeText.equals("!")){
System.out.print(basename + " ----> " + nodeText);
if (bnl.contains(basename)) bnl.remove(basename);
System.out.println(" SO removing basename: " + basename);
}
// System.out.println("G2P:"+t.getAttribute("g2p_method"));
// System.out.println("Text:"+t.getTextContent());
}
}
/*while((line =bfr.readLine()) != null){
//boolean b = m.matches();
if(Pattern.matches("rules", line))
System.out.println(basename + " LINE ---> " + line);
}*/
// System.out.println(" TO STRING: "+line);
}