[mary-dev] Phonetic transcriptions

[Fabio Tesser]

Hello,

Using the DatabaseSelector (point 6 of 
http://mary.opendfki.de/wiki/NewLanguageSupport) I was thinking that 
should be useful to have the phonetic transcriptions of the selected 
sentences into a file (i.e. selected_transcriptions.log). This because 
it should be useful to check the transcription of the selected 
sentences. This should be also a like to have feature in the 
SynthesisScriptGUI.

I would like to add this feature into the DatabaseSelector procedure in 
the Italian branch.
For the moment I have write some a procedure that get the phonetic 
transcription from maryserver using the MaryClient class.

The essential of this procedure is the following:
------
         ...
         MaryClient mary = MaryClient.getMaryClient(new 
Address(serverHost, serverPort));
         ...
         ByteArrayOutputStream baos = new ByteArrayOutputStream();
         mary.process(text, "TEXT", "ALLOPHONES", locale, null, null, baos);
         ...
         // read into mary data object
         MaryData maryData = new MaryData(MaryDataType.ALLOPHONES, null);
         maryData.readFrom(new ByteArrayInputStream(baos.toByteArray()));
         Document doc = maryData.getDocument();
         ...
         TreeWalker phWalker = MaryDomUtils.createTreeWalker(doc, doc, 
MaryXML.PHONE);
         Element ph;
         String lTranscription = "";
         while ((ph = (Element) phWalker.nextNode()) != null) {
             lTranscription = lTranscription + ph.getAttribute("p") + ' ';
         }
         ...
------

I'm pretty new in OpenMary, and so I would like to know if others 
simpler ways or other classes exist to obtain the phonetic transcription 
of a sentence.

Thanks in advance,
Fabio.