ziver/hal
1
1
Fork 0
Code Pull requests Releases Activity

Added voice control

Browse source 
Former-commit-id: 6f69079bf44f0d8f9ae40de6b0f1638d103464c2
This commit is contained in:
Ziver Koc 2015-05-13 21:14:10 +00:00
parent 35c92407a3
commit 53da641909
863 changed files with 192681 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

367
external/marytts-5.1.2/doc/examples/client/MaryClient.py vendored Normal file
View file

@ -0,0 +1,367 @@
#!/usr/bin/python
# -*- coding: utf-8 -*-
import socket, sys, types, getopt
languageNames = {'de':'German',
'en':'English',
'en_US':'US English',
'tib':'Tibetan'}
class MaryClient:
specificationVersion = "0.1"
"""Python implementation of a MARY TTS client"""
def __init__( self, host="cling.dfki.uni-sb.de", port=59125, profile=False, quiet=False ):
self.host = host
self.port = port
self.profile = profile
self.quiet = quiet
self.allVoices = None # array of Voice objects
self.voicesByLocaleMap = {} # Map locale strings to arrays of Voice objects
self.allDataTypes = None # array of DataType objects
self.inputDataTypes = None # array of DataType objects
self.outputDataTypes = None # array of DataType objects
self.serverExampleTexts = {}
self.voiceExampleTexts = {}
self.serverVersionInfo = u''
if not self.quiet:
sys.stderr.write( "MARY TTS Python Client %s\n" % ( self.specificationVersion ) )
try:
info = self.getServerVersionInfo()
except:
sys.stderr.write( "Problem connecting to mary server at %s:%i\n" % ( self.host, self.port ) )
raise
sys.stderr.write( "Connected to %s:%i, " % ( self.host, self.port ) )
sys.stderr.write( info )
sys.stderr.write( '\n' )
def __getServerInfo( self, request="", marySocket=None ):
"""Get answer to request from mary server. Returns a list of unicode strings,
each representing a line without the line break.
"""
closeSocket = False
if marySocket is None:
closeSocket = True
marySocket = socket.socket( socket.AF_INET, socket.SOCK_STREAM )
marySocket.connect( ( self.host, self.port ) )
assert isinstance(marySocket, socket.SocketType)
maryFile = marySocket.makefile( 'rwb', 1 ) # read-write, line-buffered
maryFile.write( unicode( request+"\n" ).encode( 'utf-8' ) )
result = []
while True:
got = unicode( maryFile.readline().strip(), 'utf-8' )
# read until end of file or an empty line is read:
if not got: break
result.append(got)
if closeSocket:
marySocket.close()
return result
def getServerVersionInfo( self ):
"Get version info from server. Returns a unicode string"
if self.serverVersionInfo == u'':
# need to get it from server
self.serverVersionInfo = u'\n'.join(self.__getServerInfo("MARY VERSION"))
return self.serverVersionInfo
def getAllDataTypes(self, locale=None):
"""Obtain a list of all data types known to the server. If the information is not
yet available, the server is queried. This is optional information
which is not required for the normal operation of the client, but
may help to avoid incompatibilities.
Returns an array of DataType objects
"""
if self.allDataTypes is None:
self.__fillDataTypes()
assert self.allDataTypes is not None and len( self.allDataTypes ) > 0
if locale is None:
return self.allDataTypes
else:
assert isinstance(locale, types.UnicodeType), "Unexpected type for locale: '%s'" % (type(locale))
return [d for d in self.allDataTypes if d.locale is None or d.locale == locale]
def getInputDataTypes(self,locale=None):
"""Obtain a list of input data types known to the server. If the information is not
yet available, the server is queried. This is optional information
which is not required for the normal operation of the client, but
may help to avoid incompatibilities.
Returns an arry of DataType objects
"""
if self.inputDataTypes is None:
self.__fillDataTypes()
assert self.inputDataTypes is not None and len( self.inputDataTypes ) > 0
if locale is None:
return self.inputDataTypes
else:
assert isinstance(locale, types.UnicodeType), "Unexpected type for locale: '%s'" % (type(locale))
return [d for d in self.inputDataTypes if d.locale is None or d.locale == locale]
def getOutputDataTypes(self, locale=None):
"""Obtain a list of output data types known to the server. If the information is not
yet available, the server is queried. This is optional information
which is not required for the normal operation of the client, but
may help to avoid incompatibilities.
Returns an arry of DataType objects
"""
if self.outputDataTypes is None:
self.__fillDataTypes()
assert self.outputDataTypes is not None and len( self.outputDataTypes ) > 0
if locale is None:
return self.outputDataTypes
else:
assert isinstance(locale, types.UnicodeType), "Unexpected type for locale: '%s'" % (type(locale))
return [d for d in self.outputDataTypes if d.locale is None or d.locale == locale]
def __fillDataTypes( self ):
self.allDataTypes = []
self.inputDataTypes = []
self.outputDataTypes = []
marySocket = socket.socket( socket.AF_INET, socket.SOCK_STREAM )
marySocket.connect( ( self.host, self.port ) )
# Expect a variable number of lines of the kind
# RAWMARYXML INPUT OUTPUT
# TEXT_DE LOCALE=de INPUT
# AUDIO OUTPUT
typeStrings = self.__getServerInfo( "MARY LIST DATATYPES", marySocket )
if not typeStrings or len(typeStrings) == 0:
raise IOError( "Could not get list of data types from Mary server" )
marySocket.close()
for typeString in typeStrings:
parts = typeString.split()
if len( parts ) == 0:
continue
name = parts[0]
isInputType = False
isOutputType = False
locale = None
for part in parts[1:]:
if part[:7] == "LOCALE=":
locale = part[7:]
elif part == "INPUT":
isInputType = True
elif part == "OUTPUT":
isOutputType = True
dt = DataType( name, locale, isInputType, isOutputType )
self.allDataTypes.append( dt )
if dt.isInputType:
self.inputDataTypes.append( dt )
if dt.isOutputType:
self.outputDataTypes.append( dt )
def getVoices( self, locale=None ):
"""Obtain a list of voices known to the server. If the information is not
yet available, the server is queried. This is optional information
which is not required for the normal operation of the client, but
may help to avoid incompatibilities.
Returns an array of Voice objects
"""
if self.allVoices is None:
self.__fillVoices()
assert self.allVoices is not None and len( self.allVoices ) > 0
if locale is None:
return self.allVoices
else:
assert isinstance(locale, types.UnicodeType), "Unexpected type for locale: '%s'" % (type(locale))
if self.voicesByLocaleMap.has_key(locale):
return self.voicesByLocaleMap[locale]
else:
raise Exception("No voices for locale '%s'" % (locale))
def __fillVoices( self ):
self.allVoices = []
self.voicesByLocaleMap = {}
marySocket = socket.socket( socket.AF_INET, socket.SOCK_STREAM )
marySocket.connect( ( self.host, self.port ) )
# Expect a variable number of lines of the kind
# de7 de female
# us2 en male
# dfki-stadium-emo de male limited
voiceStrings = self.__getServerInfo( "MARY LIST VOICES", marySocket )
if not voiceStrings or len(voiceStrings) == 0:
raise IOError( "Could not get list of voices from Mary server" )
marySocket.close()
for voiceString in voiceStrings:
parts = voiceString.split()
if len( parts ) < 3:
continue
name = parts[0]
locale = parts[1]
gender = parts[2]
domain = None
if len( parts ) > 3:
domain = parts[3]
voice = Voice( name, locale, gender, domain )
self.allVoices.append( voice )
localeVoices = None
if self.voicesByLocaleMap.has_key( locale ):
localeVoices = self.voicesByLocaleMap[locale]
else:
localeVoices = []
self.voicesByLocaleMap[locale] = localeVoices
localeVoices.append( voice )
def getGeneralDomainVoices( self, locale=None ):
"""Obtain a list of general domain voices known to the server. If the information is not
yet available, the server is queried. This is optional information
which is not required for the normal operation of the client, but
may help to avoid incompatibilities.
Returns an array of Voice objects
"""
return [v for v in self.getVoices( locale ) if not v.isLimitedDomain]
def getLimitedDomainVoices( self, locale=None ):
"""Obtain a list of limited domain voices known to the server. If the information is not
yet available, the server is queried. This is optional information
which is not required for the normal operation of the client, but
may help to avoid incompatibilities.
Returns an array of Voice objects
"""
return [v for v in self.getVoices( locale ) if v.isLimitedDomain]
def getAvailableLanguages(self):
""" Check available voices and return a list of tuples (abbrev, name)
representing the available languages -- e.g. [('en', 'English'),('de', 'German')].
"""
if self.allVoices is None:
self.__fillVoices()
assert self.allVoices is not None and len( self.allVoices ) > 0
languages = []
for l in self.voicesByLocaleMap.keys():
if languageNames.has_key(l):
languages.append((l,languageNames[l]))
else:
languages.append((l, l))
return languages
def getServerExampleText( self, dataType ):
"""Request an example text for a given data type from the server.
dataType the string representation of the data type,
e.g. "RAWMARYXML". This is optional information
which is not required for the normal operation of the client, but
may help to avoid incompatibilities."""
if not self.serverExampleTexts.has_key( dataType ):
exampleTexts = self.__getServerInfo( "MARY EXAMPLETEXT %s" % ( dataType ) )
if not exampleTexts or len(exampleTexts) == 0:
raise IOError( "Could not get example text for type '%s' from Mary server" % (dataType))
exampleText = u'\n'.join(exampleTexts)
self.serverExampleTexts[dataType] = exampleText
return self.serverExampleTexts[dataType]
def process( self, input, inputType, outputType, audioType=None, defaultVoiceName=None, output=sys.stdout ):
assert type( input ) in types.StringTypes
assert type( inputType ) in types.StringTypes
assert type( outputType ) in types.StringTypes
assert audioType is None or type( audioType ) in types.StringTypes
assert defaultVoiceName is None or type( defaultVoiceName ) in types.StringTypes
assert callable( getattr( output, 'write' ) )
if type( input ) != types.UnicodeType:
input = unicode( input, 'utf-8' )
maryInfoSocket = socket.socket( socket.AF_INET, socket.SOCK_STREAM )
maryInfoSocket.connect( ( self.host, self.port ) )
assert type( maryInfoSocket ) is socket.SocketType
maryInfo = maryInfoSocket.makefile( 'rwb', 1 ) # read-write, line-buffered
maryInfo.write( unicode( "MARY IN=%s OUT=%s" % ( inputType, outputType ), 'utf-8' ) )
if audioType:
maryInfo.write( unicode( " AUDIO=%s" % ( audioType ), 'utf-8' ) )
if defaultVoiceName:
maryInfo.write( unicode( " VOICE=%s" % ( defaultVoiceName ), 'utf-8' ) )
maryInfo.write( "\r\n" )
# Receive a request ID:
id = maryInfo.readline()
maryDataSocket = socket.socket( socket.AF_INET, socket.SOCK_STREAM )
maryDataSocket.connect( ( self.host, self.port ) )
assert type( maryDataSocket ) is socket.SocketType
maryDataSocket.sendall( id ) # includes newline
maryDataSocket.sendall( input.encode( 'utf-8' ) )
maryDataSocket.shutdown( 1 ) # shutdown writing
# Set mary info socket to non-blocking, so we only read somthing
# if there is something to read:
maryInfoSocket.setblocking( 0 )
while True:
try:
err = maryInfoSocket.recv( 8192 )
if err: sys.stderr.write( err )
except:
pass
got = maryDataSocket.recv( 8192 )
if not got: break
output.write( got )
maryInfoSocket.setblocking( 1 )
while True:
err = maryInfoSocket.recv( 8192 )
if not err: break
sys.stderr.write( err )
################ data representation classes ##################
class DataType:
def __init__( self, name, locale=None, isInputType=False, isOutputType=False ):
self.name = name
self.locale = locale
self.isInputType = isInputType
self.isOutputType = isOutputType
def isTextType( self ):
return self.name != "AUDIO"
class Voice:
def __init__( self, name, locale, gender, domain="general" ):
self.name = name
self.locale = locale
self.gender = gender
self.domain = domain
if not domain or domain == "general":
self.isLimitedDomain = False
else:
self.isLimitedDomain = True
def __str__(self):
if languageNames.has_key(self.locale):
langName = languageNames[self.locale]
else:
langName = self.locale
if self.isLimitedDomain:
return "%s (%s, %s %s)" % (self.name, self.domain, langName, self.gender)
else:
return "%s (%s %s)" % (self.name, langName, self.gender)
##################### Main #########################
if __name__ == '__main__':
serverHost = "cling.dfki.uni-sb.de"
serverPort = 59125
inputType = "TEXT"
outputType = "AUDIO"
audioType = "WAVE"
defaultVoice = None
inputEncoding = 'utf-8'
( options, rest ) = getopt.getopt( sys.argv[1:], '', \
['server.host=', 'server.port=', 'input.type=', 'output.type=', \
'audio.type=', 'voice.default=', 'input.encoding='] )
for ( option, value ) in options:
if option == '--server.host': serverHost = value
elif option == '--server.port': serverPort = int( value )
elif option == '--input.type': inputType = value
elif option == '--output.type': outputType = value
elif option == '--audio.type': audioType = value
elif option == '--voice.default': defaultVoice = value
elif option == '--input.encoding': inputEncoding = value
if len( rest )>0: # have input file
inputFile = file( rest[0] )
else:
inputFile = sys.stdin
input = unicode( ''.join( inputFile.readlines() ), inputEncoding )
if len( rest )>1: # also have output file
outputFile = file( rest[1] )
else:
outputFile = sys.stdout
maryClient = MaryClient( serverHost, serverPort )
maryClient.process( input, inputType, outputType, audioType, defaultVoice, outputFile )

102
external/marytts-5.1.2/doc/examples/client/MaryClientUser.java vendored Normal file
View file

@ -0,0 +1,102 @@
/**
* Copyright 2000-2006 DFKI GmbH.
* All Rights Reserved. Use is subject to license terms.
*
* Permission is hereby granted, free of charge, to use and distribute
* this software and its documentation without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of this work, and to
* permit persons to whom this work is furnished to do so, subject to
* the following conditions:
*
* 1. The code must retain the above copyright notice, this list of
* conditions and the following disclaimer.
* 2. Any modifications must be clearly marked as such.
* 3. Original authors' names are not deleted.
* 4. The authors' names are not used to endorse or promote products
* derived from this software without specific prior written
* permission.
*
* DFKI GMBH AND THE CONTRIBUTORS TO THIS WORK DISCLAIM ALL WARRANTIES WITH
* REGARD TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS, IN NO EVENT SHALL DFKI GMBH NOR THE
* CONTRIBUTORS BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL
* DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
* PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
* ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF
* THIS SOFTWARE.
*/
import java.io.ByteArrayInputStream;
import java.io.ByteArrayOutputStream;
import java.io.IOException;
import java.net.UnknownHostException;
import java.util.Locale;
import javax.sound.sampled.AudioInputStream;
import javax.sound.sampled.AudioSystem;
import javax.sound.sampled.LineEvent;
import javax.sound.sampled.LineListener;
import javax.sound.sampled.UnsupportedAudioFileException;
import marytts.util.data.audio.AudioPlayer;
import marytts.client.MaryClient;
import marytts.util.http.Address;
/**
* A demo class illustrating how to use the MaryClient class.
* This will connect to a MARY server, version 4.x.
* It requires maryclient.jar from MARY 4.0.
* This works transparently with MARY servers in both http and socket server mode.
*
* Compile this as follows:
* <code>javac -cp maryclient.jar MaryClientUser.java</code>
*
* And run as:
* <code>java -cp .:maryclient.jar MaryClientUser</code>
*
* @author marc
*
*/
public class MaryClientUser {
public static void main(String[] args)
throws IOException, UnknownHostException, UnsupportedAudioFileException,
InterruptedException
{
String serverHost = System.getProperty("server.host", "cling.dfki.uni-sb.de");
int serverPort = Integer.getInteger("server.port", 59125).intValue();
MaryClient mary = MaryClient.getMaryClient(new Address(serverHost, serverPort));
String text = "Willkommen in der Welt der Sprachsynthese!";
// If the given locale is not supported by the server, it returns
// an ambigous exception: "Problem processing the data."
String locale = "de"; // or US English (en-US), Telugu (te), Turkish (tr), ...
String inputType = "TEXT";
String outputType = "AUDIO";
String audioType = "WAVE";
String defaultVoiceName = null;
ByteArrayOutputStream baos = new ByteArrayOutputStream();
mary.process(text, inputType, outputType, locale, audioType, defaultVoiceName, baos);
// The byte array constitutes a full wave file, including the headers.
// And now, play the audio data:
AudioInputStream ais = AudioSystem.getAudioInputStream(
new ByteArrayInputStream(baos.toByteArray()));
LineListener lineListener = new LineListener() {
public void update(LineEvent event) {
if (event.getType() == LineEvent.Type.START) {
System.err.println("Audio started playing.");
} else if (event.getType() == LineEvent.Type.STOP) {
System.err.println("Audio stopped playing.");
} else if (event.getType() == LineEvent.Type.OPEN) {
System.err.println("Audio line opened.");
} else if (event.getType() == LineEvent.Type.CLOSE) {
System.err.println("Audio line closed.");
}
}
};
AudioPlayer ap = new AudioPlayer(ais, lineListener);
ap.start();
}
}

45
external/marytts-5.1.2/doc/examples/client/c++/Makefile vendored Normal file
View file

@ -0,0 +1,45 @@
##########################################################################
# Copyright (C) 2000-2006 DFKI GmbH.
# All rights reserved. Use is subject to license terms.
#
# Permission is hereby granted, free of charge, to use and distribute
# this software and its documentation without restriction, including
# without limitation the rights to use, copy, modify, merge, publish,
# distribute, sublicense, and/or sell copies of this work, and to
# permit persons to whom this work is furnished to do so, subject to
# the following conditions:
#
# 1. The code must retain the above copyright notice, this list of
# conditions and the following disclaimer.
# 2. Any modifications must be clearly marked as such.
# 3. Original authors' names are not deleted.
# 4. The authors' names are not used to endorse or promote products
# derived from this software without specific prior written
# permission.
#
# DFKI GMBH AND THE CONTRIBUTORS TO THIS WORK DISCLAIM ALL WARRANTIES WITH
# REGARD TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF
# MERCHANTABILITY AND FITNESS, IN NO EVENT SHALL DFKI GMBH NOR THE
# CONTRIBUTORS BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL
# DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
# PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
# ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF
# THIS SOFTWARE.
##########################################################################
CC=g++
CFLAGS=-Wall -w -O3 -g
ICUDIR=/usr/local/icu
ICULIBS=-Wl,-R,$(ICUDIR)/lib -L$(ICUDIR)/lib -licuuc -licui18n -ldl
all: MaryDemo
MaryDemo: MaryClient.o MaryDemo.o
$(CC) $(CFLAGS) *.o -o MaryDemo $(LIBS)
%.o: %.cc
$(CC) $(CFLAGS) $(RFLAGS) -o $@ -c $<
clean:
rm -rf *.o ./MaryDemo

277
external/marytts-5.1.2/doc/examples/client/c++/MaryClient.cc vendored Normal file
View file

@ -0,0 +1,277 @@
/**
* Copyright 2000-2006 DFKI GmbH.
* All Rights Reserved. Use is subject to license terms.
*
* Permission is hereby granted, free of charge, to use and distribute
* this software and its documentation without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of this work, and to
* permit persons to whom this work is furnished to do so, subject to
* the following conditions:
*
* 1. The code must retain the above copyright notice, this list of
* conditions and the following disclaimer.
* 2. Any modifications must be clearly marked as such.
* 3. Original authors' names are not deleted.
* 4. The authors' names are not used to endorse or promote products
* derived from this software without specific prior written
* permission.
*
* DFKI GMBH AND THE CONTRIBUTORS TO THIS WORK DISCLAIM ALL WARRANTIES WITH
* REGARD TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS, IN NO EVENT SHALL DFKI GMBH NOR THE
* CONTRIBUTORS BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL
* DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
* PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
* ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF
* THIS SOFTWARE.
*/
#include <netdb.h>
#include <stdlib.h>
#include <iostream>
#include <string.h>
#include "MaryClient.h"
using namespace std;
/**
* A C++ implementation of a simple client to the MARY TTS system.
* result: an empty string serving as the container for the output.
* It will return text or audio data; text data will be encoded as UTF-8.
* inputText: the UTF-8 encoded text (or XML document) to send as a request
* maryInFormat: the input type of the data in inputText, e.g. TEXT
* maryOutFormat: the output type to produce, e.g. MBROLA, AUDIO
* locale: the language of the input, e.g. EN-US, DE
* audioType: for AUDIO output, the type of audio data to produce,
* e.g. WAVE or MP3.
* voice: the voice to be used, e.g. cmu-slt-hsmm, bits3.
* effects: the list of effects to be generated.
* return value: 0 on success, negative on failure.
*/
int
MaryClient::maryQuery( int server_port,
string server_host,
string& result,
string inputText,
string maryInFormat,
string maryOutFormat,
string locale,
string audioType,
string voice,
string effects ) {
// prepare the request
string query = "MARY";
query += " IN=" + maryInFormat;
query += " OUT=" + maryOutFormat;
query += " LOCALE=" + locale; // remove this line, if using an older version than MARY 4.0
query += " AUDIO=" + audioType;
query += " VOICE=" + voice;
if (effects != "") {
query += " EFFECTS=" + effects;
}
query += "\012\015";
//cout << "Constructed query: " << query << endl;
// declare connection stuff
struct sockaddr_in maryServer;
struct sockaddr_in maryClient;
struct hostent* hostInfo;
// declare variables
int maryInfoSocket;
int maryDataSocket;
// set configuration parameters
// get host information
hostInfo = gethostbyname (server_host.c_str());
if (hostInfo == NULL)
{
return -2;
}
// create a tcp connection to the mary server
maryInfoSocket = socket (AF_INET, SOCK_STREAM, 0);
// verify that the socket could be opened successfully
if (maryInfoSocket == -1)
{
return -2;
}
else
// autoflush stdout, bind and connect
{
maryClient.sin_family = AF_INET;
maryClient.sin_port = htons (0);
maryClient.sin_addr.s_addr = INADDR_ANY;
int status = bind (maryInfoSocket, (struct sockaddr*) &maryClient, sizeof (maryClient));
if (status != 0)
{
return -2;
}
maryServer.sin_family = AF_INET;
maryServer.sin_port = htons (server_port);
memcpy ((char*) &maryServer.sin_addr.s_addr, hostInfo->h_addr_list [0], hostInfo->h_length);
status = connect (maryInfoSocket, (struct sockaddr*) &maryServer, sizeof (maryServer));
if (status != 0)
{
return -2;
}
}
// send request to the Mary server
if (send (maryInfoSocket, query.c_str (), query.size (), 0) == -1)
{
return -2;
}
// receive the request id
char id [32] = "";
if (recv (maryInfoSocket, id, 32, 0) == -1)
{
return -2;
}
//cout << "Read id: " << id << endl;
// create a tcp connection to the mary server
maryDataSocket = socket (AF_INET, SOCK_STREAM, 0);
// verify that the socket could be opened successfully
if (maryDataSocket == -1)
{
return -2;
}
else
// autoflush stdout, bind and connect
{
maryClient.sin_family = AF_INET;
maryClient.sin_port = htons (0);
maryClient.sin_addr.s_addr = INADDR_ANY;
int status = bind (maryDataSocket, (struct sockaddr*) &maryClient, sizeof (maryClient));
if (status != 0)
{
return -2;
}
maryServer.sin_family = AF_INET;
maryServer.sin_port = htons (server_port);
memcpy ((char*) &maryServer.sin_addr.s_addr, hostInfo->h_addr_list [0], hostInfo->h_length);
status = connect (maryDataSocket, (struct sockaddr*) &maryServer, sizeof (maryServer));
if (status != 0)
{
return -2;
}
}
// send the request id to the Mary server
if (send (maryDataSocket, id, strlen (id), 0) == -1)
{
return -2;
}
//cout << "Sending request: " << inputText << endl;
// send the query to the Mary server
if (send (maryDataSocket, inputText.c_str (), inputText.size (), 0) == -1)
{
return -2;
}
if (send (maryDataSocket, "\012\015", 2, 0) == -1)
{
return -2;
}
// shutdown data socket
shutdown (maryDataSocket, 1);
//cout << "Reading result" << endl;
unsigned int total_bytes = 0;
int recv_bytes = 0;
char data [1024] = "";
result [0] = '\0';
// receive the request result
do
{
data [0] = '\0';
recv_bytes = recv (maryDataSocket, data, 1024, 0);
if (recv_bytes == -1)
{
return -2;
}
else if (recv_bytes > 0)
{
//cout << "("<<recv_bytes<<")" << endl;
total_bytes += recv_bytes;
data [recv_bytes] = '\0';
if (maryOutFormat == "AUDIO")
{
for (unsigned int i=0; i<recv_bytes; i++)
{
result += data [i];
}
}
else
{
result += data;
}
}
} while (recv_bytes != 0);
if (result.size () != total_bytes)
{
cerr << "error: total bytes received != result bytes!" << endl;
cerr << " total bytes received = " << total_bytes << endl;
cerr << " result bytes = " << result.size () << endl;
}
// receive the request error
do
{
data [0] = '\0';
recv_bytes = recv (maryInfoSocket, data, 1024, 0);
if (recv_bytes == -1)
{
return -2;
}
else if (recv_bytes > 0)
{
cerr << endl << "Mary error code: " << data << endl;
return -3;
}
} while (recv_bytes != 0);
// close all open sockets
close (maryInfoSocket);
close (maryDataSocket);
return 0;
}

290
external/marytts-5.1.2/doc/examples/client/c++/MaryClient.cc.win vendored Normal file
View file

@ -0,0 +1,290 @@
/**
* Copyright 2000-2006 DFKI GmbH.
* All Rights Reserved. Use is subject to license terms.
*
* Permission is hereby granted, free of charge, to use and distribute
* this software and its documentation without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of this work, and to
* permit persons to whom this work is furnished to do so, subject to
* the following conditions:
*
* 1. The code must retain the above copyright notice, this list of
* conditions and the following disclaimer.
* 2. Any modifications must be clearly marked as such.
* 3. Original authors' names are not deleted.
* 4. The authors' names are not used to endorse or promote products
* derived from this software without specific prior written
* permission.
*
* DFKI GMBH AND THE CONTRIBUTORS TO THIS WORK DISCLAIM ALL WARRANTIES WITH
* REGARD TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS, IN NO EVENT SHALL DFKI GMBH NOR THE
* CONTRIBUTORS BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL
* DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
* PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
* ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF
* THIS SOFTWARE.
*/
#ifdef _WIN32
// use compiler option -L/<path to>/libwsock32.a
#include <winsock.h>
#else
#include <netdb.h>
#endif
#include <stdlib.h>
#include <iostream>
#include <string.h>
#include "MaryClient.h"
using namespace std;
/**
* A C++ implementation of a simple client to the MARY TTS system.
* result: an empty string serving as the container for the output.
* It will return text or audio data; text data will be encoded as UTF-8.
* inputText: the UTF-8 encoded text (or XML document) to send as a request
* maryInFormat: the input type of the data in inputText, e.g. TEXT
* maryOutFormat: the output type to produce, e.g. MBROLA, AUDIO
* locale: the language of the input, e.g. EN-US, DE
* audioType: for AUDIO output, the type of audio data to produce,
* e.g. WAVE or MP3.
* voice: the voice to be used, e.g. cmu-slt-hsmm, bits3.
* effects: the list of effects to be generated.
* return value: 0 on success, negative on failure.
*/
int
MaryClient::maryQuery( int server_port,
string server_host,
string& result,
string inputText,
string maryInFormat,
string maryOutFormat,
string locale,
string audioType,
string voice,
string effects ) {
// prepare the request
string query = "MARY";
query += " IN=" + maryInFormat;
query += " OUT=" + maryOutFormat;
query += " LOCALE=" + locale; // remove this line, if using an older version than MARY 4.0
query += " AUDIO=" + audioType;
query += " VOICE=" + voice;
if (effects != "") {
query += " EFFECTS=" + effects;
}
query += "\012\015";
//cout << "Constructed query: " << query << endl;
// declare connection stuff
struct sockaddr_in maryServer;
struct sockaddr_in maryClient;
struct hostent* hostInfo;
// declare variables
int maryInfoSocket;
int maryDataSocket;
// set configuration parameters
// get host information
hostInfo = gethostbyname (server_host.c_str());
if (hostInfo == NULL)
{
return -2;
}
// create a tcp connection to the mary server
maryInfoSocket = socket (AF_INET, SOCK_STREAM, 0);
// verify that the socket could be opened successfully
if (maryInfoSocket == -1)
{
return -2;
}
else
// autoflush stdout, bind and connect
{
maryClient.sin_family = AF_INET;
maryClient.sin_port = htons (0);
maryClient.sin_addr.s_addr = INADDR_ANY;
int status = bind (maryInfoSocket, (struct sockaddr*) &maryClient, sizeof (maryClient));
if (status != 0)
{
return -2;
}
maryServer.sin_family = AF_INET;
maryServer.sin_port = htons (server_port);
memcpy ((char*) &maryServer.sin_addr.s_addr, hostInfo->h_addr_list [0], hostInfo->h_length);
status = connect (maryInfoSocket, (struct sockaddr*) &maryServer, sizeof (maryServer));
if (status != 0)
{
return -2;
}
}
// send request to the Mary server
if (send (maryInfoSocket, query.c_str (), query.size (), 0) == -1)
{
return -2;
}
// receive the request id
char id [32] = "";
if (recv (maryInfoSocket, id, 32, 0) == -1)
{
return -2;
}
//cout << "Read id: " << id << endl;
// create a tcp connection to the mary server
maryDataSocket = socket (AF_INET, SOCK_STREAM, 0);
// verify that the socket could be opened successfully
if (maryDataSocket == -1)
{
return -2;
}
else
// autoflush stdout, bind and connect
{
maryClient.sin_family = AF_INET;
maryClient.sin_port = htons (0);
maryClient.sin_addr.s_addr = INADDR_ANY;
int status = bind (maryDataSocket, (struct sockaddr*) &maryClient, sizeof (maryClient));
if (status != 0)
{
return -2;
}
maryServer.sin_family = AF_INET;
maryServer.sin_port = htons (server_port);
memcpy ((char*) &maryServer.sin_addr.s_addr, hostInfo->h_addr_list [0], hostInfo->h_length);
status = connect (maryDataSocket, (struct sockaddr*) &maryServer, sizeof (maryServer));
if (status != 0)
{
return -2;
}
}
// send the request id to the Mary server
if (send (maryDataSocket, id, strlen (id), 0) == -1)
{
return -2;
}
//cout << "Sending request: " << inputText << endl;
// send the query to the Mary server
if (send (maryDataSocket, inputText.c_str (), inputText.size (), 0) == -1)
{
return -2;
}
if (send (maryDataSocket, "\012\015", 2, 0) == -1)
{
return -2;
}
// shutdown data socket
shutdown (maryDataSocket, 1);
//cout << "Reading result" << endl;
unsigned int total_bytes = 0;
int recv_bytes = 0;
char data [1024] = "";
result [0] = '\0';
// receive the request result
do
{
data [0] = '\0';
recv_bytes = recv (maryDataSocket, data, 1024, 0);
if (recv_bytes == -1)
{
return -2;
}
else if (recv_bytes > 0)
{
//cout << "("<<recv_bytes<<")" << endl;
total_bytes += recv_bytes;
data [recv_bytes] = '\0';
if (maryOutFormat == "AUDIO")
{
for (unsigned int i=0; i<recv_bytes; i++)
{
result += data [i];
}
}
else
{
result += data;
}
}
} while (recv_bytes != 0);
if (result.size () != total_bytes)
{
cerr << "error: total bytes received != result bytes!" << endl;
cerr << " total bytes received = " << total_bytes << endl;
cerr << " result bytes = " << result.size () << endl;
}
// receive the request error
do
{
data [0] = '\0';
recv_bytes = recv (maryInfoSocket, data, 1024, 0);
if (recv_bytes == -1)
{
return -2;
}
else if (recv_bytes > 0)
{
cerr << endl << "Mary error code: " << data << endl;
return -3;
}
} while (recv_bytes != 0);
#ifdef _WIN32
// close all open sockets Windows
closesocket (maryInfoSocket);
closesocket (maryDataSocket);
#else
// close all open sockets UNIX
close (maryInfoSocket);
close (maryDataSocket);
#endif
return 0;
}

49
external/marytts-5.1.2/doc/examples/client/c++/MaryClient.h vendored Normal file
View file

@ -0,0 +1,49 @@
/**
* Copyright 2000-2006 DFKI GmbH.
* All Rights Reserved. Use is subject to license terms.
*
* Permission is hereby granted, free of charge, to use and distribute
* this software and its documentation without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of this work, and to
* permit persons to whom this work is furnished to do so, subject to
* the following conditions:
*
* 1. The code must retain the above copyright notice, this list of
* conditions and the following disclaimer.
* 2. Any modifications must be clearly marked as such.
* 3. Original authors' names are not deleted.
* 4. The authors' names are not used to endorse or promote products
* derived from this software without specific prior written
* permission.
*
* DFKI GMBH AND THE CONTRIBUTORS TO THIS WORK DISCLAIM ALL WARRANTIES WITH
* REGARD TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS, IN NO EVENT SHALL DFKI GMBH NOR THE
* CONTRIBUTORS BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL
* DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
* PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
* ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF
* THIS SOFTWARE.
*/
#ifndef __MARYCLIENT_H_
#define __MARYCLIENT_H_
class MaryClient {
public:
// send query to the MaryServer
int maryQuery( int server_port,
std::string server_host,
std::string& result,
std::string inputText,
std::string maryInFormat,
std::string maryOutFormat,
std::string locale,
std::string audioType,
std::string voice,
std::string effects = "" );
};
#endif

87
external/marytts-5.1.2/doc/examples/client/c++/MaryDemo.cc vendored Normal file
View file

@ -0,0 +1,87 @@
/**
* Copyright 2000-2006 DFKI GmbH.
* All Rights Reserved. Use is subject to license terms.
*
* Permission is hereby granted, free of charge, to use and distribute
* this software and its documentation without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of this work, and to
* permit persons to whom this work is furnished to do so, subject to
* the following conditions:
*
* 1. The code must retain the above copyright notice, this list of
* conditions and the following disclaimer.
* 2. Any modifications must be clearly marked as such.
* 3. Original authors' names are not deleted.
* 4. The authors' names are not used to endorse or promote products
* derived from this software without specific prior written
* permission.
*
* DFKI GMBH AND THE CONTRIBUTORS TO THIS WORK DISCLAIM ALL WARRANTIES WITH
* REGARD TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS, IN NO EVENT SHALL DFKI GMBH NOR THE
* CONTRIBUTORS BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL
* DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
* PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
* ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF
* THIS SOFTWARE.
*/
// This version, adapted to MARY 4.0, provided by Sebastian Ptock.
#include <iostream>
#include <fstream>
#include <stdlib.h>
#include "MaryClient.h"
using namespace std;
/**
* Demonstration code for using the MaryClient.
+ Call this as:
* ./MaryDemo
* or
* ./MaryDemo > output.wav
*/
int main() {
int server_port = 59125;
string server_host = "localhost";
string inputText = "Welcome to the world of speech synthesis!";
string maryInFormat = "TEXT";
string maryOutFormat = "AUDIO";
//string maryOutFormat = "REALISED_DURATIONS";
string locale = "en-US";
string audioType = "WAV_FILE";
string voice = "cmu-slt-hsmm";
string effects;
// effects += "Volume(amount:5.0;)+";
// effects += "TractScaler(amount:1.5;)+";
// effects += "F0Scale(f0Scale:2.0;)+";
// effects += "F0Add(f0Add:50.0;)+";
// effects += "Rate(durScale:1.5;)+";
// effects += "Robot(amount:100.0;)+";
// effects += "Whisper(amount:100.0;)+";
// effects += "Stadium(amount:100.0)+";
// effects += "Chorus(delay1:466;amp1:0.54;delay2:600;amp2:-0.10;delay3:250;amp3:0.30)+";
// effects += "FIRFilter(type:3;fc1:500.0;fc2:2000.0)+";
// effects += "JetPilot";
string result;
MaryClient maryClient;
maryClient.maryQuery( server_port, server_host, result, inputText, maryInFormat, maryOutFormat, locale, audioType, voice, effects);
if (maryOutFormat == "AUDIO") {
// write result into a file
const char *filename = "output.wav";
ofstream file( filename );
file << result;
// play output
//system("play output.wav");
} else {
cout << "RESULT: " << endl << result << endl;
}
return 0;
}

4
external/marytts-5.1.2/doc/examples/client/c++/README.txt vendored Normal file
View file

@ -0,0 +1,4 @@
Start MARY as a socket server:
maryserver -Dserver=socket
(or change entry 'server' in conf/marybase.config)

185
external/marytts-5.1.2/doc/examples/client/maryclient-http.py vendored Normal file
View file

@ -0,0 +1,185 @@
#!/usr/bin/env python
import httplib, urllib
# A basic mary client in Python,
# kindly donated to the MARY TTS project
# by Hugh Sasse. Thanks Hugh!
# A very basic Python class for accessing
# the MARY TTS system using the modern
# HTTP server.
# Warning, this is probably ghastly Python,
# most of my time of late has been with
# other languages, so I'm not up to date
# with all the stylistic conventions of
# modern Python.
# This does seem to work OK though.
class maryclient:
"""A basic handler for MARY-TTS HTTP clients
At present, there is no checking for
allowed voices, locales, and so on.
Most of the useful parameters can be
accessed by get_ and set_ methods.
Relying on winsound, this is Windows
specific.
"""
def __init__(self):
"""Set up useful defaults (for
people in England, anyway)"""
self.host = "127.0.0.1"
self.port = 59125
self.input_type = "TEXT"
self.output_type = "AUDIO"
self.audio = "WAVE_FILE"
self.locale = "en_GB"
self.voice = "dfki-prudence-hsmm"
def set_host(self, a_host):
"""Set the host for the TTS server."""
self.host = a_host
def get_host(self):
"""Get the host for the TTS server."""
self.host
def set_port(self, a_port):
"""Set the port for the TTS server."""
self.port = a_port
def get_port(self):
"""Get the port for the TTS server."""
self.port
def set_input_type(self, type):
"""Set the type of input being
supplied to the TTS server
(such as 'TEXT')."""
self.input_type = type
def get_input_type(self):
"""Get the type of input being
supplied to the TTS server
(such as 'TEXT')."""
self.input_type
def set_output_type(self, type):
"""Set the type of input being
supplied to the TTS server
(such as 'AUDIO')."""
self.output_type = type
def get_output_type(self):
"""Get the type of input being
supplied to the TTS server
(such as "AUDIO")."""
self.output_type
def set_locale(self, a_locale):
"""Set the locale
(such as "en_GB")."""
self.locale = a_locale
def get_locale(self):
"""Get the locale
(such as "en_GB")."""
self.locale
def set_audio(self, audio_type):
"""Set the audio type for playback
(such as "WAVE_FILE")."""
self.audio = audio_type
def get_audio(self):
"""Get the audio type for playback
(such as "WAVE_FILE")."""
self.audio
def set_voice(self, a_voice):
"""Set the voice to speak with
(such as "dfki-prudence-hsmm")."""
self.voice = a_voice
def get_voice(self):
"""Get the voice to speak with
(such as "dfki-prudence-hsmm")."""
self.voice
def generate(self, message):
"""Given a message in message,
return a response in the appropriate
format."""
raw_params = {"INPUT_TEXT": message,
"INPUT_TYPE": self.input_type,
"OUTPUT_TYPE": self.output_type,
"LOCALE": self.locale,
"AUDIO": self.audio,
"VOICE": self.voice,
}
params = urllib.urlencode(raw_params)
headers = {}
# Open connection to self.host, self.port.
conn = httplib.HTTPConnection(self.host, self.port)
# conn.set_debuglevel(5)
conn.request("POST", "/process", params, headers)
response = conn.getresponse()
if response.status != 200:
print response.getheaders()
raise RuntimeError("{0}: {1}".format(response.status,
response.reason))
return response.read()
# If this is invoked as a program, just give
# a greeting to show it is working.
# The platform specific code is moved to this
# part so that this file may be imported without
# bringing platform specific code in.
if __name__ == "__main__":
# For handling command line arguments:
import sys
import platform
# check we are on Windows:
system = platform.system().lower()
if (system == "windows"):
import winsound
class Player:
def __init__(self):
pass
def play(self, a_sound):
winsound.PlaySound(a_sound, winsound.SND_MEMORY)
#if ("cygwin" in system):
else:
# Not sure how to do audio on cygwin,
# portably for python. So have a sound
# player class that doesn't play sounds.
# A null object, if you like.
class Player:
def __init__(self):
pass
def play(self, a_sound):
print("Here I would play a sound if I knew how")
pass
# Probably want to parse arguments to
# set the voice, etc., here
client = maryclient()
client.set_audio("WAVE_FILE") # for example
player = Player()
the_sound = client.generate("hello from Mary Text to Speech, with Python.")
if client.output_type == "AUDIO":
player.play(the_sound)
# vi:set sw=4 et:

177
external/marytts-5.1.2/doc/examples/client/maryclient.cgi vendored Normal file
View file

@ -0,0 +1,177 @@
#!/usr/bin/perl -T
# -*- Mode: Perl -*-
# MARY Text-to-Speech System
# CGI Script implementing a simple mary client,
# can be used for web pages.
##########################################################################
# Copyright (C) 2000-2006 DFKI GmbH.
# All rights reserved. Use is subject to license terms.
#
# Permission is hereby granted, free of charge, to use and distribute
# this software and its documentation without restriction, including
# without limitation the rights to use, copy, modify, merge, publish,
# distribute, sublicense, and/or sell copies of this work, and to
# permit persons to whom this work is furnished to do so, subject to
# the following conditions:
#
# 1. The code must retain the above copyright notice, this list of
# conditions and the following disclaimer.
# 2. Any modifications must be clearly marked as such.
# 3. Original authors' names are not deleted.
# 4. The authors' names are not used to endorse or promote products
# derived from this software without specific prior written
# permission.
#
# DFKI GMBH AND THE CONTRIBUTORS TO THIS WORK DISCLAIM ALL WARRANTIES WITH
# REGARD TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF
# MERCHANTABILITY AND FITNESS, IN NO EVENT SHALL DFKI GMBH NOR THE
# CONTRIBUTORS BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL
# DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
# PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
# ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF
# THIS SOFTWARE.
##########################################################################
# Author: Marc Schroeder
use strict;
use IO::Socket;
use CGI;
# variables getting their values from form:
my ($inputtext, $in, $out, $audiotype, $voice);
# little helpers:
my ($var, $tmp);
# contacting the mary server:
my ($host, $port, $maryInfoSocket, $maryDataSocket, $id);
# helping with audio output:
my ($save_to_disk, $audiosubtype, $filename);
my $cgi = new CGI;
my @param = $cgi->param();
$inputtext = $cgi->param('inputtext');
$in = $cgi->param('in');
$out = $cgi->param('out');
$audiotype = $cgi->param('audiotype');
$save_to_disk = $cgi->param('save_to_disk');
$voice = $cgi->param('voice');
my ($sec,$min,$hour,$mday,$mon,$year,$wday,$yday,$isdst)=localtime(time);
$year += 1900;
printf STDERR "[%04i-%02i-%02i %02i:%02i:%02i] ", $year, $mon, $mday, $hour, $min, $sec;
print STDERR "Request from ",$cgi->remote_user(),"@",$cgi->remote_host(),": \n";
print STDERR " in=",$in;
print STDERR " out=",$out;
print STDERR " audiotype=",$audiotype;
print STDERR " voice=",$voice;
print STDERR " save_to_disk=",$save_to_disk,"\n";
print STDERR " inputtext: ";
print STDERR $inputtext,"\n";
# Limit inputtext length to 5000 bytes:
if (length $inputtext > 5000) {
$inputtext = substr $inputtext, 0, 5000;
}
# set audio subtype
if ($out eq "AUDIO") {
if ($audiotype eq "AU") {
$audiosubtype = "basic";
$filename = "mary.au";
} elsif ($audiotype eq "AIFF") {
$audiosubtype = "x-aiff";
$filename = "mary.aiff";
} elsif ($audiotype eq "WAVE") {
$audiosubtype = "x-wav";
$filename = "mary.wav";
} elsif ($audiotype eq "MP3") {
$audiosubtype = "mp3";
$filename = "mary.mp3";
} else {
$audiosubtype = "x-wav";
$filename = "mary.wav";
}
}
# announce data type on stdout
if ($save_to_disk) {
print "Content-Type: application/octet-stream";
} else {
print "Content-Type: audio/$audiosubtype";
}
print "\nContent-Disposition: filename=\"$filename\"\n\n";
# contact mary server
$host = "cling.dfki.uni-sb.de";
$port = 59125;
# create a tcp connection to the specified host and port
$maryInfoSocket = IO::Socket::INET->new(Proto => "tcp",
PeerAddr => $host,
PeerPort => $port)
or die "can't connect to port $port on $host: $!";
# avoid buffering when writing to server:
$maryInfoSocket->autoflush(1); # so output gets there right away
########## Write input to server: ##########
# formulate the request:
print $maryInfoSocket "MARY IN=$in OUT=$out AUDIO=$audiotype";
if ($voice && $voice ne 'v') { print $maryInfoSocket " VOICE=$voice"; }
print $maryInfoSocket " LOG=\"REMOTE_HOST=$ENV{'REMOTE_HOST'}",
", REMOTE_ADDR=$ENV{'REMOTE_ADDR'}\"";
print $maryInfoSocket "\015\012";
# receive a request ID:
$id = <$maryInfoSocket>;
# open second socket for the data:
$maryDataSocket = IO::Socket::INET->new(Proto => "tcp",
PeerAddr => $host,
PeerPort => $port)
or die "can't connect to port $port on $host: $!";
# identify with request number:
print $maryDataSocket $id; # $id contains a newline character
# copy $inputtext to mary data socket
print $maryDataSocket $inputtext;
# mark end-of-request:
print $maryDataSocket "\015\012"; # that is a \n, actually
$maryDataSocket->shutdown(1); # we have stopped writing data
########## Read output from server: ##########
# copy the data socket to standard output
if ($out ne "AUDIO") { # text output
my $line;
while (defined ($line = <$maryDataSocket>)) {
print STDOUT $line;
}
} else { # audio data output
my $nr; # number of bytes read
my $buf; # buffer to read into
my $outnr; # number of bytes written
while($nr = read($maryDataSocket, $buf, 8192)) {
# (read returns no. of bytes read, 0 at eof)
print STDOUT $buf
or die "Write error on stdout";
} # while read something from socket
} # audio output
### Read complaints from server:
my $line;
while (defined ($line = <$maryInfoSocket>)) {
print STDERR $line;
}

136
external/marytts-5.1.2/doc/examples/client/maryclient.pl vendored Normal file
View file

@ -0,0 +1,136 @@
#!/usr/bin/env perl
#
# MARY Text-to-Speech System
# Minimal Socket client (for demonstration)
##########################################################################
# Copyright (C) 2000-2006 DFKI GmbH.
# All rights reserved. Use is subject to license terms.
#
# Permission is hereby granted, free of charge, to use and distribute
# this software and its documentation without restriction, including
# without limitation the rights to use, copy, modify, merge, publish,
# distribute, sublicense, and/or sell copies of this work, and to
# permit persons to whom this work is furnished to do so, subject to
# the following conditions:
#
# 1. The code must retain the above copyright notice, this list of
# conditions and the following disclaimer.
# 2. Any modifications must be clearly marked as such.
# 3. Original authors' names are not deleted.
# 4. The authors' names are not used to endorse or promote products
# derived from this software without specific prior written
# permission.
#
# DFKI GMBH AND THE CONTRIBUTORS TO THIS WORK DISCLAIM ALL WARRANTIES WITH
# REGARD TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF
# MERCHANTABILITY AND FITNESS, IN NO EVENT SHALL DFKI GMBH NOR THE
# CONTRIBUTORS BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL
# DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
# PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
# ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF
# THIS SOFTWARE.
##########################################################################
# Author: Marc Schroeder
# This is a minimal version of a socket client for the mary TtS system.
# It is intended to be used as a model for writing socket clients for
# particular applications. All input verification, command line options,
# and other luxury have been omitted.
#
# Usage:
# maryclient.pl infile.txt > outfile.wav
#
# Input/output formats and other options must be set in the perl code directly.
# See also Protocol.html for a description of the Protocol.
#
use strict;
use IO::Socket;
############################
# Package-global variables #
############################
# global settings:
my $maryInfoSocket; # handle to socket server
my $maryDataSocket; # handle to socket server
my $host; # string containing host address
my $port; # socket port on which we listen
my ($in, $out, $audiotype); # requested input / output format
my $voice; # default voice
my $id; # request ID
######################################################################
################################ main ################################
######################################################################
STDOUT->autoflush(1);
$host = "cling.dfki.uni-sb.de";
$port = 59125;
$in = "TEXT_DE";
$out = "AUDIO";
$audiotype = "MP3";
#$audiotype = "WAVE";
#$voice = "male";
$voice = "de3";
# create a tcp connection to the specified host and port
$maryInfoSocket = IO::Socket::INET->new(Proto => "tcp",
PeerAddr => $host,
PeerPort => $port)
or die "can't connect to port $port on $host: $!";
# avoid buffering when writing to server:
$maryInfoSocket->autoflush(1); # so output gets there right away
########## Write input to server: ##########
# formulate the request:
print $maryInfoSocket "MARY IN=$in OUT=$out AUDIO=$audiotype";
if ($voice) { print $maryInfoSocket " VOICE=$voice"; }
print $maryInfoSocket "\015\012";
# receive a request ID:
$id = <$maryInfoSocket>;
chomp $id; chomp $id;
# open second socket for the data:
$maryDataSocket = IO::Socket::INET->new(Proto => "tcp",
PeerAddr => $host,
PeerPort => $port)
or die "can't connect to port $port on $host: $!";
# identify with request number:
print $maryDataSocket $id, "\015\012";
# copy standard input and/or files given on the command line to the socket
while (defined (my $line = <>)) {
print $maryDataSocket $line;
}
# mark end-of-request:
print $maryDataSocket "\015\012"; # that is a \n, actually
shutdown($maryDataSocket, 1); # we have stopped writing data
########## Read output from server: ##########
# copy the data socket to standard output
if ($out ne "AUDIO") { # text output
my $line;
while (defined ($line = <$maryDataSocket>)) {
print STDOUT $line;
}
} else { # audio data output
my $nr; # number of bytes read
my $buf; # buffer to read into
my $outnr; # number of bytes written
while($nr = read($maryDataSocket, $buf, 100000)) {
# (read returns no. of bytes read, 0 at eof)
print STDOUT $buf
or die "Write error on stdout";
} # while read something from socket
} # audio output
### Read complaints from server:
my $line;
while (defined ($line = <$maryInfoSocket>)) {
print STDERR $line;
}

261
external/marytts-5.1.2/doc/examples/client/maryclient.rb vendored Normal file
View file

@ -0,0 +1,261 @@
#!/usr/bin/env ruby
#
# A basic mary client in Ruby,
# kindly donated to the MARY TTS project
# by Hugh Sasse. Thanks Hugh!
# Ruby client for the MARY TTS HTTP server.
# This is for Windows only, and relies on
# the Win32-Sound gem to access the audio.
#
#
require 'rubygems'
require 'net/http'
require 'uri'
# A fairly minimal client class for the
# MARY TTS system. This uses the modern
# HTTP interface to access the server.
# At present, this doesn't wrap the methods
# which provide documentation or lists of
# voices or features.
class MaryClient
attr_accessor :host, :port
attr_accessor :input_type, :output_type
attr_accessor :locale, :audio, :voice
# Set up the defaults for the MARY TTS
# server, which is assumed to be running
# on the local host, with British voices
# installed. These may be modified with
# the appropriate methods.
# host = 127.0.0.1)
# port = 59125
# input_type = "TEXT"
# output_type = "AUDIO"
# audio = "WAVE_FILE"
# locale = "en_GB"
# voice = "dfki-prudence-hsmm"
def initialize
@host = "127.0.0.1" # The local machine
@port = 59125
@input_type = "TEXT"
@output_type = "AUDIO"
@locale = "en_GB"
@audio = "WAVE_FILE"
@voice = "dfki-prudence-hsmm"
end
# Process a text message, which with a
# new client, will return the audio.
# This is so that platform dependent parts
# are kept separate.
def generate(message)
raw_params = {"INPUT_TEXT" => message,
"INPUT_TYPE" => @input_type,
"OUTPUT_TYPE" => @output_type,
"LOCALE" => @locale,
"AUDIO" => @audio,
"VOICE" => @voice,
}
res = Net::HTTP.post_form(URI.parse("http://#{@host}:#{@port}/process"), raw_params)
res.value # Throw an exception on failure
#puts res.body
return res.body
end
end
# If this invoked as a program with no
# argumens, just give a greeting to show
# that it is working. If arguments are
# supplied, process options to work out
# what to do with the arguments.
if __FILE__ == $0
# These files are only loaded when this is
# invoked as a program.
require 'rbconfig'
require 'getoptlong'
# PLATFORM SPECIFIC CODE.
# Needs more work [!]
case Config::CONFIG['host_os']
when /darwin/i
raise NotImplementedError.new("Don't know how to play audio on a Mac")
when /linux/i
raise NotImplementedError.new("Far too many ways to play audio on Linux, you'll need to choose something")
when /sunos|solaris/i
raise NotImplementedError.new("Have not played audio on Suns for too long to implement this.")
when /java/i
raise NotImplementedError.new("Don't know how to play audio from Java ")
when /win32|cygwin|mingw32/i
# The various things that can use the Win32
# sound gem
require 'win32/sound'
# Create a player class that will play the
# sound that the Mary TTS system returns
class Player
# Play the audio passed in.
# Possibly this should receive the audio
# type so we can check that we can play it,
# but at the moment that is the
# responsibility of the user.
def self.play(sound)
Win32::Sound.play(sound, Win32::Sound::MEMORY)
end
end
else
raise NotImplementedError.new("Haven't thought how to support this OS yet")
end
client = nil
split = ""
if ARGV.size.zero?
client = MaryClient.new()
sound = client.generate("Hello from Mary Text to Speech with Ruby.")
Player.play(sound)
else
args_mode = :words
stdout_mode = :absorb
opts = GetoptLong::new(
["--audio", "-a", GetoptLong::REQUIRED_ARGUMENT],
["--echo", "-e", GetoptLong::NO_ARGUMENT],
["--help", "-h", GetoptLong::NO_ARGUMENT],
["--host", "-H", GetoptLong::REQUIRED_ARGUMENT],
["--input-type", "-i", GetoptLong::REQUIRED_ARGUMENT],
["--locale", "-l", GetoptLong::REQUIRED_ARGUMENT],
["--read", "-r", GetoptLong::NO_ARGUMENT],
["--split", "-s", GetoptLong::REQUIRED_ARGUMENT],
["--output-type", "-o", GetoptLong::REQUIRED_ARGUMENT],
["--port", "-P", GetoptLong::REQUIRED_ARGUMENT],
["--tee", "-t", GetoptLong::NO_ARGUMENT],
["--voice", "-v", GetoptLong::REQUIRED_ARGUMENT]
)
opts.each do |opt, arg|
unless ["--help", "-h"].include?(opt)
# skip if we are only getting help
client ||= MaryClient.new()
end
case opt
when "--help", "-h"
puts <<-EOHELP
Usage: #{$0} [options] [arguments]
--audio -a
Audio format. Defualt: WAVE_FILE
--echo -e
Act as an echo command and send output
arguments to the synthesizer only (not
to standard output.
Turns off --read|-r
--help -h
Print this help, then exit.
--host -H
The host which is the server.
Default: 127.0.0.1
--input-type -i
The type of the input supplied to the
TTS system. Default: TEXT
--locale -l
The locale of the input. Default: en_GB
--output-type -o
The output type from the TTS system.
Default: AUDIO
--port -P
The port for the TTS server
Default: 59125
--read -r
Read the files passed as arguments.
Turns off --echo|-e
--split -s (lines|paragraphs)
When reading files, split the input
into lines or paragraphs. Paragraphs
mean reading up to the next double
newline. Note, the argument is literally
"lines" or "paragraphs" (or some
abbreviation of those) without the
quotes.
Default is paragraphs.
--tee -t
Act as tee: send the output to the TTS
system, and to standard output.
--voice -v
The voice to use.
Default: dfki-prudence-hsmm
EOHELP
exit(0)
when "--audio", "-a"
client.audio = arg
when "--echo", "-e"
args_mode = :words
when "--host", "-H"
client.host = arg
when "--input-type", "-i"
client.input_type = arg
when "--locale", "-l"
client.locale = arg
when "--output-type", "-o"
client.output_type = arg
when "--port", "-P"
client.port = arg.to_i
when "--read", "-r"
args_mode = :files
when "--split", "-s"
case arg
when /^p/i
split = ""
when /^l/i
split = $/
end
when "--tee", "-t"
stdout_mode = :emit
when "--voice", "-v"
client.voice = arg
end
end
client ||= MaryClient.new()
case args_mode
when :words
input_text = ARGV.join(" ")
unless input_text =~ /\A\s*\Z/m
sound = client.generate(input_text)
if client.output_type == "AUDIO"
Player.play(sound)
end
end
if stdout_mode == :emit
puts input_text
end
when :files
# Slurp in paragraphs so sentences
# don't get broken in stupid places.
$/ = split # paragraph mode
ARGF.each do |paragraph|
begin
unless paragraph =~ /\A\s*\Z/m
sound = client.generate(paragraph)
if client.output_type == "AUDIO"
# and client.audio == "WAVE_FILE"
Player.play(sound)
end
end
rescue Exception => e
puts "got error #{e} while trying to say #{paragraph.inspect}"
raise
end
if stdout_mode == :emit
puts paragraph
end # end if
end # end ARGF.each
end # end case
end # if ARGV.size.zero?
end

705
external/marytts-5.1.2/doc/examples/client/maryclient.tcl vendored Normal file
View file

@ -0,0 +1,705 @@
# Tcl/Tk MARY TTS client.
# This has been tested on Windows, and because
# of the use of sound there will be portability
# issues. However, there should be enough here
# for a reasonable start at a client, for any
# platform that supports Tcl/Tk. The platform
# specific code has, as far as possible, been
# isolated in the part of the code that detects
# whether this is being run as a program.
# Notes:
# More work will need to be done with this,
# in order to make the code clean. It should
# probably be wrapped in a package, to solve
# any namespace issues. There are a lot of
# global variables. It seems that some of
# these are necessary for the menus to work.
# Handling of temporary files could be improved.
# TODO:
# Create modifier sliders, for the effects.
# Extend the query proc to make use of them.
# Turn the Help menu into something more useful.
# Debug the actions for the Edit menu.
# Provide a means of getting example inputs
# from the server.
# Provide a means of re-loading all the
# dynamically collected information when the
# server is changed from the menu. This means
# that we need to delete the existing menu
# entries in order to add them correctly.
# How do we ensure temporary files are removed
# in the event of a problem? if {catch {}} ...?
# Maybe leaving them around is diagnostic info?
# Make that an option?
# Add error handling code for network and disk
# failures likely to beset such clients.
# Add sensible defaults for things the user must
# always set at startup, but these will be
# platform spacific. Always default to Audio
# output for example, or is it possible that
# people have no voices installed?
# This is a GUI, so:
package require Tk
# We are communicating with the Mary server
# with HTTP.
package require http
# Use the local machine in preference to the
# one in Germany.
set mary_tts_default_host "127.0.0.1"
set mary_tts_default_port 59125
# Actual host and port, and global old
# copies to allow revert on cancel in the
# dialogues. Apparently upvar #0 is the
# norm for that sort of thing [Tcl Wiki]
set mary_tts_host $mary_tts_default_host
set old_mary_tts_host $mary_tts_host
set mary_tts_port $mary_tts_default_port
set old_mary_tts_port $mary_tts_port
# Informational URLs
set informational_urls [ list \
version datatypes voices \
audioformats audioeffects ]
#######
# Obtain a static page from the server, i.e.
# no parameters are needed to get it.
proc get_page { relative_url } {
global mary_tts_host mary_tts_port
set url http://$mary_tts_host:$mary_tts_port/$relative_url
set result [::http::geturl $url]
return [::http::data $result]
}
proc list_of_lines {str} {
return [ split $str "\n" ]
}
# We will need to collect this information
# when we have the server and port chosen.
proc get_audioeffects {} {
return [list_of_lines [get_page audioeffects] ]
}
proc get_audioformats {} {
return [list_of_lines [get_page audioformats] ]
}
proc get_datatypes {} {
return [ list_of_lines [get_page datatypes] ]
}
proc get_voices {} {
return [list_of_lines [get_page voices] ]
}
# Handling post queries.
# Submit the query to the server, using the
# http POST method.
proc make_query {url encoded_params} {
set http [::http::geturl $url -query $encoded_params]
set result [::http::data $http]
return $result
}
# Get the text from the input text area
proc get_input_text {} {
return [.io.inp.input_area get 1.0 end]
}
# Get the text from the output text area
proc get_output_text {} {
return [.io.out.output_area get 1.0 end]
}
# Collect the audio data from the server.
proc collect_audio_data {text_to_process} {
global mary_tts_host mary_tts_port
global inputtype outputtype locales
global audioformat voice
set url "http://$mary_tts_host:$mary_tts_port/process"
# ::http::formatQuery converts a list of
# key value pairs into the correct format
# for http POST.
set params [::http::formatQuery INPUT_TEXT $text_to_process INPUT_TYPE $inputtype OUTPUT_TYPE $outputtype LOCALE $locales($voice) AUDIO $audioformat VOICE $voice ]
set result [make_query $url $params]
return $result
}
# Pushes the query to the server and gets
# the results back, displaying or playing
# them.
proc generate_output {text_to_process} {
global outputtype
set result [collect_audio_data $text_to_process]
if {$outputtype eq "AUDIO"} {
# call the platform dependent implementation.
play $result
} else {
clear_output
add_message $result
}
# Return the result so we can save it if
# the user requires it.
return $result
}
# These next procs are for handling the
# lists of data one gets back from the server
# which possibly have several words per line,
# separated by spaces.
# If the first word of each listed line is
# significant, extract the list of first words.
proc collect_first_words_of_phrase_list {a_list} {
for {set i 0} {$i < [llength $a_list]} {incr i} {
set data [lindex $a_list $i ]
set word [ lindex [split $data " "] 0 ]
lappend words $word
}
return $words
}
# If the second word of each listed line is
# significant, extract the list of second words.
proc collect_second_words_of_phrase_list {a_list} {
for {set i 0} {$i < [llength $a_list]} {incr i} {
set data [lindex $a_list $i ]
set word [ lindex [split $data " "] 1 ]
lappend words $word
}
return $words
}
# The list of datatypes must be separated into
# input data types and output data types so that
# interactions with the server make sense.
# This handles the inputs.
proc collect_first_words_of_input_types {a_list} {
for {set i 0} {$i < [llength $a_list]} {incr i} {
set data [lindex $a_list $i ]
if {[ string match -nocase "*input*" $data ]} {
set word [ lindex [split $data " "] 0 ]
lappend words $word
}
}
return $words
}
# The list of datatypes must be separated into
# input data types and output data types so that
# interactions with the server make sense.
# This handles the outputs.
proc collect_first_words_of_output_types {a_list} {
for {set i 0} {$i < [llength $a_list]} {incr i} {
set data [lindex $a_list $i ]
if {[string match -nocase "*output*" $data]} {
set word [ lindex [split $data " "] 0 ]
lappend words $word
}
}
return $words
}
# setup all the variables to hold voices,
# audio options, etc., based on what the
# server can do.
proc setup_globals {} {
global audioeffects audioformats voices
global inputtypes outputtypes audioformat voice
global inputtype outputtype locales
set audioeffects [get_audioeffects]
set audioformats [get_audioformats]
set audioformat [lindex $audioformats 0 ]
set datatypes_data [get_datatypes]
set inputtypes [collect_first_words_of_input_types $datatypes_data]
set inputtype [lindex $inputtypes 0]
set outputtypes [collect_first_words_of_output_types $datatypes_data]
set outputtype [lindex $outputtypes 0]
set voices_data [get_voices]
set voices [collect_first_words_of_phrase_list $voices_data]
set locales_list [collect_second_words_of_phrase_list $voices_data ]
for {set i 0} {$i < [llength $voices]} {incr i} {
set locales([lindex $voices $i]) [lindex $locales_list $i]
}
set voice [lindex $voices 0]
}
# A general procedure for filling in the
# elements of a listbox from a list.
# At present this is unused, but it could
# be useful later. [It took a while to
# figure out so I'm not ready to kill it
# with YAGNI.]
proc add_listbox_items {a_var a_widget} {
upvar $a_var var
foreach item $var {
$a_widget insert end $item
}
}
# Create the menubuttons along the top.
# Usual File, Edit and Help menus plus
# those to set attributes.
proc create_menubuttons {} {
set buttons [ list file File edit Edit \
server "Server" \
inputtype "Input type" outputtype "Output type" \
voice Voice \
audioformat "Audio format" \
textstyle "Text style" help Help ]
set count 1
foreach { menu_tag string_tag} $buttons {
menubutton .menus.$menu_tag -text $string_tag \
-menu .menus.${menu_tag}.menu -underline 0 -font ClientFont
menu .menus.${menu_tag}.menu -tearoff true
grid .menus.$menu_tag -in .menus -row 1 -column $count -sticky w
incr count
}
}
# Get the contents of a text file for reading
# or loading into a text widget, etc.
proc text_file_contents {what_for} {
set a_file [tk_getOpenFile -title $what_for ]
set the_text ""
if {$a_file != ""} {
set a_stream [open $a_file r ]
set the_text [read $a_stream]
close $a_stream
}
return $the_text
}
# Save the_text to a text file specified
# by the user, for the given reason (what_for).
# At the moment there is no error handling
# for this (disk full, write protected, etc).
proc save_text_file {the_text what_for} {
set a_file [tk_getSaveFile -title $what_for -parent .]
if {$a_file != ""} {
set a_stream [open $a_file w ]
puts $a_stream $the_text
close $a_stream
}
}
# Save the_data to a binary file specified
# by the user, for the given reason (what_for),
# a text string.
# At the moment there is no error handling
# for this (disk full, write protected, etc).
proc save_binary_file {the_data what_for} {
set a_file [tk_getSaveFile -title $what_for -parent .]
if {$a_file != ""} {
set a_stream [open $a_file w ]
fconfigure $a_stream -translation binary
puts -nonewline $a_stream $the_data
close $a_stream
}
}
# Create the menu for File operations
proc create_menu_file {} {
set fmenu .menus.file.menu
$fmenu add command -label "New" \
-font ClientFont -command {
.io.inp.input_area delete 1.0 end
}
# Replace the contents of the input text
# widget by the data from the open file.
# <FIXME>YAGNI, but is there any reason
# to allow inserting a file, rather than
# replacing the text with file contents?
# </FIXME>
$fmenu add command -label "Open" \
-font ClientFont -command {
set the_text [text_file_contents "File to load"]
if {$the_text != ""} {
.io.inp.input_area delete 1.0 end
.io.inp.input_area insert end $the_text
}
}
$fmenu add command -label "Read" \
-font ClientFont -command {
generate_output [text_file_contents "File to read"]
}
# How to make these disabled for now?
$fmenu add command -label "Save Input" \
-font ClientFont -command {
set the_text [get_input_text]
save_text_file $the_text "Save Input"
}
$fmenu add command -label "Save Output" \
-font ClientFont -command {
set the_text [get_output_text]
save_text_file $the_text "Save Output"
}
}
# Create the menu for edit operations
proc create_menu_edit {} {
set emenu .menus.edit.menu
$emenu add command -label "Select All from Input Area" \
-font ClientFont -command {
# This code says copy the selection as well.
# May be wrong for some platforms, but is
# it more useful?
.io.inp.input_area tag add sel 1.0 end
event generate .io.inp.input_area <<copy>>
}
$emenu add command -label "Select All from Output Area" \
-font ClientFont -command {
# This code says copy the selection as well.
# May be wrong for some platforms, but is
# it more useful?
.io.out.output_area tag add sel 1.0 end
event generate .io.out.output_area <<Copy>>
}
$emenu add command -label "Copy from Input Area" \
-font ClientFont -command {
# this appears not to work. FIXME
event generate .io.inp.input_area <<Copy>>
}
$emenu add command -label "Copy from Output Area" \
-font ClientFont -command {
# this appears not to work. FIXME
event generate .io.out.output_area <<copy>>
}
$emenu add command -label "Paste into Input Area" \
-font ClientFont -command {
# this appears not to work. FIXME
event generate .io.inp.input_area <<Paste>>
}
$emenu add command \
-font ClientFont -label "Insert example text into Input Area"\
-command {
}
# Add specific editing commands here later.
# For example, we would like to be able to
# add whole tags to the XML based formats,
# wrap matching tags around selected text.
# Also we need to find out what happens with
# copy cut and paste, given that X Windows
# is different from MS Windows.
# Allow example text to be inserted.
# However, my thinking is that this should not
# overwrite as it is in the Java application,
# because this rubs out edits when switching
# voices, and this can be annoying when
# exploring the system.
}
# Set the server properties, mostly just
# host and port. Maybe later protocol will
# be possible for https connections?
proc create_menu_server {} {
set smenu .menus.server.menu
$smenu add command -label "host" -font ClientFont -command {
create_entry_dialog "MARY TTS server name" "hostname/IP Address" mary_tts_host
}
$smenu add command -label "port" -font ClientFont -command {
create_entry_dialog "MARY TTS server port" "pott number" mary_tts_port
}
}
# setup the fonts for the various areas on the dipslay.
proc setup_font {family size} {
foreach win {.io .controls .entry.dialogue } {
font configure ClientFont -family $family -size $size
}
}
# Create the menu for changing the text size.
proc create_menu_textstyle {} {
set tmenu .menus.textstyle.menu
$tmenu add cascade -label "Courier" -underline 0 -menu \
$tmenu.courier -font ClientFont
$tmenu add cascade -label "Times" -underline 0 -menu \
$tmenu.times -font ClientFont
$tmenu add cascade -label "Helvetica" -underline 0 -menu \
$tmenu.helvetica -font ClientFont
foreach {name family} [list $tmenu.courier Courier \
$tmenu.times Times $tmenu.helvetica Helvetica ] {
set m1 [menu $name]
foreach pts {6 7 8 9 10 12 14 16 18 20 24 28 32 36} {
$m1 add command -label "$pts" -font ClientFont\
-command [list setup_font $family $pts ]
}
}
}
# Create the menu for Help
proc create_menu_help {} {
# This is all pretty much "wet paint"
# Is there enough to merit separate menus?
set hmenu .menus.help.menu
$hmenu add command -label "Introduction" -font ClientFont\
-command {
tk_messageBox -message "This is a basic Tcl/Tk
client for the MARY TTS system. Most of the options
are reached through the menus on the top. Some
facilities are presently lacking.
Most of the interface should be self-explanatory.
In the File menu, Read will read a given file aloud
(or at least take it as input for the present
form of processing), whereas Open will load it
into the input area. Save input and Save output
refer to the contents of the text windows. The
save button next to the play button will save
the output to a file; this is assumed to be a
text file, unless the output is audio, in which
case it is a binary file.
The Edit menu has cut and paste facilities,
but these don't seem to work reliably. The
default key bindings for text areas should
be useable.
You will need to set the input and output types
and the audio format before pressing play.
Code does not yet exist to figure out sensible
defaults for your platform.
This does not have support for the effects, yet.
Contributions from developers welcome." -type ok
}
$hmenu add command -label "About" -command {} -font ClientFont
}
# We need to create menus for the available
# voices and audio formats, etc.
# When we have the data for these menus from
# the server, create them by using the global
# lists of information.
proc create_radio_menu_from_list {what} {
global $what
set plural "${what}s"
upvar 1 $plural var
foreach item $var {
.menus.${what}.menu add radiobutton -label $item -variable $what \
-value $item -font ClientFont
}
}
proc reset_entry_and_var {a_variable} {
upvar #0 $a_variable var
upvar #0 old_$a_variable old_var
set var $old_var
destroy .entry_dialogue
}
# Create the toplevel for choosing a host
# or port, something taken from an entry.
proc create_entry_dialog {a_message a_label a_variable} {
upvar #0 $a_variable var
upvar #0 old_$a_variable old_var
toplevel .entry_dialogue
label .entry_dialogue.the_message -text $a_message \
-font ClientFont
label .entry_dialogue.the_label -text $a_label -font ClientFont
entry .entry_dialogue.the_entry -textvariable $a_variable \
-font ClientFont
button .entry_dialogue.ok -text "OK" -font ClientFont -command {
destroy .entry_dialogue
}
button .entry_dialogue.cancel -text "Cancel" -font ClientFont \
-command "reset_entry_and_var $a_variable"
grid .entry_dialogue.the_message -row 1 -column 1
grid .entry_dialogue.the_label -row 2 -column 1
grid .entry_dialogue.the_entry -row 2 -column 2
grid .entry_dialogue.ok -row 3 -column 1
grid .entry_dialogue.cancel -row 3 -column 2
}
# Add a message to the end of the output
# text widget.
proc add_message {a_message} {
.io.out.output_area configure -state normal
.io.out.output_area insert end $a_message
.io.out.output_area configure -state disabled
}
# Clear the text in the output text widget.
proc clear_output {} {
.io.out.output_area configure -state normal
.io.out.output_area delete 1.0 end
.io.out.output_area configure -state disabled
}
# Sound generation is platform dependent.
# This provides an "abstract" function to
# be overridden by the platform dependent
# code. In this case it alerts the user
# in the output window that nothing is going
# to happen.
proc play {sound} {
add_message \
"play sound not implemented on this platform apparently"
}
# Graphical stuff.
# In order to be able to scale the font, define a font.
font create ClientFont -family [font actual TkDefaultFont -family] \
-size [font actual TkDefaultFont -size]
frame .menus
create_menubuttons
create_menu_file
create_menu_edit
create_menu_server
create_menu_textstyle
create_menu_help
# Fill in the other menus at runtime.
# .io communicates text with the user,
# through an input and output window.
frame .io
frame .io.inp
frame .io.out
# .controls will hold the play button and
# the effects controls.
frame .controls
# Draw the controls in .io
label .io.inp.input_label -text "Input Area" -font ClientFont
text .io.inp.input_area -height 10 -width 40 \
-xscrollcommand ".io.inp.input_x set" \
-yscrollcommand ".io.inp.input_y set" -font ClientFont
scrollbar .io.inp.input_x -orient horizontal \
-command ".io.inp.input_area xview"
scrollbar .io.inp.input_y -orient vertical \
-command ".io.inp.input_area yview"
label .io.out.output_label -text "Output Area" -font ClientFont
text .io.out.output_area -height 10 -width 40 -state disabled \
-xscrollcommand ".io.out.output_x set" \
-yscrollcommand ".io.out.output_y set" -font ClientFont
scrollbar .io.out.output_x -orient horizontal \
-command ".io.out.output_area xview"
scrollbar .io.out.output_y -orient vertical \
-command ".io.out.output_area yview"
grid .io.inp -in .io -row 1 -column 1
grid .io.out -in .io -row 1 -column 2
grid .io.inp.input_label -in .io.inp -row 1 -column 1
grid .io.inp.input_area -in .io.inp -row 2 -column 1
grid .io.inp.input_y -in .io.inp -row 2 -column 2 -sticky ns
grid .io.inp.input_x -in .io.inp -row 3 -column 1 -sticky ew
grid .io.out.output_label -in .io.out -row 1 -column 1
grid .io.out.output_area -in .io.out -row 2 -column 1
grid .io.out.output_y -in .io.out -row 2 -column 2 -sticky ns
grid .io.out.output_x -in .io.out -row 3 -column 1 -sticky ew
button .controls.play -text "play" -font ClientFont -command {
generate_output [get_input_text]
}
grid .controls.play -in .controls -row 1 -column 1
button .controls.save -text "save" -font ClientFont -command {
global outputtype
set input_text [get_input_text]
if { $outputtype eq "AUDIO" } {
save_binary_file [collect_audio_data $input_text ] "Save audio file"
} else {
save_text_file [collect_audio_data $input_text ] "Save output to file"
}
}
grid .controls.save -in .controls -row 1 -column 2
pack .menus .io .controls -in . -side top
# Detect whether this is the main program
# This test was taken from the Tcl Wiki, and
# seems to work OK.
if {[info exists argv0] && [file tail [info script]] eq [file tail $argv0]} {
# Try to find the temporary files directory.
catch { set tmpdir "/tmp" }
catch { set tmpdir $::env(TRASH_FOLDER) }
catch { set tmpdir $::env(TMP) }
catch { set tmpdir $::env(TEMP) }
# <FIXME>This needs better handling of
# possible alternatives</FIXME>
# This is needed for Windows sound only.
# Do the platform dependent things.
if {$tcl_platform(platform) eq "windows"} {
package require twapi
proc play {sound} {
global tmpdir
# Write sound to a temporary file
set sndfile [file join $tmpdir "MARYTTS_sound.[pid].wav" ]
set stream [open $sndfile w]
# Make sure the file is binary:
fconfigure $stream -translation binary
puts -nonewline $stream $sound
close $stream
# Play the file.
::twapi::play_sound $sndfile
# Remove the file.
file delete $sndfile
}
}
# Put other platforms here.
# Setup the globals with reference to the
# server, which is assumed to be working.
# Since we have options to alter this with
# menu items, there probably needs to be
# some way to reload all this. But we need
# to know how to delete the existing menu
# entries to do that.
setup_globals
create_radio_menu_from_list inputtype
create_radio_menu_from_list outputtype
create_radio_menu_from_list voice
create_radio_menu_from_list audioformat
# Note, at the moment voices holds locales,
# gender, and voice type
# At the moment this is just diagnostic:
## add_message [ join $voices "\n" ]
# it tells us we have a basically working
# system and the list of voices has been
# picked up and manipulated correctly.
# So it is commented out now.
}

81
external/marytts-5.1.2/doc/examples/client/texttospeechdemo.html vendored Normal file
View file

@ -0,0 +1,81 @@
<html>
<!--
Copyright 2000-2006 DFKI GmbH.
All Rights Reserved. Use is subject to license terms.
Permission is hereby granted, free of charge, to use and distribute
this software and its documentation without restriction, including
without limitation the rights to use, copy, modify, merge, publish,
distribute, sublicense, and/or sell copies of this work, and to
permit persons to whom this work is furnished to do so, subject to
the following conditions:
1. The code must retain the above copyright notice, this list of
conditions and the following disclaimer.
2. Any modifications must be clearly marked as such.
3. Original authors' names are not deleted.
4. The authors' names are not used to endorse or promote products
derived from this software without specific prior written
permission.
DFKI GMBH AND THE CONTRIBUTORS TO THIS WORK DISCLAIM ALL WARRANTIES WITH
REGARD TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS, IN NO EVENT SHALL DFKI GMBH NOR THE
CONTRIBUTORS BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL
DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF
THIS SOFTWARE.
-->
<head>
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
<link rel=stylesheet type="text/css" href="mary.css">
<script type="text/javascript">
</head>
<body>
<h1 align="center">The MARY Text-to-Speech System: Online Demo</h1>
<div align="center">
<p>Synthesise any text online (max. 5000 characters):
</p>
<form name="InputForm" method="get" action="maryclient.cgi">
<!-- Set the language of your input here: TEXT_DE=German, TEXT_EN=English... -->
<input type="hidden" name="in" value="TEXT"/>
<input type="hidden" name="out" value="AUDIO"/>
<p>Input text:</p>
<textarea name="inputtext" rows=8 cols=50 wrap=virtual>
Welcome to the world of speech synthesis!
</textarea><br>
<table align="center">
<tr>
<td>
Voice:
<select name="voice">
<option value="us1">us1 (American English, female)</option>
<option value="us2">us2 (American English, male)</option>
</select>
</td><td>
Audio format:
<select name="audiotype">
<option value="AU">Sun Audio, 16 bit</option>
<option value="AIFF">AIFF, 16 bit</option>
<option value="WAVE" selected>WAV, 16 bit</option>
<option value="MP3">mp3</option>
</select>
</td>
</tr>
<tr>
<td align="right">
<input type=submit value="Speak">
</td><td align="left">
<input type=reset value="Reset">
</td>
</tr>
</table>
<p>
<input type=radio name="save_to_disk" value="0" checked>Hear directly
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
<input type=radio name="save_to_disk" value="1">Save to disk
</form>
</div>
</body>
</html>