mirror of
https://github.com/fhem/fhem-mirror.git
synced 2025-01-31 06:39:11 +00:00
2015 lines
81 KiB
Perl
2015 lines
81 KiB
Perl
|
||
##############################################
|
||
# $Id$
|
||
#
|
||
# 98_Text2Speech.pm
|
||
#
|
||
# written by Tobias Faust 2013-10-23
|
||
# e-mail: tobias dot faust at gmx dot net
|
||
#
|
||
##############################################
|
||
|
||
##############################################
|
||
# EDITOR=nano
|
||
# visudo
|
||
# ALL ALL = NOPASSWD: /usr/bin/mplayer
|
||
##############################################
|
||
|
||
package main;
|
||
use strict;
|
||
use warnings;
|
||
use Blocking;
|
||
use HttpUtils;
|
||
# use Data::Dumper;
|
||
|
||
use lib ('./FHEM/lib', './lib');
|
||
|
||
sub Text2Speech_OpenDev($);
|
||
sub Text2Speech_CloseDev($);
|
||
|
||
|
||
# SetParamName -> Anzahl Paramter
|
||
my %sets = (
|
||
"tts" => "1",
|
||
"volume" => "1"
|
||
);
|
||
|
||
# path to mplayer
|
||
my $mplayer = 'sudo /usr/bin/mplayer';
|
||
my $mplayerOpts = '-nolirc -noconsolecontrols';
|
||
my $mplayerNoDebug = '-really-quiet';
|
||
my $mplayerAudioOpts = '-ao alsa:device=';
|
||
|
||
my %ttsHost = ("Google" => "translate.google.com",
|
||
"VoiceRSS" => "api.voicerss.org"
|
||
);
|
||
my %ttsLang = ("Google" => "tl=",
|
||
"VoiceRSS" => "hl="
|
||
);
|
||
my %ttsQuery = ("Google" => "q=",
|
||
"VoiceRSS" => "src="
|
||
);
|
||
my %ttsPath = ("Google" => "/translate_tts?",
|
||
"VoiceRSS" => "/?"
|
||
);
|
||
my %ttsAddon = ("Google" => "client=tw-ob&ie=UTF-8",
|
||
"VoiceRSS" => ""
|
||
);
|
||
my %ttsAPIKey = ("Google" => "", # kein APIKey nötig
|
||
"VoiceRSS" => "key=",
|
||
"maryTTS" => ''
|
||
);
|
||
my %ttsUser = ("Google" => "", # kein Username nötig
|
||
"VoiceRSS" => "" # kein Username nötig
|
||
);
|
||
my %ttsSpeed = ("Google" => "",
|
||
"VoiceRSS" => "r="
|
||
);
|
||
my %ttsQuality = ("Google" => "",
|
||
"VoiceRSS" => "f="
|
||
);
|
||
my %ttsMaxChar = ("Google" => 200,
|
||
"VoiceRSS" => 300,
|
||
"SVOX-pico" => 1000,
|
||
"Amazon-Polly" => 3000,
|
||
"maryTTS" => 3000
|
||
);
|
||
my %language = ("Google" => { "Deutsch" => "de",
|
||
"English-US" => "en-us",
|
||
"Schwedisch" => "sv",
|
||
"France" => "fr",
|
||
"Spain" => "es",
|
||
"Italian" => "it",
|
||
"Chinese" => "cn",
|
||
"Dutch" => "nl"
|
||
},
|
||
"VoiceRSS" => { "Deutsch" => "de-de",
|
||
"English-US" => "en-us",
|
||
"Schwedisch" => "sv-se",
|
||
"France" => "fr-fr",
|
||
"Spain" => "es-es",
|
||
"Italian" => "it-it",
|
||
"Chinese" => "zh-cn",
|
||
"Dutch" => "nl-nl"
|
||
},
|
||
"SVOX-pico" => { "Deutsch" => "de-DE",
|
||
"English-US" => "en-US",
|
||
"France" => "fr-FR",
|
||
"Spain" => "es-ES",
|
||
"Italian" => "it-IT"
|
||
},
|
||
"Amazon-Polly"=> {"Deutsch" => "Marlene",
|
||
"English-US" => "Joanna",
|
||
"Schwedisch" => "Astrid",
|
||
"France" => "Celine",
|
||
"Spain" => "Conchita",
|
||
"Italian" => "Carla",
|
||
"Chinese" => "Zhiyu",
|
||
"Dutch" => "Lotte"
|
||
}
|
||
);
|
||
|
||
##########################
|
||
sub Text2Speech_Initialize($)
|
||
{
|
||
my ($hash) = @_;
|
||
$hash->{WriteFn} = "Text2Speech_Write";
|
||
$hash->{ReadyFn} = "Text2Speech_Ready";
|
||
$hash->{DefFn} = "Text2Speech_Define";
|
||
$hash->{SetFn} = "Text2Speech_Set";
|
||
$hash->{UndefFn} = "Text2Speech_Undefine";
|
||
$hash->{RenameFn} = "Text2Speech_Rename";
|
||
$hash->{AttrFn} = "Text2Speech_Attr";
|
||
$hash->{AttrList} = "disable:0,1".
|
||
" TTS_Delimiter".
|
||
" TTS_Ressource:ESpeak,SVOX-pico,Amazon-Polly,maryTTS,". join(",", sort keys %ttsHost).
|
||
" TTS_APIKey".
|
||
" TTS_User".
|
||
" TTS_Quality:".
|
||
"48khz_16bit_stereo,".
|
||
"48khz_16bit_mono,".
|
||
"48khz_8bit_stereo,".
|
||
"48khz_8bit_mono".
|
||
"44khz_16bit_stereo,".
|
||
"44khz_16bit_mono,".
|
||
"44khz_8bit_stereo,".
|
||
"44khz_8bit_mono".
|
||
"32khz_16bit_stereo,".
|
||
"32khz_16bit_mono,".
|
||
"32khz_8bit_stereo,".
|
||
"32khz_8bit_mono".
|
||
"24khz_16bit_stereo,".
|
||
"24khz_16bit_mono,".
|
||
"24khz_8bit_stereo,".
|
||
"24khz_8bit_mono".
|
||
"22khz_16bit_stereo,".
|
||
"22khz_16bit_mono,".
|
||
"22khz_8bit_stereo,".
|
||
"22khz_8bit_mono".
|
||
"16khz_16bit_stereo,".
|
||
"16khz_16bit_mono,".
|
||
"16khz_8bit_stereo,".
|
||
"16khz_8bit_mono".
|
||
"8khz_16bit_stereo,".
|
||
"8khz_16bit_mono,".
|
||
"8khz_8bit_stereo,".
|
||
"8khz_8bit_mono".
|
||
" TTS_Speed:-10,-9,-8,-7,-6,-5,-4,-3,-2,-1,0,1,2,3,4,5,6,7,8,9,10".
|
||
" TTS_TimeOut".
|
||
" TTS_CacheFileDir".
|
||
" TTS_UseMP3Wrap:0,1".
|
||
" TTS_MplayerCall".
|
||
" TTS_SentenceAppendix".
|
||
" TTS_FileMapping".
|
||
" TTS_FileTemplateDir".
|
||
" TTS_VolumeAdjust".
|
||
" TTS_noStatisticsLog:1,0".
|
||
" TTS_Language:".join(",", sort keys %{$language{"Google"}}).
|
||
" TTS_Language_Custom".
|
||
" TTS_SpeakAsFastAsPossible:1,0".
|
||
" TTS_OutputFile".
|
||
" TTS_AWS_HomeDir".
|
||
" TTS_RemotePlayerCall".
|
||
" ".$readingFnAttributes;
|
||
}
|
||
|
||
|
||
##########################
|
||
# Define <tts> Text2Speech <alsa-device>
|
||
# Define <tts> Text2Speech host[:port][:SSL] [portpassword]
|
||
##########################
|
||
sub Text2Speech_Define($$)
|
||
{
|
||
my ($hash, $def) = @_;
|
||
my @a = split("[ \t]+", $def);
|
||
|
||
#$a[0]: Name
|
||
#$a[1]: Type/Alias -> Text2Speech
|
||
#$a[2]: definition
|
||
#$a[3]: optional: portpasswd
|
||
if(int(@a) < 3) {
|
||
my $msg = "wrong syntax: define <name> Text2Speech <alsa-device>\n".
|
||
"see at /etc/asound.conf\n".
|
||
"or remote syntax: define <name> Text2Speech host[:port][:SSL] [portpassword]";
|
||
Log3 $hash, 2, $msg;
|
||
return $msg;
|
||
}
|
||
|
||
my $dev = $a[2];
|
||
if($dev =~ m/^([0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}).*/ ) {
|
||
# Ein RemoteDevice ist angegeben
|
||
# zb: 192.168.10.24:7272:SSL mypasswd
|
||
|
||
if($dev =~ m/^(.*):SSL$/) {
|
||
$dev = $1;
|
||
$hash->{SSL} = 1;
|
||
}
|
||
if($dev !~ m/^.+:[0-9]+$/) { # host:port
|
||
$dev = "$dev:7072";
|
||
}
|
||
$hash->{Host} = $dev;
|
||
$hash->{portpassword} = $a[3] if(@a == 4);
|
||
|
||
$hash->{MODE} = "REMOTE";
|
||
} elsif (lc($dev) eq "none") {
|
||
# Ein DummyDevice, Serverdevice. Nur Generierung der mp3 TTS Dateien
|
||
$hash->{MODE} = "SERVER";
|
||
undef $hash->{ALSADEVICE};
|
||
} else {
|
||
# Ein Alsadevice ist angegeben
|
||
# pruefen, ob Alsa-Device in /etc/asound.conf definiert ist
|
||
$hash->{MODE} = "DIRECT";
|
||
$hash->{ALSADEVICE} = $a[2];
|
||
}
|
||
|
||
BlockingKill($hash->{helper}{RUNNING_PID}) if(defined($hash->{helper}{RUNNING_PID}));
|
||
delete($hash->{helper}{RUNNING_PID});
|
||
|
||
$hash->{STATE} = "Initialized";
|
||
|
||
my $ret = Text2Speech_loadmodules($hash, "");
|
||
if ($ret) {
|
||
Log3 $hash->{NAME}, 3, $ret;
|
||
}
|
||
Text2Speech_AddExtension( $hash->{NAME}, \&Text2Speech_getLastMp3, "$hash->{TYPE}/$hash->{NAME}/last.mp3" );
|
||
return undef;
|
||
}
|
||
|
||
##########################
|
||
# Überprüfung und Einladen der notwendigen Module
|
||
##########################
|
||
sub Text2Speech_loadmodules($$) {
|
||
my ($hash, $TTS_Ressource) = @_;
|
||
eval {
|
||
require IO::File;
|
||
IO::File->import;
|
||
1;
|
||
} or return "IO::File Module not installed. Please install.";
|
||
|
||
eval {
|
||
require Digest::MD5;
|
||
Digest::MD5->import;
|
||
1;
|
||
} or return "Digest::MD5 Module not installed. Please install.";
|
||
|
||
eval {
|
||
require URI::Escape;
|
||
URI::Escape->import;
|
||
1;
|
||
} or return "URI::Escape Module not installed. Please install.";
|
||
|
||
eval {
|
||
require Text::Iconv;
|
||
Text::Iconv->import;
|
||
1;
|
||
} or return "Text::Iconv Module not installed. Please install.";
|
||
|
||
eval {
|
||
require Encode::Guess;
|
||
Encode::Guess->import;
|
||
1;
|
||
} or return "Encode::Guess Module not installed. Please install.";
|
||
|
||
eval {
|
||
require MP3::Info;
|
||
MP3::Info->import;
|
||
1;
|
||
} or return "MP3::Info Module not installed. Please install.";
|
||
|
||
if ($TTS_Ressource eq "Amazon-Polly") {
|
||
# Module werden nur benötigt mit der Polly Engine
|
||
eval {
|
||
require Paws::Polly;
|
||
Paws::Polly->import;
|
||
1;
|
||
} or return "Paws Module not installed. Please install via 'sudo cpan Paws'.";
|
||
|
||
eval {
|
||
require File::HomeDir;
|
||
File::HomeDir->import;
|
||
1;
|
||
} or return "File::HomeDir Module not installed. Please install";
|
||
}
|
||
|
||
return undef;
|
||
}
|
||
|
||
#####################################
|
||
sub Text2Speech_Undefine($$)
|
||
{
|
||
my ($hash, $arg) = @_;
|
||
|
||
RemoveInternalTimer($hash);
|
||
BlockingKill($hash->{helper}{RUNNING_PID}) if(defined($hash->{helper}{RUNNING_PID}));
|
||
Text2Speech_RemoveExtension( "$hash->{TYPE}/$hash->{NAME}/last.mp3" );
|
||
Text2Speech_CloseDev($hash);
|
||
|
||
return undef;
|
||
}
|
||
|
||
sub Text2Speech_Rename(@) {
|
||
my ( $newname, $oldname ) = @_;
|
||
my $hash = $defs{$newname};
|
||
my $type = $hash->{TYPE};
|
||
Text2Speech_RemoveExtension( "$type/$oldname/last.mp3" );
|
||
Text2Speech_AddExtension( $newname, \&Text2Speech_getLastMp3, "$type/$newname/last.mp3" );
|
||
}
|
||
|
||
sub Text2Speech_Attr(@) {
|
||
my @a = @_;
|
||
my $do = 0;
|
||
my $hash = $defs{$a[1]};
|
||
my $value = $a[3];
|
||
|
||
my $TTS_FileTemplateDir = AttrVal($hash->{NAME}, "TTS_FileTemplateDir", "templates");
|
||
my $TTS_CacheFileDir = AttrVal($hash->{NAME}, "TTS_CacheFileDir", "cache");
|
||
my $TTS_FileMapping = AttrVal($hash->{NAME}, "TTS_FileMapping", ""); # zb, silence:silence.mp3 ring:myringtone.mp3;
|
||
my $TTS_AWS_HomeDir = AttrVal($hash->{NAME}, "TTS_AWS_HomeDir", "/home/fhem");
|
||
|
||
if($a[2] eq "TTS_Delimiter" && $a[0] ne "del") {
|
||
return "wrong Delimiter syntax: [+-]a[lfn]. Please see CommandRef for Notation. \n".
|
||
" Example 1: +an~\n".
|
||
" Example 2: +al." if($value !~ m/^([+-]a[lfn]){0,1}(.){1}$/i);
|
||
return "This Attribute is only available in direct or server mode" if($hash->{MODE} !~ m/(DIRECT|SERVER)/ );
|
||
|
||
} elsif ($a[0] eq "set" && $a[2] eq "TTS_Ressource" && $value eq "Amazon-Polly") {
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}. ": Wechsele auf Amazon Polly, Lade Librarys nach.";
|
||
my $ret = Text2Speech_loadmodules($hash, $a[2]);
|
||
if ($ret) {return $ret;} # breche ab wenn Module fehlen
|
||
|
||
if (! -e $TTS_AWS_HomeDir."/.aws/credentials"){
|
||
return "No AWS credentials in FHEM Homedir found, please check ".$TTS_AWS_HomeDir."/.aws/credentials \n please refer https://metacpan.org/pod/Paws#AUTHENTICATION \n\n Please check Attribute 'TTS_AWS_HomeDir' too";
|
||
}
|
||
|
||
} elsif ($a[0] eq "set" && $a[2] eq "TTS_Ressource") {
|
||
return "This Attribute is only available in direct or server mode" if($hash->{MODE} !~ m/(DIRECT|SERVER)/ );
|
||
|
||
} elsif ($a[0] eq "set" && $a[2] eq "TTS_CacheFileDir") {
|
||
return "This Attribute is only available in direct or server mode" if($hash->{MODE} !~ m/(DIRECT|SERVER)/ );
|
||
|
||
} elsif ($a[0] eq "set" && $a[2] eq "TTS_SpeakAsFastAsPossible") {
|
||
return "This Attribute is only available in direct or server mode" if($hash->{MODE} !~ m/(DIRECT|SERVER)/ );
|
||
|
||
} elsif ($a[0] eq "set" && $a[2] eq "TTS_UseMP3Wrap") {
|
||
return "This Attribute is only available in direct or server mode" if($hash->{MODE} !~ m/(DIRECT|SERVER)/ );
|
||
return "Attribute TTS_UseMP3Wrap is required by Attribute TTS_SentenceAppendix! Please delete it first."
|
||
if(($a[0] eq "del") && (AttrVal($hash->{NAME}, "TTS_SentenceAppendix", undef)));
|
||
|
||
} elsif ($a[0] eq "set" && $a[2] eq "TTS_SentenceAppendix") {
|
||
return "This Attribute is only available in direct or server mode" if($hash->{MODE} !~ m/(DIRECT|SERVER)/ );
|
||
return "Attribute TTS_UseMP3Wrap is required!" unless(AttrVal($hash->{NAME}, "TTS_UseMP3Wrap", undef));
|
||
|
||
my $file = $TTS_CacheFileDir ."/". $value;
|
||
return "File <".$file."> does not exists in CacheFileDir" if(! -e $file);
|
||
|
||
} elsif ($a[0] eq "set" && $a[2] eq "TTS_FileTemplateDir") {
|
||
# Verzeichnis beginnt mit /, dann absoluter Pfad, sonst Unterpfad von $TTS_CacheFileDir
|
||
my $newDir;
|
||
if($value =~ m/^\/.*/) { $newDir = $value; } else { $newDir = $TTS_CacheFileDir ."/". $value;}
|
||
unless(-e ($newDir) or mkdir ($newDir)) {
|
||
#Verzeichnis anlegen gescheitert
|
||
return "Could not create directory: <$value>";
|
||
}
|
||
|
||
} elsif ($a[0] eq "set" && $a[2] eq "TTS_TimeOut") {
|
||
return "Only Numbers allowed" if ($value !~ m/[0-9]+/);
|
||
|
||
} elsif ($a[0] eq "set" && $a[2] eq "TTS_AWS_HomeDir") {
|
||
return "Your HomeDir cannot be found." if (! -e $value)
|
||
|
||
} elsif ($a[0] eq "set" && $a[2] eq "TTS_FileMapping") {
|
||
#Bsp: silence:silence.mp3 pling:mypling,mp3
|
||
#ueberpruefen, ob mp3 Template existiert
|
||
my @FileTpl = split(" ", $TTS_FileMapping);
|
||
my $newDir;
|
||
for(my $j=0; $j<(@FileTpl); $j++) {
|
||
my @FileTplPc = split(/:/, $FileTpl[$j]);
|
||
if($TTS_FileTemplateDir =~ m/^\/.*/) { $newDir = $TTS_FileTemplateDir; } else { $newDir = $TTS_CacheFileDir ."/". $TTS_FileTemplateDir;}
|
||
return "file does not exist: <".$newDir ."/". $FileTplPc[1] .">"
|
||
unless (-e $newDir ."/". $FileTplPc[1]);
|
||
}
|
||
} elsif ($a[0] eq "set" && $a[2] eq "TTS_RemotePlayerCall") {
|
||
if( $init_done ) {
|
||
eval $value ;
|
||
return "Text2Speach_Attr evaluating TTS_RemotePlayerCall error: $@" if ( $@ );
|
||
|
||
}
|
||
}
|
||
|
||
if($a[0] eq "set" && $a[2] eq "disable") {
|
||
$do = (!defined($a[3]) || $a[3]) ? 1 : 2;
|
||
}
|
||
$do = 2 if($a[0] eq "del" && (!$a[2] || $a[2] eq "disable"));
|
||
return if(!$do);
|
||
|
||
$hash->{STATE} = ($do == 1 ? "disabled" : "Initialized");
|
||
|
||
return undef;
|
||
}
|
||
|
||
#####################################
|
||
sub Text2Speech_Ready($)
|
||
{
|
||
my ($hash) = @_;
|
||
return Text2speech_OpenDev($hash, 1);
|
||
}
|
||
|
||
########################
|
||
sub Text2Speech_OpenDev($) {
|
||
my ($hash) = @_;
|
||
my $dev = $hash->{Host};
|
||
my $name = $hash->{NAME};
|
||
|
||
Log3 $name, 4, "Text2Speech opening $name at $dev";
|
||
|
||
my $conn;
|
||
if($hash->{SSL}) {
|
||
eval "use IO::Socket::SSL";
|
||
Log3 $name, 1, $@ if($@);
|
||
$conn = IO::Socket::SSL->new(PeerAddr => "$dev", MultiHomed => 1) if(!$@);
|
||
} else {
|
||
$conn = IO::Socket::INET->new(PeerAddr => $dev, MultiHomed => 1);
|
||
}
|
||
|
||
if(!$conn) {
|
||
Log3($name, 3, $hash->{NAME}.": Can't connect to $dev: $!");
|
||
$hash->{STATE} = "disconnected";
|
||
return "";
|
||
} else {
|
||
$hash->{STATE} = "Initialized";
|
||
}
|
||
|
||
$hash->{TCPDev} = $conn;
|
||
$hash->{FD} = $conn->fileno();
|
||
|
||
Log3 $name, 4, "Text2Speech device opened ($name)";
|
||
|
||
syswrite($hash->{TCPDev}, $hash->{portpassword} . "\n")
|
||
if($hash->{portpassword});
|
||
|
||
return undef;
|
||
}
|
||
|
||
########################
|
||
sub Text2Speech_CloseDev($) {
|
||
my ($hash) = @_;
|
||
my $name = $hash->{NAME};
|
||
my $dev = $hash->{Host};
|
||
return if(!$dev);
|
||
|
||
if($hash->{TCPDev}) {
|
||
$hash->{TCPDev}->close();
|
||
Log3 $hash, 4, "Text2speech Device closed ($name)";
|
||
}
|
||
|
||
delete($hash->{TCPDev});
|
||
delete($hash->{FD});
|
||
}
|
||
|
||
########################
|
||
sub Text2Speech_Write($$) {
|
||
my ($hash,$msg) = @_;
|
||
my $name = $hash->{NAME};
|
||
my $dev = $hash->{Host};
|
||
|
||
#my $call = "set tts tts Das ist ein Test.";
|
||
my $call = "set $name $msg";
|
||
|
||
#Prüfen ob PRESENCE vorhanden und present
|
||
my $isPresent = 0;
|
||
my $hasPRESENCE = 0;
|
||
my $devname="";
|
||
if ($hash->{MODE} eq "REMOTE") {
|
||
foreach $devname (devspec2array("TYPE=PRESENCE")) {
|
||
if (defined $defs{$devname}->{ADDRESS} && $dev) {
|
||
if ($dev =~ $defs{$devname}->{ADDRESS}) {
|
||
$hasPRESENCE = 1;
|
||
$isPresent = 1 if (ReadingsVal($devname,"presence","unknown") eq "present");
|
||
last;
|
||
}
|
||
}
|
||
}
|
||
}
|
||
if ($hasPRESENCE) {
|
||
Log3 $hash, 4, $name.": found PRESENCE Device $devname for host: $dev, it\'s state is: ".($isPresent ? "present" : "absent");
|
||
Text2Speech_OpenDev($hash) if(!$hash->{TCPDev} && $isPresent);
|
||
#lets try again
|
||
Text2Speech_OpenDev($hash) if(!$hash->{TCPDev} && $isPresent);
|
||
} else {
|
||
Log3 $hash, 4, $name.": no proper PRESENCE Device for host: $dev";
|
||
Text2Speech_OpenDev($hash) if(!$hash->{TCPDev});
|
||
#lets try again
|
||
Text2Speech_OpenDev($hash) if(!$hash->{TCPDev});
|
||
}
|
||
|
||
if($hash->{TCPDev}) {
|
||
Log3 $hash, 4, $name.": Write remote message to $dev: $call";
|
||
Log3 $hash, 3, $name.": Could not write remote message ($call) at " .$hash->{Host} if(!defined(syswrite($hash->{TCPDev}, "$call\n")));
|
||
Text2Speech_CloseDev($hash);
|
||
}
|
||
|
||
}
|
||
|
||
|
||
###########################################################################
|
||
|
||
sub Text2Speech_Set($@)
|
||
{
|
||
my ($hash, @a) = @_;
|
||
my $me = $hash->{NAME};
|
||
my $TTS_APIKey = AttrVal($hash->{NAME}, "TTS_APIKey", undef);
|
||
my $TTS_User = AttrVal($hash->{NAME}, "TTS_User", undef);
|
||
my $TTS_Ressource = AttrVal($hash->{NAME}, "TTS_Ressource", "Google");
|
||
my $TTS_TimeOut = AttrVal($hash->{NAME}, "TTS_TimeOut", 60);
|
||
|
||
|
||
return "no set argument specified" if(int(@a) < 2);
|
||
|
||
return "No APIKey specified" if (!defined($TTS_APIKey) && ($ttsAPIKey{$TTS_Ressource} || length($ttsAPIKey{$TTS_Ressource})>0));
|
||
return "No Username for TTS Access specified" if ( $TTS_Ressource ne 'maryTTS' && !defined($TTS_User) && ($ttsUser{$TTS_Ressource} || length($ttsUser{$TTS_Ressource})>0));
|
||
|
||
my $ret = Text2Speech_loadmodules($hash, $TTS_Ressource);
|
||
if ($ret) {
|
||
# breche ab wenn Module fehlen
|
||
Log3 $me, 3, $ret;
|
||
return $ret;
|
||
}
|
||
|
||
my $cmd = shift(@a); # Dummy
|
||
$cmd = shift(@a); # DevName
|
||
|
||
if(!defined($sets{$cmd})) {
|
||
my $r = "Unknown argument $cmd, choose one of ".join(" ",sort keys %sets);
|
||
return $r;
|
||
}
|
||
|
||
if($cmd ne "tts") {
|
||
return "$cmd needs $sets{$cmd} parameter(s)" if(@a-$sets{$cmd} != 0);
|
||
} else {
|
||
return "$cmd needs text parameter" if(@a-$sets{$cmd} < 0);
|
||
}
|
||
|
||
# Abbruch falls Disabled
|
||
return "no set cmd on a disabled device !" if(IsDisabled($me));
|
||
|
||
if($cmd eq "tts") {
|
||
|
||
if($hash->{MODE} eq "DIRECT" || $hash->{MODE} eq "SERVER") {
|
||
$hash->{VOLUME} = ReadingsNum($me, "volume", 100);
|
||
readingsSingleUpdate($hash, "playing", "1", 1);
|
||
Text2Speech_PrepareSpeech($hash, join(" ", @a));
|
||
$hash->{helper}{RUNNING_PID} = BlockingCall("Text2Speech_DoIt", $hash, "Text2Speech_Done", $TTS_TimeOut, "Text2Speech_AbortFn", $hash) unless(exists($hash->{helper}{RUNNING_PID}));
|
||
} elsif ($hash->{MODE} eq "REMOTE") {
|
||
Text2Speech_Write($hash, "tts " . join(" ", @a));
|
||
} else {return undef;}
|
||
} elsif($cmd eq "volume") {
|
||
my $vol = join(" ", @a);
|
||
return "volume level expects 0..100 percent" if($vol !~ m/^([0-9]{1,3})$/ or $vol > 100);
|
||
|
||
if($hash->{MODE} eq "DIRECT") {
|
||
$hash->{VOLUME} = $vol if($vol <= 100);
|
||
delete($hash->{VOLUME}) if($vol > 100);
|
||
} elsif ($hash->{MODE} eq "REMOTE") {
|
||
Text2Speech_Write($hash, "volume $vol");
|
||
} else {return undef;}
|
||
|
||
readingsSingleUpdate($hash, "volume", (($vol>100)?0:$vol), 1);
|
||
}
|
||
|
||
return undef;
|
||
}
|
||
|
||
#####################################
|
||
# Bereitet den gesamten String vor.
|
||
# Bei Nutzung Google wird dieser in ein Array
|
||
# zerlegt mit jeweils einer maximalen
|
||
# Stringlänge von 100Chars
|
||
#
|
||
# param1: $hash
|
||
# param2: string to speech
|
||
#
|
||
#####################################
|
||
###################################
|
||
# Angabe des Delimiters: zb.: +af~
|
||
# + -> erzwinge das Trennen, auch wenn Textbaustein < 100Zeichen
|
||
# - -> Trenne nur wenn Textbaustein > 100Zeichen
|
||
# af -> add first -> füge den Delimiter am Satzanfang wieder hinzu
|
||
# al -> add last -> füge den Delimiter am Satzende wieder hinzu
|
||
# an -> add nothing -> Delimiter nicht wieder hinzufügen
|
||
# ~ -> der Delimiter
|
||
###################################
|
||
sub Text2Speech_PrepareSpeech($$) {
|
||
my ($hash, $t) = @_;
|
||
my $me = $hash->{NAME};
|
||
|
||
my $TTS_Ressource = AttrVal($hash->{NAME}, "TTS_Ressource", "Google");
|
||
my $TTS_Delimiter = AttrVal($hash->{NAME}, "TTS_Delimiter", undef);
|
||
my $TTS_FileTpl = AttrVal($hash->{NAME}, "TTS_FileMapping", ""); # zb, silence:silence.mp3 ring:myringtone.mp3; im Text: mein Klingelton :ring: ist laut.
|
||
my $TTS_FileTemplateDir = AttrVal($hash->{NAME}, "TTS_FileTemplateDir", "templates");
|
||
|
||
my $TTS_ForceSplit = 0;
|
||
my $TTS_AddDelimiter;
|
||
|
||
# Cleanup string
|
||
$hash->{helper}{TTS_PlayerOptions} = "";
|
||
while ($t =~ s/^ //isg) {};
|
||
$t =~ s/^'(.*)'$/$1/;
|
||
$t =~ s/^"(.*)"$/$1/;
|
||
|
||
# Check text for command string
|
||
if ($t =~ /^\[(.*?)\](.*?)$/) {
|
||
($hash->{helper}{TTS_PlayerOptions}, $t) = ($1, $2);
|
||
while ($t =~ s/^ //isg) {};
|
||
}
|
||
|
||
if($TTS_Delimiter && $TTS_Delimiter =~ m/^[+-]a[lfn]/i) {
|
||
$TTS_ForceSplit = 1 if(substr($TTS_Delimiter,0,1) eq "+");
|
||
$TTS_ForceSplit = 0 if(substr($TTS_Delimiter,0,1) eq "-");
|
||
|
||
$TTS_AddDelimiter = substr($TTS_Delimiter,1,2); # af, al oder an
|
||
|
||
$TTS_Delimiter = substr($TTS_Delimiter,3);
|
||
|
||
} elsif (!$TTS_Delimiter) { # Default wenn Attr nicht gesetzt
|
||
$TTS_Delimiter = "(?<=[\\.!?])\\s*";
|
||
$TTS_ForceSplit = 0;
|
||
$TTS_AddDelimiter = "";
|
||
}
|
||
|
||
#-- we may have problems with umlaut characters
|
||
# ersetze Sonderzeichen die Google nicht auflösen kann
|
||
my $converter;
|
||
|
||
# wandle per standard alles nach UTF8
|
||
# check only ascii, utf8 and UTF-(16|32) with BOM, if not enough use function set_suspects
|
||
# Encode::Guess->set_suspects(qw/euc-jp shiftjis 7bit-jis/); # for japanese codepages
|
||
my $enc = guess_encoding($t);
|
||
if ($enc->name ne "utf8") {
|
||
Log3 $hash, 4, "$me: ermittelte CodePage: " .$enc->name. " , konvertiere nach UTF-8";
|
||
$converter = Text::Iconv->new($enc->name, "utf-8");
|
||
$t = $converter->convert($t);
|
||
}
|
||
|
||
#if($TTS_Ressource eq "Google") {
|
||
# Google benötigt UTF-8
|
||
# $t =~ s/ä/ae/g;
|
||
# $t =~ s/ö/oe/g;
|
||
# $t =~ s/ü/ue/g;
|
||
# $t =~ s/Ä/Ae/g;
|
||
# $t =~ s/Ö/Oe/g;
|
||
# $t =~ s/Ü/Ue/g;
|
||
# $t =~ s/ß/ss/g;
|
||
#}
|
||
|
||
if ($TTS_Ressource eq "Amazon-Polly") {
|
||
# Amazon benötigt ISO-8859-1 bei Nutzung Region eu-central-1
|
||
$converter = Text::Iconv->new("utf-8", "iso-8859-1");
|
||
$t = $converter->convert($t);
|
||
}
|
||
|
||
my @text;
|
||
push(@text, $t);
|
||
|
||
# hole alle Filetemplates
|
||
my @FileTpl = split(" ", $TTS_FileTpl);
|
||
my @FileTplPc;
|
||
|
||
# bei Angabe direkter MP3-Files wird hier ein temporäres Template vergeben
|
||
for(my $i=0; $i<(@text); $i++) {
|
||
@FileTplPc = ($text[$i] =~ /:([\w-]+?\.(?:mp3|ogg|wav)):/g);
|
||
for(my $j=0; $j<(@FileTplPc); $j++) {
|
||
my $tpl = "FileTpl_#".$i."_".$j; #eindeutige Templatedefinition schaffen
|
||
Log3 $hash, 4, "$me: Angabe einer direkten MP3-Datei gefunden: $FileTplPc[$j] => $tpl";
|
||
push(@FileTpl, $tpl.":".$FileTplPc[$j]); #zb: FileTpl_123645875_#0:/ring.mp3
|
||
$text[$i] =~ s/$FileTplPc[$j]/$tpl/g; # Ersetze die DateiDefinition gegen ein Template
|
||
}
|
||
}
|
||
|
||
#iteriere durch die Sprachbausteine und splitte den Text bei den Filetemplates auf
|
||
for(my $i=0; $i<(@text); $i++) {
|
||
my $cutter = '#!#'; #eindeutigen Cutter als Delimiter bei den Filetemplates vergeben
|
||
@FileTplPc = ($text[$i] =~ /:([^:]+):/g);
|
||
for(my $j=0; $j<(@FileTplPc); $j++) {
|
||
$text[$i] =~ s/:$FileTplPc[$j]:/$cutter$FileTplPc[$j]$cutter/g;
|
||
}
|
||
@text = Text2Speech_SplitString(\@text, 0, $cutter, 1, "");
|
||
}
|
||
|
||
Log3 $hash, 4, "$me: MaxChar = $ttsMaxChar{$TTS_Ressource}, Delimiter = $TTS_Delimiter, ForceSplit = $TTS_ForceSplit, AddDelimiter = $TTS_AddDelimiter";
|
||
|
||
@text = Text2Speech_SplitString(\@text, $ttsMaxChar{$TTS_Ressource}, $TTS_Delimiter, $TTS_ForceSplit, $TTS_AddDelimiter);
|
||
@text = Text2Speech_SplitString(\@text, $ttsMaxChar{$TTS_Ressource}, "(?<=[.!?])\\s*", 0, "");
|
||
@text = Text2Speech_SplitString(\@text, $ttsMaxChar{$TTS_Ressource}, ",", 0, "al");
|
||
@text = Text2Speech_SplitString(\@text, $ttsMaxChar{$TTS_Ressource}, ";", 0, "al");
|
||
@text = Text2Speech_SplitString(\@text, $ttsMaxChar{$TTS_Ressource}, "und", 0, "af");
|
||
@text = Text2Speech_SplitString(\@text, $ttsMaxChar{$TTS_Ressource}, ":", 0, "al");
|
||
@text = Text2Speech_SplitString(\@text, $ttsMaxChar{$TTS_Ressource}, "\\bund\\b", 0, "af");
|
||
@text = Text2Speech_SplitString(\@text, $ttsMaxChar{$TTS_Ressource}, " ", 0, "");
|
||
|
||
Log3 $hash, 4, "$me: Auflistung der Textbausteine nach Aufbereitung:";
|
||
for(my $i=0; $i<(@text); $i++) {
|
||
# entferne führende und abschließende Leerzeichen aus jedem Textbaustein
|
||
$text[$i] =~ s/^\s+|\s+$//g;
|
||
for(my $j=0; $j<(@FileTpl); $j++) {
|
||
# ersetze die FileTemplates mit den echten MP3-Files
|
||
@FileTplPc = split(/:/, $FileTpl[$j]);
|
||
$text[$i] = $TTS_FileTemplateDir ."/". $FileTplPc[1] if($text[$i] eq $FileTplPc[0]);
|
||
}
|
||
Log3 $hash, 4, "$me: $i => ".$text[$i];
|
||
}
|
||
|
||
push( @{$hash->{helper}{Text2Speech}}, @text );
|
||
}
|
||
|
||
#####################################
|
||
# param1: array : Text 2 Speech
|
||
# param2: string: MaxChar
|
||
# param3: string: Delimiter
|
||
# param4: int : 1 -> es wird am Delimiter gesplittet
|
||
# 0 -> es wird nur gesplittet, wenn Stringlänge länger als MaxChar
|
||
# param5: string: Add Delimiter to String? [al|af|<empty>] (AddLast/AddFirst)
|
||
#
|
||
# Splittet die Texte aus $hash->{helper}->{Text2Speech} anhand des
|
||
# Delimiters, wenn die Stringlänge MaxChars übersteigt.
|
||
# Ist "AddDelimiter" angegeben, so wird der Delimiter an den
|
||
# String wieder angefügt
|
||
#####################################
|
||
sub Text2Speech_SplitString($$$$$){
|
||
my @text = @{shift()};
|
||
my $MaxChar = shift;
|
||
my $Delimiter = shift;
|
||
my $ForceSplit = shift;
|
||
my $AddDelimiter = shift;
|
||
my @newText;
|
||
|
||
for(my $i=0; $i<(@text); $i++) {
|
||
if((length($text[$i]) <= $MaxChar) && (!$ForceSplit)) { #Google kann nur 100zeichen
|
||
push(@newText, $text[$i]);
|
||
next;
|
||
}
|
||
|
||
my @b;
|
||
if($Delimiter =~/^ $/) {
|
||
@b = split(' ', $text[$i]);
|
||
}
|
||
else {
|
||
@b = split(/$Delimiter/, $text[$i]);
|
||
}
|
||
if((@b)>1) {
|
||
# setze zu kleine Textbausteine wieder zusammen bis MaxChar erreicht ist
|
||
if(length($Delimiter)==1) {
|
||
for(my $k=0; $k<(@b); ) {
|
||
if($k+1<(@b) && length($b[$k])+length($b[$k+1]) <= $MaxChar) {
|
||
$b[$k] = join($Delimiter, $b[$k], $b[$k+1]);
|
||
splice(@b, $k+1, 1);
|
||
}
|
||
else {
|
||
$k++;
|
||
}
|
||
}
|
||
}
|
||
for(my $j=0; $j<(@b); $j++) {
|
||
(my $boundaryDelimiter = $Delimiter) =~ s/^\\b(.+)\\b$/$1/g;
|
||
$b[$j] = $b[$j] . $boundaryDelimiter if($AddDelimiter eq "al"); # Am Satzende wieder hinzufügen.
|
||
$b[$j+1] = $boundaryDelimiter . $b[$j+1] if(($AddDelimiter eq "af") && ($b[$j+1])); # Am Satzanfang des nächsten Satzes wieder hinzufügen.
|
||
push(@newText, $b[$j]);
|
||
}
|
||
}
|
||
elsif((@b)==1) {
|
||
push(@newText, $text[$i]);
|
||
}
|
||
}
|
||
return @newText;
|
||
}
|
||
|
||
#####################################
|
||
# param1: hash : Hash
|
||
# param2: string: Datei
|
||
#
|
||
# Erstellt den Commandstring für den Systemaufruf
|
||
#####################################
|
||
sub Text2Speech_BuildMplayerCmdString($$) {
|
||
my ($hash, $file) = @_;
|
||
my $cmd;
|
||
|
||
my $TTS_MplayerCall = AttrVal($hash->{NAME}, "TTS_MplayerCall", $mplayer);
|
||
my $TTS_VolumeAdjust = AttrVal($hash->{NAME}, "TTS_VolumeAdjust", 110);
|
||
my $verbose = AttrVal($hash->{NAME}, "verbose", 3);
|
||
|
||
if($hash->{VOLUME}) { # per: set <name> volume <..>
|
||
$mplayerOpts .= " -softvol -softvol-max ". $TTS_VolumeAdjust ." -volume " . $hash->{VOLUME};
|
||
}
|
||
|
||
my $AlsaDevice = $hash->{ALSADEVICE};
|
||
if($AlsaDevice eq "default") {
|
||
$AlsaDevice = "";
|
||
$mplayerAudioOpts = "";
|
||
}
|
||
|
||
my $NoDebug = $mplayerNoDebug;
|
||
$NoDebug = "" if($verbose >= 5);
|
||
# anstatt mplayer wird ein anderer Player verwendet
|
||
if ($TTS_MplayerCall !~ m/mplayer/) {
|
||
$TTS_MplayerCall =~ s/{device}/$AlsaDevice/g;
|
||
$TTS_MplayerCall =~ s/{volume}/$hash->{VOLUME}/g;
|
||
$TTS_MplayerCall =~ s/{volumeadjust}/$TTS_VolumeAdjust/g;
|
||
$TTS_MplayerCall =~ s/{file}/$file/g;
|
||
$TTS_MplayerCall =~ s/{options}/$hash->{helper}{TTS_PlayerOptions}/g;
|
||
|
||
$cmd = $TTS_MplayerCall;
|
||
} else {
|
||
$cmd = $TTS_MplayerCall . " " . $mplayerAudioOpts . $AlsaDevice . " " .$NoDebug. " " . $mplayerOpts . " " . $file;
|
||
}
|
||
|
||
|
||
my $mp3Duration = Text2Speech_CalcMP3Duration($hash, $file);
|
||
BlockingInformParent("Text2Speech_readingsSingleUpdateByName", [$hash->{NAME}, "duration", "$mp3Duration"], 0);
|
||
return $cmd;
|
||
}
|
||
|
||
#####################################
|
||
# Benutzt um Infos aus dem Blockingprozess
|
||
# in die Readings zu schreiben
|
||
#####################################
|
||
sub Text2Speech_readingsSingleUpdateByName($$$) {
|
||
my ($devName, $readingName, $readingVal) = @_;
|
||
my $hash = $defs{$devName};
|
||
Log3 $hash, 5, $hash->{NAME}.": readingsSingleUpdateByName: Dev:$devName Reading:$readingName Val:$readingVal";
|
||
readingsSingleUpdate($hash, $readingName, $readingVal, 1);
|
||
}
|
||
|
||
#####################################
|
||
# param1: string: MP3 Datei inkl. Pfad
|
||
#
|
||
# Ermittelt die Abspieldauer einer MP3 und gibt die Zeit in Sekunden zurück.
|
||
# Die Abspielzeit wird auf eine ganze Zahl gerundet
|
||
#####################################
|
||
sub Text2Speech_CalcMP3Duration($$) {
|
||
my $time;
|
||
my ($hash, $file) = @_;
|
||
eval {
|
||
my $tag = get_mp3info($file);
|
||
if ($tag && defined($tag->{SECS})) {
|
||
$time = int($tag->{SECS}+0.5);
|
||
Log3 $hash, 4, $hash->{NAME}.": $file hat eine Länge von $time Sekunden.";
|
||
}
|
||
};
|
||
|
||
if ($@) {
|
||
Log3 $hash, 2, $hash->{NAME}.": Bei der MP3-Längenermittlung ist ein Fehler aufgetreten: $@";
|
||
return undef;
|
||
}
|
||
return $time;
|
||
}
|
||
|
||
|
||
#####################################
|
||
# param1: hash : Hash
|
||
# param2: string: Dateiname
|
||
# param2: string: Text
|
||
#
|
||
# Holt den Text mithilfe der entsprechenden TTS_Ressource
|
||
#####################################
|
||
sub Text2Speech_Download($$$) {
|
||
my ($hash, $file, $text) = @_;
|
||
|
||
my $TTS_Ressource = AttrVal($hash->{NAME}, "TTS_Ressource", "Google");
|
||
my $TTS_User = AttrVal($hash->{NAME}, "TTS_User", "");
|
||
my $TTS_APIKey = AttrVal($hash->{NAME}, "TTS_APIKey", "");
|
||
my $TTS_Language = AttrVal($hash->{NAME}, "TTS_Language_Custom", $language{$TTS_Ressource}{AttrVal($hash->{NAME}, "TTS_Language", "Deutsch")});
|
||
my $TTS_Quality = AttrVal($hash->{NAME}, "TTS_Quality", "");
|
||
my $TTS_Speed = AttrVal($hash->{NAME}, "TTS_Speed", "");
|
||
my $cmd;
|
||
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": Verwende ".$TTS_Ressource." Resource zur TTS-Generierung";
|
||
|
||
if($TTS_Ressource =~ m/(Google|VoiceRSS)/) {
|
||
my $HttpResponse;
|
||
my $HttpResponseErr;
|
||
my $fh;
|
||
|
||
my $url = "https://" . $ttsHost{$TTS_Ressource} . $ttsPath{$TTS_Ressource};
|
||
$url .= $ttsLang{$TTS_Ressource} . $TTS_Language;
|
||
$url .= "&" . $ttsAddon{$TTS_Ressource} if(length($ttsAddon{$TTS_Ressource})>0);
|
||
$url .= "&" . $ttsUser{$TTS_Ressource} . $TTS_User if(length($ttsUser{$TTS_Ressource})>0);
|
||
$url .= "&" . $ttsAPIKey{$TTS_Ressource} . $TTS_APIKey if(length($ttsAPIKey{$TTS_Ressource})>0);
|
||
$url .= "&" . $ttsQuality{$TTS_Ressource} . $TTS_Quality if(length($ttsQuality{$TTS_Ressource})>0);
|
||
$url .= "&" . $ttsSpeed{$TTS_Ressource} . $TTS_Speed if(length($ttsSpeed{$TTS_Ressource})>0);
|
||
$url .= "&" . $ttsQuery{$TTS_Ressource} . uri_escape($text);
|
||
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": Hole URL: ". $url;
|
||
#$HttpResponse = GetHttpFile($ttsHost, $ttsPath . $ttsLang . $TTS_Language . "&" . $ttsQuery . uri_escape($text));
|
||
my $param = {
|
||
url => $url,
|
||
timeout => 5,
|
||
hash => $hash, # Muss gesetzt werden, damit die Callback funktion wieder $hash hat
|
||
method => "GET" # Lesen von Inhalten
|
||
#httpversion => "1.1",
|
||
#header => "User-Agent:Mozilla/5.0 (Windows NT 6.2; WOW64) AppleWebKit/537.22 (KHTML, like Gecko) Chrome/25.0.1364.172 Safari/537.22m" # Den Header gemäss abzufragender Daten ändern
|
||
#header => "agent: Mozilla/1.22\r\nUser-Agent: Mozilla/1.22"
|
||
};
|
||
($HttpResponseErr, $HttpResponse) = HttpUtils_BlockingGet($param);
|
||
|
||
if(length($HttpResponseErr) > 0) {
|
||
Log3 $hash->{NAME}, 3, $hash->{NAME}.": Fehler beim abrufen der Daten von " .$TTS_Ressource. " Translator";
|
||
Log3 $hash->{NAME}, 3, $hash->{NAME}.": " . $HttpResponseErr;
|
||
}
|
||
|
||
$fh = new IO::File ">$file";
|
||
if(!defined($fh)) {
|
||
Log3 $hash->{NAME}, 2, $hash->{NAME}.": mp3 Datei <$file> konnte nicht angelegt werden.";
|
||
return undef;
|
||
}
|
||
|
||
$fh->print($HttpResponse);
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": Schreibe mp3 in die Datei $file mit ".length($HttpResponse)." Bytes";
|
||
close($fh);
|
||
}
|
||
elsif ($TTS_Ressource eq "ESpeak") {
|
||
my $FileWav = $file . ".wav";
|
||
|
||
$cmd = "sudo espeak -vde+f3 -k5 -s150 \"" . $text . "\" -w \"" . $FileWav . "\"";
|
||
Log3 $hash, 4, $hash->{NAME}.":" .$cmd;
|
||
system($cmd);
|
||
|
||
$cmd = "lame \"" . $FileWav . "\" \"" . $file . "\"";
|
||
Log3 $hash, 4, $hash->{NAME}.":" .$cmd;
|
||
system($cmd);
|
||
unlink $FileWav;
|
||
}
|
||
elsif ($TTS_Ressource eq "SVOX-pico") {
|
||
my $FileWav = $file . ".wav";
|
||
|
||
$cmd = "pico2wave --lang=" . $TTS_Language . " --wave=\"" . $FileWav . "\" \"" . $text . "\"";
|
||
Log3 $hash, 4, $hash->{NAME}.":" .$cmd;
|
||
system($cmd);
|
||
|
||
$cmd = "lame \"" . $FileWav . "\" \"" . $file . "\"";
|
||
Log3 $hash, 4, $hash->{NAME}.":" .$cmd;
|
||
system($cmd);
|
||
unlink $FileWav;
|
||
}
|
||
elsif ($TTS_Ressource eq "Amazon-Polly") {
|
||
# with awscli
|
||
# aws polly synthesize-speech --output-format mp3 --voice-id Marlene --text '%text%' abc.mp3
|
||
#$cmd = "aws polly synthesize-speech --output-format json --speech-mark-types='[\"viseme\"]' --voice-id " . $TTS_Language . " --text '" . $text . "' " . $file;
|
||
#Log3 $hash, 4, $hash->{NAME}.":" .$cmd;
|
||
#system($cmd);
|
||
my $fh;
|
||
my $texttype = "text";
|
||
|
||
$texttype = "ssml" if($text =~ m/^<speak>.*<\/speak>$/);
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": Folgender TextTyp wurde für ".$TTS_Ressource." erkannt: ".$texttype;
|
||
|
||
my $polly = Paws->service('Polly', region => 'eu-central-1');
|
||
my $res = $polly->SynthesizeSpeech(
|
||
VoiceId => $TTS_Language,
|
||
Text => $text,
|
||
TextType => $texttype,
|
||
OutputFormat => 'mp3',
|
||
);
|
||
|
||
$fh = new IO::File ">$file";
|
||
if(!defined($fh)) {
|
||
Log3 $hash->{NAME}, 2, $hash->{NAME}.": mp3 Datei <$file> konnte nicht angelegt werden.";
|
||
return undef;
|
||
}
|
||
|
||
$fh->print($res->AudioStream);
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": Schreibe mp3 in die Datei $file mit ". $res->RequestCharacters ." Chars";
|
||
close($fh);
|
||
} elsif ( $TTS_Ressource eq 'maryTTS' ) {
|
||
my $mTTSurl = $TTS_User;
|
||
my($unnamed, $named) = parseParams($mTTSurl);
|
||
$named->{host} //= shift @{$unnamed} // '127.0.0.1';
|
||
$named->{port} //= shift @{$unnamed} // '59125';
|
||
$named->{lang} //= shift @{$unnamed} // !$TTS_Language || $TTS_Language eq 'Deutsch' ? 'de_DE' : $TTS_Language;
|
||
$named->{voice} //= shift @{$unnamed} // 'de_DE/thorsten_low';
|
||
$named->{endpoint} //= shift @{$unnamed} // 'process';
|
||
|
||
$mTTSurl = "http://$named->{host}:$named->{port}/$named->{endpoint}?INPUT_TYPE=TEXT&OUTPUT_TYPE=AUDIO&AUDIO=WAVE_FILE&LOCALE=$named->{lang}&VOICE=$named->{voice}&INPUT_TEXT="; # https://github.com/marytts/marytts-txt2wav/blob/python/txt2wav.py#L21
|
||
$mTTSurl .= uri_escape($text);
|
||
|
||
Log3( $hash->{NAME}, 4, "$hash->{NAME}: Hole URL: $mTTSurl" );
|
||
my $param = { url => $mTTSurl,
|
||
timeout => 5,
|
||
hash => $hash, # Muss gesetzt werden, damit die Callback funktion wieder $hash hat
|
||
method => 'GET' # POST can be found in https://github.com/marytts/marytts-txt2wav/blob/python/txt2wav.py#L33
|
||
};
|
||
my ($maryTTSResponseErr, $maryTTSResponse) = HttpUtils_BlockingGet($param);
|
||
|
||
if(length($maryTTSResponseErr) > 0) {
|
||
Log3($hash->{NAME}, 3, "$hash->{NAME}: Fehler beim Abrufen der Daten von $TTS_Ressource: $maryTTSResponseErr");
|
||
return;
|
||
}
|
||
|
||
my $FileWav2 = $file . '.wav';
|
||
my $fh2 = new IO::File ">$FileWav2";
|
||
if ( !defined $FileWav2 ) {
|
||
Log3($hash->{NAME}, 2, "$hash->{NAME}: wav Datei <$FileWav2> konnte nicht angelegt werden.");
|
||
return;
|
||
}
|
||
|
||
$fh2->print($maryTTSResponse);
|
||
Log3($hash->{NAME}, 4, "$hash->{NAME}: Schreibe wav in die Datei $FileWav2 mit ".length $maryTTSResponse . ' Bytes');
|
||
close $fh2;
|
||
$cmd = qq(lame "$FileWav2" "$file");
|
||
Log3($hash, 4, "$hash->{NAME}:$cmd");
|
||
system $cmd;
|
||
return unlink $FileWav2;
|
||
}
|
||
}
|
||
|
||
#####################################
|
||
sub Text2Speech_DoIt($) {
|
||
my ($hash) = @_;
|
||
|
||
my $TTS_CacheFileDir = AttrVal($hash->{NAME}, "TTS_CacheFileDir", "cache");
|
||
my $TTS_Ressource = AttrVal($hash->{NAME}, "TTS_Ressource", "Google");
|
||
my $TTS_Language = AttrVal($hash->{NAME}, "TTS_Language", "Deutsch");
|
||
my $TTS_SentenceAppendix = AttrVal($hash->{NAME}, "TTS_SentenceAppendix", undef); #muss eine mp3-Datei sein, ohne Pfadangabe
|
||
my $TTS_FileTemplateDir = AttrVal($hash->{NAME}, "TTS_FileTemplateDir", "templates");
|
||
my $TTS_OutputFile = AttrVal($hash->{NAME}, "TTS_OutputFile", undef);
|
||
|
||
my $myFileTemplateDir;
|
||
if($TTS_FileTemplateDir =~ m/^\/.*/) { $myFileTemplateDir = $TTS_FileTemplateDir; } else { $myFileTemplateDir = $TTS_CacheFileDir ."/". $TTS_FileTemplateDir;}
|
||
|
||
my $verbose = AttrVal($hash->{NAME}, "verbose", 3);
|
||
my $cmd;
|
||
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": Verwende TTS Spracheinstellung: ".$TTS_Language;
|
||
|
||
my $filename;
|
||
my $file;
|
||
|
||
unless(-e $TTS_CacheFileDir or mkdir $TTS_CacheFileDir) {
|
||
#Verzeichnis anlegen gescheitert
|
||
Log3 $hash->{NAME}, 2, $hash->{NAME}.": Angegebenes Verzeichnis $TTS_CacheFileDir konnte erstmalig nicht angelegt werden.";
|
||
return undef;
|
||
}
|
||
|
||
my @Mp3WrapFiles;
|
||
my @Mp3WrapText;
|
||
|
||
$TTS_SentenceAppendix = $myFileTemplateDir ."/". $TTS_SentenceAppendix if($TTS_SentenceAppendix);
|
||
undef($TTS_SentenceAppendix) if($TTS_SentenceAppendix && (! -e $TTS_SentenceAppendix));
|
||
|
||
#Abspielliste erstellen
|
||
my $AnzahlDownloads = 0;
|
||
foreach my $t (@{$hash->{helper}{Text2Speech}}) {
|
||
if(-e $t) {
|
||
# falls eine bestimmte mp3-Datei mit absolutem Pfad gespielt werden soll
|
||
$filename = $t;
|
||
$file = $filename;
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": $filename als direkte MP3 Datei erkannt!";
|
||
} elsif(-e $TTS_CacheFileDir."/".$t) {
|
||
# falls eine bestimmte mp3-Datei mit relativem Pfad gespielt werden soll
|
||
$filename = $t;
|
||
$file = $TTS_CacheFileDir."/".$filename;
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": $filename als direkte MP3 Datei erkannt!";
|
||
} else {
|
||
$filename = md5_hex($TTS_Ressource ."|". $t) . ".mp3";
|
||
$file = $TTS_CacheFileDir."/".$filename;
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": Textbaustein ist keine direkte MP3 Datei, ermittle MD5 CacheNamen: $filename";
|
||
}
|
||
|
||
if(-e $file) {
|
||
push(@Mp3WrapFiles, $file);
|
||
push(@Mp3WrapText, $t);
|
||
} else {
|
||
# es befindet sich noch Text zum Download in der Queue
|
||
if (AttrVal($hash->{NAME}, "TTS_SpeakAsFastAsPossible", 0) == 0 || (AttrVal($hash->{NAME}, "TTS_SpeakAsFastAsPossible", 0) == 1 && $AnzahlDownloads == 0)) {
|
||
# nur Download wenn kein TTS_SpeakAsFastAsPossible gesetzt ist oder der erste Download erfolgen soll
|
||
Text2Speech_Download($hash, $file, $t);
|
||
$AnzahlDownloads ++;
|
||
if(-e $file) {
|
||
push(@Mp3WrapFiles, $file);
|
||
push(@Mp3WrapText, $t);
|
||
}
|
||
} else {
|
||
last;
|
||
}
|
||
}
|
||
|
||
last if (AttrVal($hash->{NAME}, "TTS_UseMP3Wrap", 0) == 0);
|
||
# ohne mp3wrap darf nur ein Textbaustein verarbeitet werden
|
||
}
|
||
|
||
push(@Mp3WrapFiles, $TTS_SentenceAppendix) if($TTS_SentenceAppendix);
|
||
|
||
if (AttrVal($hash->{NAME}, "TTS_UseMP3Wrap", 0) == 1) {
|
||
# benutze das Tool MP3Wrap um bereits einzelne vorhandene Sprachdateien
|
||
# zusammenzuführen. Ziel: sauberer Sprachfluss
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": Bearbeite per MP3Wrap jetzt den Text: ". join(" ", @Mp3WrapText);
|
||
|
||
my $Mp3WrapFile;
|
||
my $Mp3WrapPrefix;
|
||
$Mp3WrapPrefix = md5_hex(join("|", @Mp3WrapFiles));
|
||
|
||
if ($TTS_OutputFile) {
|
||
if ($TTS_OutputFile !~ m/^\//) {
|
||
$TTS_OutputFile = $TTS_CacheFileDir ."/".$TTS_OutputFile;
|
||
}
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": Verwende fixen Dateinamen: $TTS_OutputFile";
|
||
$Mp3WrapFile = $TTS_OutputFile;
|
||
unlink($Mp3WrapFile);
|
||
} else {
|
||
$Mp3WrapFile = $TTS_CacheFileDir ."/". $Mp3WrapPrefix . ".mp3";
|
||
}
|
||
|
||
if (scalar(@Mp3WrapFiles) == 1) {
|
||
# wenn nur eine Datei, dann wird diese genutzt
|
||
$Mp3WrapFile = $Mp3WrapFiles[0];
|
||
} elsif(! -e $Mp3WrapFile) {
|
||
$cmd = "mp3wrap " .$Mp3WrapFile. " " .join(" ", @Mp3WrapFiles);
|
||
$cmd .= " >/dev/null" if($verbose < 5);
|
||
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": " .$cmd;
|
||
system($cmd);
|
||
|
||
my $t = substr($Mp3WrapFile, 0, length($Mp3WrapFile)-4)."_MP3WRAP.mp3";
|
||
if(-e $t){
|
||
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": Benenne Datei um von <".$t."> nach <".$Mp3WrapFile.">";
|
||
rename($t, $Mp3WrapFile);
|
||
#falls die Datei existiert den ID3V1 und ID3V2 Tag entfernen
|
||
my $ret = eval{ remove_mp3tag( $Mp3WrapFile, 'ALL' ) };
|
||
Log3 $hash, 1, $hash->{NAME}.": Fehle beim entfernen der ID3 Tags: $@" if ( $@ );
|
||
Log3 $hash, 4, $hash->{NAME}.": Die ID3 Tags ( $ret Bytes ) von $Mp3WrapFile wurden geloescht." if ( $ret > 0 );
|
||
Log3 $hash, 4, $hash->{NAME}.": Die ID3 Tags ( 0 Bytes ) von $Mp3WrapFile wurden geloescht." if ( $ret == -1 );
|
||
Log3 $hash->{NAME}, 3, "MP3::Info Modul fehlt, konnte MP3 Tags nicht entfernen!" if ( !$ret );
|
||
|
||
} else {
|
||
|
||
Log3 $hash->{NAME}, 3, $hash->{NAME}.": MP3WRAP Fehler!, Datei wurde nicht generiert.";
|
||
|
||
}
|
||
}
|
||
|
||
if ($TTS_OutputFile && $TTS_OutputFile ne $Mp3WrapFile) {
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": Benenne Datei um von <".$Mp3WrapFile."> nach <".$TTS_OutputFile.">";
|
||
rename($Mp3WrapFile, $TTS_OutputFile);
|
||
$Mp3WrapFile = $TTS_OutputFile;
|
||
}
|
||
|
||
if ($hash->{MODE} ne "SERVER") {
|
||
# im Server Mode, nicht die Datei abspielen
|
||
if(-e $Mp3WrapFile) {
|
||
$cmd = Text2Speech_BuildMplayerCmdString($hash, $Mp3WrapFile);
|
||
$cmd .= " >/dev/null" if($verbose < 5);
|
||
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": " .$cmd;
|
||
system($cmd);
|
||
} else {
|
||
Log3 $hash->{NAME}, 2, $hash->{NAME}.": Mp3Wrap Datei konnte nicht gefunden werden.";
|
||
}
|
||
}
|
||
|
||
return $hash->{NAME} ."|".
|
||
($TTS_SentenceAppendix ? scalar(@Mp3WrapFiles)-1: scalar(@Mp3WrapFiles)) ."|".
|
||
$Mp3WrapFile;
|
||
}
|
||
|
||
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": Bearbeite jetzt den Text: ". $hash->{helper}{Text2Speech}[0];
|
||
|
||
if(! -e $file) { # Datei existiert noch nicht im Cache
|
||
Text2Speech_Download($hash, $file, $hash->{helper}{Text2Speech}[0]);
|
||
} else {
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.": $file gefunden, kein Download";
|
||
}
|
||
|
||
if(-e $file && $hash->{MODE} ne "SERVER") {
|
||
# Datei existiert jetzt
|
||
# im Falls Server, nicht die Datei abspielen
|
||
$cmd = Text2Speech_BuildMplayerCmdString($hash, $file);
|
||
$cmd .= " >/dev/null" if($verbose < 5);
|
||
|
||
Log3 $hash->{NAME}, 4, $hash->{NAME}.":" .$cmd;
|
||
system($cmd);
|
||
}
|
||
return $hash->{NAME}. "|".
|
||
"1" ."|".
|
||
$file;
|
||
}
|
||
|
||
####################################################
|
||
# Rückgabe der Blockingfunktion
|
||
# param1: HashName
|
||
# param2: Anzahl der abgearbeiteten Textbausteine
|
||
# param3: Dateiname der abgespielt wurde
|
||
####################################################
|
||
sub Text2Speech_Done($) {
|
||
my ($string) = @_;
|
||
return unless(defined($string));
|
||
|
||
my @a = split("\\|",$string);
|
||
my $hash = $defs{shift(@a)};
|
||
my $tts_done = shift(@a);
|
||
my $filename = shift(@a);
|
||
|
||
my $TTS_TimeOut = AttrVal($hash->{NAME}, "TTS_TimeOut", 60);
|
||
|
||
if($filename) {
|
||
my @text;
|
||
for(my $i=0; $i<$tts_done; $i++) {
|
||
push(@text, $hash->{helper}{Text2Speech}[$i]);
|
||
}
|
||
Text2Speech_WriteStats($hash, 1, $filename, join(" ", @text)) if (AttrVal($hash->{NAME},"TTS_noStatisticsLog", "0")==0);
|
||
|
||
readingsBeginUpdate( $hash );
|
||
readingsBulkUpdate($hash, 'lastFilename', $filename );
|
||
# Update der Dauer im Servermode
|
||
readingsBulkUpdate( $hash, 'duration', Text2Speech_CalcMP3Duration( $hash, $filename ) ) if( $hash->{MODE} eq "SERVER" );
|
||
readingsEndUpdate( $hash, 1 );
|
||
|
||
# Aufruf eine eines Abspielgerätes wenn das Attibut gesetzt ist
|
||
my $playercall = AttrVal( $hash->{NAME}, 'TTS_RemotePlayerCall', '' );
|
||
eval $playercall if( $playercall );
|
||
Log3( $hash, 1, $hash->{NAME}." TTS_RemotePlayerCall: eval error $@.") if( $@ );
|
||
}
|
||
|
||
delete($hash->{helper}{RUNNING_PID});
|
||
splice(@{$hash->{helper}{Text2Speech}}, 0, $tts_done);
|
||
|
||
# erneutes aufrufen da ev. weiterer Text in der Warteschlange steht
|
||
if(@{$hash->{helper}{Text2Speech}} > 0) {
|
||
# es wurde nur ein Teil abgearbeitet
|
||
Log3($hash,4, $hash->{NAME}.": Es wurde nur ein Teil ausgegeben und weitere Teile folgen!");
|
||
|
||
$hash->{helper}{RUNNING_PID} = BlockingCall("Text2Speech_DoIt", $hash, "Text2Speech_Done", $TTS_TimeOut, "Text2Speech_AbortFn", $hash);
|
||
} else {
|
||
|
||
# alles wurde bearbeitet
|
||
Log3($hash,4, $hash->{NAME}.": Es wurden alle Teile ausgegeben und der Befehl ist abgearbeitet.");
|
||
readingsSingleUpdate($hash, "playing", "0", 1);
|
||
}
|
||
}
|
||
|
||
#####################################
|
||
sub Text2Speech_AbortFn($) {
|
||
my ($hash) = @_;
|
||
|
||
delete($hash->{helper}{RUNNING_PID});
|
||
Log3 $hash->{NAME}, 2, $hash->{NAME}.": BlockingCall for ".$hash->{NAME}." was aborted";
|
||
readingsSingleUpdate($hash, "playing", "0", 1);
|
||
}
|
||
|
||
#####################################
|
||
# Hiermit werden Statistken per DbLogModul gesammelt
|
||
# Wichitg zur Entscheidung welche Dateien aus dem Cache lange
|
||
# nicht benutzt und somit gelöscht werden koennen.
|
||
#
|
||
# param1: hash
|
||
# param2: int: 0=indirekt (über mp3wrap); 1=direkt abgespielt
|
||
# param3: string: Datei
|
||
# param4: string: Text der als mp3 abgespielt wird
|
||
#####################################
|
||
sub Text2Speech_WriteStats($$$$){
|
||
my($hash, $typ, $file, $text) = @_;
|
||
my $DbLogDev;
|
||
|
||
#suche ein DbLogDevice
|
||
return undef unless($modules{"DbLog"} && $modules{"DbLog"}{"LOADED"});
|
||
foreach my $key (keys(%defs)) {
|
||
if($defs{$key}{TYPE} eq "DbLog") {
|
||
$DbLogDev = $key;
|
||
last;
|
||
}
|
||
}
|
||
return undef if($defs{$DbLogDev}{STATE} !~ m/(active|connected)/); # muss active sein!
|
||
return undef if(AttrVal($defs{$DbLogDev}, "DbLogType", "History") !~ /Current/); # muss die Tabelle Current nutzen
|
||
|
||
my $logdevice = $hash->{NAME} ."|". $file;
|
||
# den letzten Value von "Usage" ermitteln um dann die Statistik um 1 zu erhoehen.
|
||
my @LastValue = DbLog_Get($defs{$DbLogDev}, "", "current", "array", "-", "-", $logdevice.":Usage");
|
||
my $NewValue = 1;
|
||
$NewValue = $LastValue[0]{value} + 1 if($LastValue[0]);
|
||
|
||
my $cmd;
|
||
if ($NewValue == 1) {
|
||
$cmd = "INSERT INTO current (TIMESTAMP, DEVICE, TYPE, EVENT, READING, VALUE, UNIT) VALUES (\
|
||
'".TimeNow()."','".$logdevice."','".$hash->{TYPE}."','".$text."','Usage','".$NewValue."','')";
|
||
} else {
|
||
$cmd = "UPDATE current SET VALUE = '".$NewValue."', TIMESTAMP = '".TimeNow()."' WHERE DEVICE ='".$logdevice."'";
|
||
}
|
||
DbLog_ExecSQL($defs{$DbLogDev}, $cmd);
|
||
}
|
||
|
||
#########################
|
||
sub Text2Speech_readMp3(@) {
|
||
my ($hash) = @_;
|
||
my $name = $hash->{NAME};
|
||
my $type = $hash->{TYPE};
|
||
my $iam = "$type $name Text2Speech_readMp3:";
|
||
my $filename = ReadingsVal( $name, 'lastFilename', '' );
|
||
|
||
if ( $filename and -e $filename ) {
|
||
|
||
if( open my $fh, '<:raw', $filename ) {
|
||
|
||
my $content = '';
|
||
|
||
while (1) {
|
||
my $success = read $fh, $content, 1024, length( $content );
|
||
if( not defined $success ) {
|
||
close $fh;
|
||
Log3 $name, 1, "$iam read file \"$filename\" error: $!.";
|
||
return undef;
|
||
}
|
||
|
||
last if not $success;
|
||
}
|
||
|
||
close $fh;
|
||
Log3 $name, 4, "$iam file \"$filename\" content length: ".length( $content );
|
||
return \$content;
|
||
|
||
} else {
|
||
Log3 $name, 1, "$iam open file \"$filename\" error: $!.";
|
||
}
|
||
|
||
} else {
|
||
Log3 $name, 2, "$iam file \"$filename\" does not exist.";
|
||
}
|
||
return undef;
|
||
}
|
||
|
||
#########################
|
||
sub Text2Speech_getLastMp3 {
|
||
my ($request) = @_;
|
||
|
||
if ( $request =~ /^\/(Text2Speech)\/(\w+)\/last.mp3/ ) {
|
||
|
||
my $type = $1;
|
||
my $name = $2;
|
||
my $hash = $defs{$name};
|
||
my $audioData = Text2Speech_readMp3( $hash );
|
||
return ( "text/plain; charset=utf-8","${type} ${name}: No MP3 file for webhook $request" ) if ( !$audioData );
|
||
my $audioMime = 'audio/mpeg';
|
||
|
||
return ( $audioMime, $$audioData );
|
||
}
|
||
return ( "text/plain; charset=utf-8", "No Text2Speech device for webhook $request" );
|
||
}
|
||
|
||
#########################
|
||
sub Text2Speech_AddExtension(@) {
|
||
my ( $name, $func, $link ) = @_;
|
||
my $hash = $defs{$name};
|
||
my $type = $hash->{TYPE};
|
||
|
||
my $url = "/$link";
|
||
Log3( $name, 2, "Registering $type $name for URL $url..." );
|
||
$::data{FWEXT}{$url}{deviceName} = $name;
|
||
$::data{FWEXT}{$url}{FUNC} = $func;
|
||
$::data{FWEXT}{$url}{LINK} = $link;
|
||
|
||
return;
|
||
}
|
||
|
||
#########################
|
||
sub Text2Speech_RemoveExtension(@) {
|
||
my ($link) = @_;
|
||
my $url = "/$link";
|
||
my $name = $::data{FWEXT}{$url}{deviceName};
|
||
|
||
Log3( $name, 2, "Unregistering URL $url..." );
|
||
delete $::data{FWEXT}{$url};
|
||
|
||
return;
|
||
}
|
||
|
||
|
||
1;
|
||
|
||
=pod
|
||
=item helper
|
||
=item summary A module that converts text to speech and also plays \
|
||
the result on a local or remote loudspeaker
|
||
|
||
=item summary_DE Modul, das Text in Sprache umwandelt und das Ergebnis \
|
||
über einen lokalen oder entfernten Lautsprecher wiedergibt
|
||
=begin html
|
||
|
||
<a id="Text2Speech"></a>
|
||
<h3>Text2Speech</h3>
|
||
<ul>
|
||
<br>
|
||
<a id="Text2Speech-define"></a>
|
||
<h4>Define</h4>
|
||
<ul>
|
||
<b>Local : </b><code>define <name> Text2Speech <alsadevice></code><br>
|
||
<b>Remote: </b><code>define <name> Text2Speech <host>[:<portnr>][:SSL] [portpassword]</code>
|
||
<b>Server: </b><code>define <name> Text2Speech none</code><br>
|
||
<p>
|
||
This module converts any text into speech with several possible providers. The Device can be defined as locally
|
||
or remote instance.
|
||
</p>
|
||
|
||
<li>
|
||
<b>Local Device</b><br>
|
||
<ul>
|
||
The output will be sent to any connected audio device. For example speakers connected per jack,
|
||
network, WiFI or Bluetooth. Playback can be done using MPlayer or any other application.<br>
|
||
<br>
|
||
Mplayer installation under Debian/Ubuntu/Raspbian:<br>
|
||
<code>apt-get install mplayer</code><br>
|
||
The given alsa device has to be configured in <code>/etc/asound.conf</code>
|
||
<p>
|
||
<b>Special AlsaDevice: </b><i>default</i><br>
|
||
The internal Mplayer command will be without any audio directive if the given alsa device is <i>default</i>.
|
||
In this case Mplayer is using the standard audio device.
|
||
</p>
|
||
<p>
|
||
<b>Example:</b><br>
|
||
<code>define MyTTS Text2Speech hw=0.0</code><br>
|
||
<code>define MyTTS Text2Speech default</code>
|
||
</p>
|
||
</ul>
|
||
</li>
|
||
|
||
<li>
|
||
<b>Remote Device</b><br>
|
||
<ul>
|
||
This module can be configured as remote-device for client-server environments. The Client has to be configured
|
||
as local device.<br>
|
||
Notice: the Name of the locally instance has to be the same!
|
||
<ul>
|
||
<li>Host: setting up IP-adress</li>
|
||
<li>PortNr: setting up TelnetPort of FHEM; default: 7072</li>
|
||
<li>SSL: setting up if connect over SSL; default: no SSL</li>
|
||
<li>PortPassword: setting up the configured target telnet passwort</li>
|
||
</ul>
|
||
<p>
|
||
<b>Example:</b><br>
|
||
<code>define MyTTS Text2Speech 192.168.178.10:7072 fhempasswd</code>
|
||
<code>define MyTTS Text2Speech 192.168.178.10</code>
|
||
</p>
|
||
If a PRESENCE Device is avilable for the host IP-address, than this will be used to detect the reachability instead of the blocking internal method.
|
||
</ul>
|
||
</li>
|
||
|
||
<li>
|
||
<b>Server Device</b>
|
||
<ul>
|
||
In case of an usage as a server, only the mp3 file will be generated and displayed as lastFilename reading. It makes no sense to use the attribute <i>TTS_speakAsFastAsPossible</i> here.
|
||
Its recommend, to use the attribute <i>TTS_useMP3Wrap</i>. Otherwise only the last audiobrick will be shown in reading <i>lastFilename</i>.
|
||
</ul>
|
||
<p>
|
||
<b>Example:</b><br>
|
||
<code>define MyTTS Text2Speech none</code>
|
||
</p>
|
||
</li>
|
||
|
||
</ul>
|
||
</ul>
|
||
|
||
<a id="Text2Speech-set"></a>
|
||
<h4>Set</h4>
|
||
<ul>
|
||
<a id="Text2Speech-set-tts"></a><li><b>tts</b>:<br>
|
||
Definition of text for voice output. To output mp3 files directly, they must be specified with
|
||
leading and closing colons. Therefore, the text itself must not contain any double punctuation.
|
||
The mp3 files must be stored in the <i>TTS_FileTemplateDir</i> directory.
|
||
SSML can be used for the Amazon Polly language engine. See examples.
|
||
</li>
|
||
<a id="Text2Speech-set-volume"></a><li><b>volume</b>:<br>
|
||
Setting up the volume audio response.<br>
|
||
Notice: Only available in locally instances!
|
||
</li>
|
||
</ul><br>
|
||
|
||
<a id="Text2Speech-get"></a>
|
||
<h4>Get</h4>
|
||
<ul>N/A</ul><br>
|
||
|
||
<a id="Text2Speech-attr"></a>
|
||
<h4>Attributes</h4>
|
||
<ul>
|
||
<a id="Text2Speech-attr-TTS_Delimiter"></a><li>TTS_Delimiter<br>
|
||
Optional: By using the Google engine, its not possible to convert more than 100 characters in a single audio brick.
|
||
With a delimiter the audio brick will be split at this character. A delimiter must be a single character!<br>
|
||
By default, each audio brick will be split at sentence end. Is a single sentence longer than 100 characters,
|
||
the sentence will be split additionally at comma, semicolon and the word <i>and</i>.<br>
|
||
Notice: Only available in locally instances with Google engine!
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_Ressource"></a><li>TTS_Ressource<br>
|
||
Optional: Selection of the Translator Engine<br>
|
||
Notice: Only available in locally instances!
|
||
<ul>
|
||
<li>Google<br>
|
||
Google Engine. Prerequisite: Active Internet connection<br>
|
||
This engine is recommended for its quality and is used by default.
|
||
</li>
|
||
<li>VoiceRSS<br>
|
||
VoiceRSS Engine. Prerequisite: Active Internet connection<br>
|
||
Free of charge up to 350 requests per day. If you need more, you have to pay.
|
||
This engine is also recommended due to its quality. To use this engine, you need an APIKey (see TTS_APIKey)
|
||
</li>
|
||
<li>ESpeak<br>
|
||
eSpeak Engine. Prerequisite: Installation of Espeak and lame<br>
|
||
eSpeak is an open source software speech synthesizer for English and other languages.
|
||
</li>
|
||
<li>SVOX-pico<br>
|
||
SVOX-Pico TTS-Engine (from the AOSP). Prerequisite: Installation of SVOX-Pico and lame<br>
|
||
<code>sudo apt-get install libttspico-utils lame</code><br><br>
|
||
On ARM/Raspbian the package <code>libttspico-utils</code>,<br>
|
||
so you may have to compile it yourself or use the precompiled package from <a target="_blank" href"http://www.robotnet.de/2014/03/20/sprich-freund-und-tritt-ein-sprachausgabe-fur-den-rasberry-pi/">this guide</a>, in short:<br>
|
||
<code>sudo apt-get install libpopt-dev lame</code><br>
|
||
<code>cd /tmp</code><br>
|
||
<code>wget http://www.dr-bischoff.de/raspi/pico2wave.deb</code><br>
|
||
<code>sudo dpkg --install pico2wave.deb</code>
|
||
</li>
|
||
<li>Amazon-Polly<br>
|
||
Amazon Polly Engine. Prerequisite: Active Internet connection, Perl package Paws<br>
|
||
Amazon service that turns text into lifelike speech. An AWS Access and Polly Aws User is required.<br>
|
||
<code>cpan paws</code><br>
|
||
The credentials to your AWS Polly are expected at ~/.aws/credentials<br>
|
||
<code>[default]
|
||
aws_secret_access_key = xxxxxxxxxxxxxxxxxxxxxx
|
||
aws_access_key_id = xxxxxxxxxxxxxxx
|
||
</code>
|
||
</li>
|
||
<li>maryTTS<br>
|
||
<a target="_blank" href"https://github.com/marytts/marytts">maryTTS</a> or <a target="_blank" href"https://github.com/MycroftAI/mimic3">Mimic 3</a> Engine. Prerequisite: Installation of respective server and lame, appropriate settings in <a href="#Text2Speech-attr-TTS_User">TTS_User</a> attribute (if other than default settings shall be applied).<br>
|
||
Both are open source software speech synthesizers for English and other languages.
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_Language"></a><li>TTS_Language<br>
|
||
Selection of different languages
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_Language_Custom"></a><li>TTS_Language_Custom<br>
|
||
If you want another engine and speech of default languages, you can insert this here.<br>
|
||
The definition depends on the used engine. This attribute overrides an TTS_Language attribute.<br>
|
||
Please refer to the specific API reference.
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_APIKey"></a><li>TTS_APIKey<br>
|
||
An APIKey is needed if you want to use VoiceRSS. You have to register at the following page:<br>
|
||
http://www.voicerss.org/registration.aspx
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_User"></a><li>TTS_User<br>
|
||
Actual only used for maryTTS (and Mimic 3). Needed in case if a TTS Engine needs a username and an APIKey for a request. <br>
|
||
<p>(Full) example for maryTTS (values are defaults and may be left out):</p>
|
||
<p><code>attr t2s TTS_User host=127.0.0.1 port=59125 lang=de_DE voice=de_DE/thorsten_low</code></p>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_CacheFileDir"></a><li>TTS_CacheFileDir<br>
|
||
Optional: The downloaded Google audio bricks are saved in this folder.
|
||
No automatic delete/cleanup available.<br>
|
||
Default: <i>cache/</i><br>
|
||
Notice: Available on local instances only!
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_UseMP3Wrap"></a><li>TTS_UseMP3Wrap<br>
|
||
For best voice output, it is recommended that the individual downloads are combined into a single file.
|
||
Each downloaded audio bricks are concatinated to a single audio file to play with Mplayer.<br>
|
||
Installtion of the mp3wrap package is required.<br>
|
||
<code>apt-get install mp3wrap</code><br>
|
||
Notice: Available on local instances only!
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_MplayerCall"></a><li>TTS_MplayerCall<br>
|
||
Optional: Definition of the system call to Mplayer or a different tool.<br>
|
||
If a tool other than Mplayer is used, the following templates apply:<br>
|
||
<ul>
|
||
<li>{device}</li>
|
||
<li>{volume}</li>
|
||
<li>{volumeadjust}</li>
|
||
<li>{file}</li>
|
||
<li>{options}</li>
|
||
</ul>
|
||
{options} are provided inside the text in parentheses during the set command.
|
||
Used for example to set special parameters for each call separately<br>
|
||
Example: <code>set myTTS tts [192.168.0.1:7000] This is my text</code><br><br>
|
||
|
||
Examples:<br>
|
||
<code>attr myTTS TTS_MplayerCall sudo /usr/bin/mplayer</code><br>
|
||
<code>attr myTTS TTS_MplayerCall AUDIODEV={device} play -q -v {volume} {file}</code><br>
|
||
<code>attr myTTS TTS_MplayerCall player {file} {options}</code><br>
|
||
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_SentenceAppendix"></a><li>TTS_SentenceAppendix<br>
|
||
Optional: Definition of one mp3-file to append each time of audio response.<br>
|
||
Mp3Wrap is required. The audio chunks must be downloaded to the CacheFileDir beforehand.
|
||
Example: <code>silence.mp3</code>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_FileMapping"></a><li>TTS_FileMapping<br>
|
||
Definition of mp3files with a custom template definition. Separated by space.
|
||
All template definitions can be used in audiobricks by <i>tts</i> command.
|
||
The definition must begin and end with a colon.
|
||
The mp3files must be saved in <i>TTS_FIleTemplateDir</i>.<br>
|
||
<code>attr myTTS TTS_FileMapping ring:ringtone.mp3 beep:MyBeep.mp3</code><br>
|
||
<code>set MyTTS tts Attention: This is my ringtone :ring: Its loud?</code>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_FileTemplateDir"></a><li>TTS_FileTemplateDir<br>
|
||
Directory to save all mp3-files are defined in <i>TTS_FileMapping</i> und <i>TTS_SentenceAppendix</i><br>
|
||
Optional, Default: <code>cache/templates</code>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_VolumeAdjust"></a><li>TTS_VolumeAdjust<br>
|
||
Basic volume increase<br>
|
||
Default: 110<br>
|
||
<code>attr myTTS TTS_VolumeAdjust 400</code>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_noStatisticsLog"></a><li>TTS_noStatisticsLog<br>
|
||
If set to <b>1</b>, it prevents logging statistics to DbLog Devices, default is <b>0</b><br>
|
||
Note: This logging is important to be able to delete cache files that have not been used for a longer period of time.
|
||
If you disable this, you will have to clean your cache directory manually.
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_speakAsFastAsPossible"></a><li>TTS_speakAsFastAsPossible<br>
|
||
Trying to get a speech as fast as possible. In case of not present audio bricks, you can
|
||
hear a short break as the audio brick will be downloaded at this time.
|
||
In case of a presentation of all audio bricks at local cache, this attribute has no impact.<br>
|
||
Attribute is only valid on local or server instances.
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_OutputFile"></a><li>TTS_OutputFile<br>
|
||
Definition of a fixed file name as mp3 output. The attribute is only relevant in conjunction with TTS_UseMP3Wrap.
|
||
If a file name is specified, then TTS_CacheFileDir is also taken into account.<br>
|
||
<code>attr myTTS TTS_OutputFile output.mp3</code><br>
|
||
<code>attr myTTS TTS_OutputFile /media/miniDLNA/output.mp3</code><br>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_RemotePlayerCall"></a><li>TTS_RemotePlayerCall<br>
|
||
The Text2Speech devices provide a URL to the last generated mp3 file:
|
||
<code><protocol>://<fhem server ip or name>:<fhem port>/fhem/Text2Speech/<device name>/last.mp3.</code><br>
|
||
If this attibute contains a remote player call, it will be executed after the last mp3 file is generated.<br>
|
||
<code>attr <device name> TTS_RemotePlayerCall GetFileFromURL('<protocol>://<remote player name or ip>:<remote player port>/?cmd=playSound&url=<protocol>://<fhem server name orip>:<fhem port>/fhem/Text2Speech/<device name>/last.mp3&loop=false&password=<password>')</code><br>
|
||
</li>
|
||
|
||
<li><a href="#readingFnAttributes">readingFnAttributes</a></li><br>
|
||
|
||
<li><a href="#disable">disable</a><br>
|
||
If this attribute is activated, the sound output will be disabled.<br>
|
||
Possible values: 0 => not disabled , 1 => disabled<br>
|
||
Default Value is 0 (not disabled)<br><br>
|
||
</li>
|
||
|
||
<li><a href="#verbose">verbose</a><br>
|
||
<b>4:</b> each step will be logged<br>
|
||
<b>5:</b> Additionally the individual debug information from Mplayer and mp3wrap will be logged
|
||
</li>
|
||
|
||
</ul><br>
|
||
|
||
<a id="Text2Speech-examples"></a>
|
||
<h4>Examples</h4>
|
||
<ul>
|
||
<code>define TTS_EG_WZ Text2Speech hw=/dev/snd/controlC3</code><br>
|
||
<code>attr TTS_EG_WZ TTS_Language English</code><br>
|
||
<code>attr TTS_EG_WZ TTS_Ressource Amazon-Polly</code><br>
|
||
<code>attr TTS_EG_WZ TTS_UseMP3Wrap 1</code><br><br>
|
||
<code>set MyTTS tts <speak>Mary had a little lamb.</speak></code>
|
||
<br>
|
||
<code>define MyTTS Text2Speech hw=0.0</code><br>
|
||
<code>set MyTTS tts The alarm system is ready.</code><br>
|
||
<code>set MyTTS tts :beep.mp3:</code><br>
|
||
<code>set MyTTS tts :mytemplates/alarm.mp3:The alarm system is ready.:ring.mp3:</code>
|
||
<br>
|
||
Example of MaryTTS and using SSML: <br>
|
||
<code>
|
||
define T2S Text2Speech default
|
||
attr T2S TTS_MplayerCall /usr/bin/mplayer
|
||
attr T2S TTS_Ressource maryTTS
|
||
attr T2S TTS_User host=192.168.100.1 port=59125 lang=de_DE voice=de_DE/thorsten_low ssml=1
|
||
set T2S tts '<voice name="de_DE/m-ailabs_low#rebecca_braunert_plunkett">Das ist ein Test in deutsch </voice><voice name="en_US/vctk_low#p236">and this is an test in english.</voice>'
|
||
</code>
|
||
</ul>
|
||
|
||
=end html
|
||
=begin html_DE
|
||
|
||
<a id="Text2Speech"></a>
|
||
<h3>Text2Speech</h3>
|
||
<ul>
|
||
<br>
|
||
<a id="Text2Speech-define"></a>
|
||
<h4>Define</h4>
|
||
<ul>
|
||
<b>Local : </b><code>define <name> Text2Speech <alsadevice></code><br>
|
||
<b>Remote: </b><code>define <name> Text2Speech <host>[:<portnr>][:SSL] [portpassword]</code>
|
||
<b>Server : </b><code>define <name> Text2Speech none</code><br>
|
||
<p>
|
||
Das Modul wandelt Text mittels verschiedener Provider/Ressourcen in Sprache um. Dabei kann das Device als
|
||
Remote, Lokales Device oder als Server konfiguriert werden.
|
||
</p>
|
||
|
||
<li>
|
||
<b>Local Device</b><br>
|
||
<ul>
|
||
Die Ausgabe wird an jedes angeschlossene Audiogerät gesendet. Zum Beispiel an einen lokalen Lautsprecher oder an
|
||
entfernte Geräte via Netzwerk, WiFI oder Bluetooth. Die Wiedergabe kann über MPlayer oder jede andere
|
||
Anwendung erfolgen.<br>
|
||
<br>
|
||
Mplayer-Installation unter Debian/Ubuntu/Raspbian:<br>
|
||
<code>apt-get install mplayer</code><br>
|
||
Das angegebene Alsa-Device ist in der <code>/etc/asound.conf</code> zu konfigurieren.
|
||
<p>
|
||
<b>Special AlsaDevice: </b><i>default</i><br>
|
||
Ist als Alsa-Device <i>default</i> angegeben, so wird Mplayer ohne eine Audiodevice-Angabe aufgerufen.
|
||
Dementsprechend verwendet Mplayer dann das Standard-Audio Ausgabedevice.
|
||
</p>
|
||
<p>
|
||
<b>Beispiel:</b><br>
|
||
<code>define MyTTS Text2Speech hw=0.0</code><br>
|
||
<code>define MyTTS Text2Speech default</code>
|
||
</p>
|
||
</ul>
|
||
</li>
|
||
|
||
<li>
|
||
<b>Remote Device</b><br>
|
||
<ul>
|
||
Das Modul ist Client-Server fäas bedeutet, das auf der Haupt-FHEM Installation eine Text2Speech-Instanz
|
||
als Remote definiert wird. Auf dem Client wird Text2Speech als Local definiert. Die Sprachausgabe erfolgt auf
|
||
der lokalen Instanz.<br>
|
||
Zu beachten ist, dass die Text2Speech Instanz (Definition als Local-Device) auf dem Zieldevice identisch benannt ist.
|
||
<ul>
|
||
<li>Host: Angabe der IP-Adresse</li>
|
||
<li>PortNr: Angabe des Telnet-Ports von FHEM; default: 7072</li>
|
||
<li>SSL: Angabe, ob der Zugriff per SSL erfolgen soll oder nicht; default: kein SSL</li>
|
||
<li>PortPassword: Angabe des in der Ziel-FHEM-Installation angegebenen Telnet-Passworts</li>
|
||
</ul>
|
||
<p>
|
||
<b>Beispiel:</b><br>
|
||
<code>define MyTTS Text2Speech 192.168.178.10:7072 fhempasswd</code>
|
||
<code>define MyTTS Text2Speech 192.168.178.10</code>
|
||
</p>
|
||
Wenn ein PRESENCE Gerät die Host-IP-Adresse abfragt, wird die blockierende interne Prüfung auf Erreichbarkeit umgangen und das PRESENCE Gerät genutzt.
|
||
</ul>
|
||
</li>
|
||
|
||
<li>
|
||
<b>Server Device</b>
|
||
<ul>
|
||
Im Falle der Verwendung als Server wird nur die MP3-Datei erstellt und als Reading lastFilename dargestellt.
|
||
Es ergibt keinen Sinn hier das Attribut <i>TTS_speakAsFastAsPossible</i> zu verwenden.
|
||
Die Verwendung des Attributs <i>TTS_useMP3Wrap</i> wird dringend empfohlen.
|
||
Ansonsten wird hier nur der letzte Teiltext als mp3 Datei im Reading dargestellt.
|
||
</ul>
|
||
<p>
|
||
<b>Beispiel:</b><br>
|
||
<code>define MyTTS Text2Speech none</code>
|
||
</p>
|
||
</li>
|
||
|
||
</ul>
|
||
</ul>
|
||
|
||
<a id="Text2Speech-set"></a>
|
||
<h4>Set</h4>
|
||
<ul>
|
||
<li><b>tts</b>:<br>
|
||
Setzen eines Textes zur Sprachausgabe. Um mp3-Dateien direkt auszugeben, müssen diese mit führenden
|
||
und schließenden Doppelpunkten angegebenen sein. Die MP3-Dateien müssen unterhalb des Verzeichnisses <i>TTS_FileTemplateDir</i> gespeichert sein.<br>
|
||
Der Text selbst darf deshalb selbst keine Doppelpunkte beinhalten. <br>
|
||
Für die Sprachengine Amazon Polly kann auch SSML verwendet werden. Siehe Beispiele.
|
||
</li>
|
||
<li><b>volume</b>:<br>
|
||
Setzen der Ausgabe Lautstärke.<br>
|
||
Achtung: Nur bei einem lokal definierter Text2Speech Instanz möglich!
|
||
</li>
|
||
</ul><br>
|
||
|
||
<a id="Text2Speech-get"></a>
|
||
<h4>Get</h4>
|
||
<ul>N/A</ul><br>
|
||
|
||
<a id="Text2Speech-attr"></a>
|
||
<h4>Attribute</h4>
|
||
<ul>
|
||
<a id="Text2Speech-attr-TTS_Delimiter"></a><li>TTS_Delimiter<br>
|
||
Optional: Wird ein Delimiter angegeben, so wird der Sprachbaustein an dieser Stelle geteilt.
|
||
Als Delimiter ist nur ein einzelnes Zeichen zulässig.
|
||
Hintergrund ist die Tatsache, dass die einige Sprachengines nur eine bestimmte Anzahl an Zeichen (z. B. Google nur 100Zeichen) zulässt.<br>
|
||
Im Standard wird nach jedem Satzende geteilt. Ist ein einzelner Satz länger als 100 Zeichen,
|
||
so wird zusätzlich nach Kommata, Semikolon und dem Verbindungswort <i>und</i> geteilt.<br>
|
||
Achtung: Nur bei einem lokal definierter Text2Speech Instanz möglich!<br>
|
||
<b>Notation</b><br>
|
||
+ -> erzwinge das Trennen, auch wenn Textbaustein < x Zeichen<br>
|
||
- -> trenne, nur wenn Textbaustein > x Zeichen
|
||
af -> add first -> füge den Delimiter am Satzanfang wieder hinzu<br>
|
||
al -> add last -> füge den Delimiter am Satzende wieder hinzu<br>
|
||
an -> add nothing -> Delimiter nicht wieder hinzufügen<br>
|
||
~ -> der Delimiter<br>
|
||
<b>Beispiel</b><br>
|
||
<code>attr myTTS TTS_Delimiter -al.</code>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_Ressource"></a><li>TTS_Ressource<br>
|
||
Optional: Auswahl der Sprachengine<br>
|
||
Achtung: Nur bei einem lokal definierter Text2Speech Instanz möglich!
|
||
<ul>
|
||
<li>Google<br>
|
||
Google Sprachengine. Voraussetzung: Aktive Internetverbindung<br>
|
||
Aufgrund der Qualität ist der Einsatz der Engine empfohlen und daher der Standard.
|
||
</li>
|
||
<li>VoiceRSS<br>
|
||
VoiceRSS Sprachengine. Voraussetzung: Aktive Internetverbindung<br>
|
||
Die Nutzung ist frei bis zu 350 Anfragen pro Tag. Wenn mehr benötigt werden, ist ein Bezahlmodell wählbar.
|
||
Aufgrund der Qualität ist der Einsatz dieser Engine ebenfalls empfohlen.
|
||
Wird diese Engine benutzt, ist ein APIKey notwendig (siehe TTS_APIKey)
|
||
</li>
|
||
<li>ESpeak<br>
|
||
ESpeak Sprachengine. Voraussetzung: Installation von Espeak und lame<br>
|
||
eSpeak ist ein Open-Source-Software-Sprachsynthesizer für Englisch und andere Sprachen.
|
||
Die Qualitä ist schlechter als die der Google Engine<br>
|
||
</li>
|
||
<li>SVOX-pico<br>
|
||
SVOX-Pico TTS-Engine (aus dem AOSP). Voraussetzung: Installation von SVOX-Pico and lame<br>
|
||
Die Sprachengine sowie <code>lame</code> müssen installiert sein:<br>
|
||
<code>sudo apt-get install libttspico-utils lame</code><br><br>
|
||
Für ARM/Raspbian sind die <code>libttspico-utils</code> leider nicht verfügbar,<br>
|
||
deswegen müsste man diese selbst kompilieren oder das vorkompilierte Paket aus <a target="_blank" href"http://www.robotnet.de/2014/03/20/sprich-freund-und-tritt-ein-sprachausgabe-fur-den-rasberry-pi/">dieser Anleitung</a> verwenden, in aller Kürze:<br>
|
||
<code>sudo apt-get install libpopt-dev lame</code><br>
|
||
<code>cd /tmp</code><br>
|
||
<code>wget http://www.dr-bischoff.de/raspi/pico2wave.deb</code><br>
|
||
<code>sudo dpkg --install pico2wave.deb</code>
|
||
</li>
|
||
<li>Amazon-Polly<br>
|
||
Amazon Polly Sprachengine. Voraussetzung: Aktive Internetverbindung und Perl Package Paws<br>
|
||
Amazon-Dienst, der Text in lebensechte Sprache umwandelt. Ein AWS Konto und ein Polly AWS User müssen verfügbar sein<br>
|
||
<code>cpan paws</code><br>
|
||
Die Zugangsdaten zum eigenen AWS Konto müssen unter ~/.aws/credentials liegen. <br>
|
||
<code>[default]
|
||
aws_secret_access_key = xxxxxxxxxxxxxxxxxxxxxx
|
||
aws_access_key_id = xxxxxxxxxxxxxxx
|
||
</code>
|
||
</li>
|
||
<li>maryTTS<br>
|
||
<a target="_blank" href"https://github.com/marytts/marytts">maryTTS</a> oder <a target="_blank" href"https://github.com/MycroftAI/mimic3">Mimic 3</a> Sprachsynthesizer, der betr. Server sowie lame muss separat installiert werden. Beides sind open source Lösungen für English und andere Sprachen.
|
||
Über das Attribut <a href="#Text2Speech-attr-TTS_User">TTS_User</a> können ergänzende Angaben zu Server, Port und verwendeten Stimme etc. gemacht werden.<br>
|
||
</li>
|
||
</ul>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_Language"></a><li>TTS_Language<br>
|
||
Auswahl verschiedener Standardsprachen.
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_Language_Custom"></a><li>TTS_Language_Custom<br>
|
||
Möchte man eine Sprache und Stimme abweichend der Standardsprachen verwenden, so kann man diese hier eintragen. <br>
|
||
Die Definition ist abhängig der verwendeten Sprachengine. Dieses Attribut überschreibt ein ev. vorhandenes TTS_Langugae Attribut.<br>
|
||
Siehe in die jeweilige API Referenz
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_APIKey"></a><li>TTS_APIKey<br>
|
||
Wenn VoiceRSS genutzt wird, ist ein APIKey notwendig. Um diesen zu erhalten ist eine vorherige
|
||
Registrierung notwendig. Anschließend erhält man den APIKey <br>
|
||
http://www.voicerss.org/registration.aspx <br>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_User"></a><li>TTS_User<br>
|
||
Derzeit nur für maryTTS (bzw. Mimic 3) genutzt. Falls eine Sprachengine zusätzlich zum APIKey einen Usernamen im Request verlangt.<br>
|
||
<p>(Vollständiges) Beispiel für maryTTS (die angegebenen Werte entsprechen den defaults):</p>
|
||
<p><code>attr t2s TTS_User host=127.0.0.1 port=59125 lang=de_DE voice=de_DE/thorsten_low</code></p>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_CacheFileDir"></a><li>TTS_CacheFileDir<br>
|
||
Optional: Die per Google geladenen Sprachbausteine werden in diesem Verzeichnis zur Wiederverwendung abgelegt.
|
||
Es findet zurzeit keine automatisierte Löschung statt.<br>
|
||
Default: <i>cache/</i><br>
|
||
Achtung: Nur bei einer lokal definierten Text2Speech-Instanz möglich!
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_UseMP3Wrap"></a><li>TTS_UseMP3Wrap<br>
|
||
Optional: Für eine flüssige Sprachausgabe ist es zu empfehlen, die einzelnen vorher
|
||
geladenen Sprachbausteine zu einem einzelnen Sprachbaustein zusammenfassen zu lassen bevor dieses per
|
||
Mplayer ausgegeben werden. Dazu muss Mp3Wrap installiert werden.<br>
|
||
<code>apt-get install mp3wrap</code><br>
|
||
Achtung: Nur bei einer lokal definierten Text2Speech-Instanz möglich!
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_MplayerCall"></a><li>TTS_MplayerCall<br>
|
||
Optional: Angabe des Systemaufrufs für einen alternativen Player. Wird der Aufruf gesetzt,<br>
|
||
können folgende Templates genutzt werden: <br>
|
||
<ul>
|
||
<li>{device}</li>
|
||
<li>{volume}</li>
|
||
<li>{volumeadjust}</li>
|
||
<li>{file}</li>
|
||
<li>{options}</li>
|
||
</ul>
|
||
{options} werden als Text in Klammern bei der Ausführung von set gesetzt, um beispielsweise spezielle
|
||
Parameter für jeden Aufruf separat zu setzen<br>
|
||
Beispiel: <code>set myTTS tts [192.168.0.1:7000] Das ist mein Text</code><br><br>
|
||
|
||
Beispiel der Definition:<br>
|
||
<code>attr myTTS TTS_MplayerCall sudo /usr/bin/mplayer</code><br>
|
||
<code>attr myTTS TTS_MplayerCall AUDIODEV={device} play -q -v {volume} {file}</code><br>
|
||
<code>attr myTTS TTS_MplayerCall player {file} {options}</code><br>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_SentenceAppendix"></a><li>TTS_SentenceAppendix<br>
|
||
Optional: Angabe einer mp3-Datei die mit jeder Sprachausgabe am Ende ausgegeben wird.<br>
|
||
Voraussetzung ist die Nutzung von MP3Wrap. Die Sprachbausteine müssen bereits als mp3 im
|
||
CacheFileDir vorliegen.
|
||
Beispiel: <code>silence.mp3</code>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_FileMapping"></a><li>TTS_FileMapping<br>
|
||
Angabe von möglichen MP3-Dateien mit deren Template-Definition. Getrennt durch Leerzeichen.
|
||
Die Template-Definitionen können in den per <i>tts</i> übergebenen Sprachbausteinen verwendet werden
|
||
und müssen mit einem beginnenden und endenden Doppelpunkt angegeben werden.
|
||
Die Dateien müssen im Verzeichnis <i>TTS_FileTemplateDir</i> gespeichert sein.<br>
|
||
<code>attr myTTS TTS_FileMapping ring:ringtone.mp3 beep:MyBeep.mp3</code><br>
|
||
<code>set MyTTS tts Achtung: hier kommt mein Klingelton :ring: War der laut?</code>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_FileTemplateDir"></a><li>TTS_FileTemplateDir<br>
|
||
Verzeichnis, in dem die per <i>TTS_FileMapping</i> und <i>TTS_SentenceAppendix</i> definierten
|
||
MP3-Dateien gespeichert sind.<br>
|
||
Optional, Default: <code>cache/templates</code>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_VolumeAdjust"></a><li>TTS_VolumeAdjust<br>
|
||
Anhebung der Grundlautstärke zur Anpassung an die angeschlossenen Lautsprecher. <br>
|
||
Default: 110<br>
|
||
<code>attr myTTS TTS_VolumeAdjust 400</code><br>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_noStatisticsLog"></a><li>TTS_noStatisticsLog<br>
|
||
<b>1</b>, verhindert das Loggen von Statistikdaten in DbLog Geräten. Default ist <b>0</b><br>
|
||
Hinweis: Das Logging ist wichtig um alte, lang nicht genutzte Cachedateien automatisiert zu löschen.
|
||
Wird die Option hier aktiviert, muss sich der Nutzer selbst darum küümmern.
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_speakAsFastAsPossible"></a><li>TTS_speakAsFastAsPossible<br>
|
||
Es wird versucht, so schnell als möglich eine Sprachausgabe zu erzielen. Bei Sprachbausteinen
|
||
die nicht bereits lokal vorliegen, ist eine kurze Pause wahrnehmbar. Dann wird der benötigte
|
||
Sprachbaustein nachgeladen. Liegen alle Sprachbausteine im Cache vor, so hat dieses Attribut keine Auswirkung.<br>
|
||
Attribut nur verfügbar bei einer lokalen oder Server Instanz
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_OutputFile"></a><li>TTS_OutputFile<br>
|
||
Angabe eines fixen Dateinamens als mp3 Output. Das Attribut ist nur relevant in Verbindung mit TTS_UseMP3Wrap.<br>
|
||
Wenn ein Dateiname angegeben wird, so wird zusätzlich TTS_CacheFileDir beachtet. Bei einer absoluten Pfadangabe
|
||
muss der Dateipfad durch FHEM schreibbar sein.<br>
|
||
<code>attr myTTS TTS_OutputFile output.mp3</code><br>
|
||
<code>attr myTTS TTS_OutputFile /media/miniDLNA/output.mp3</code><br>
|
||
</li>
|
||
|
||
<a id="Text2Speech-attr-TTS_RemotePlayerCall"></a><li>TTS_RemotePlayerCall<br>
|
||
Die Text2Speech Geräte stellen eine URL bereit, die auf die letzte erzeugte mp3 Datei zeigt:<br>
|
||
<code><protocol>://<fhem server name or ip>:<fhem port>/fhem/Text2Speech/<device name>/last.mp3</code><br>
|
||
Wenn dieses Attribut den Aufruf eines Remoteplayers enthält, wird er nach dem Erzeugen der letzten mp3 Datei ausgeführt.<br>
|
||
Beispiel zum Abspielen einer Datei auf einem Smartphone oder Tablet mit Fully Kiosk Browser App.<br>
|
||
<code>attr <device name> TTS_RemotePlayerCall GetFileFromURL('<protocol>://<remote player name or ip>:2323/?cmd=playSound&url=<protocol>://<fhem server name or ip>:<fhem port>/fhem/Text2Speech/<device name>/last.mp3&loop=false&password=<password>')</code><br>
|
||
</li>
|
||
|
||
<li><a href="#readingFnAttributes">readingFnAttributes</a>
|
||
</li><br>
|
||
|
||
<li><a href="#disable">disable</a><br>
|
||
Wird das Attribut aktiviert, wird die Audioausgabe deaktiviert.<br>
|
||
Mögliche Werte: 0 => nicht deaktiviert , 1 => deaktiviert<br>
|
||
Standardwert ist 0 (nicht deaktiviert)<br><br>
|
||
</li>
|
||
|
||
<li><a href="#verbose">verbose</a><br>
|
||
<b>4:</b> Alle Zwischenschritte der Verarbeitung werden ausgegeben<br>
|
||
<b>5:</b> Zusätzlich werden auch die Meldungen von Mplayer und Mp3Wrap ausgegeben
|
||
</li>
|
||
|
||
</ul><br>
|
||
|
||
<a id="Text2Speech-examples"></a>
|
||
<h4>Beispiele</h4>
|
||
<ul>
|
||
<code>define TTS_EG_WZ Text2Speech hw=/dev/snd/controlC3</code><br>
|
||
<code>attr TTS_EG_WZ TTS_Language Deutsch</code><br>
|
||
<code>attr TTS_EG_WZ TTS_MplayerCall /usr/bin/mplayer</code><br>
|
||
<code>attr TTS_EG_WZ TTS_Ressource Amazon-Polly</code><br>
|
||
<code>attr TTS_EG_WZ TTS_UseMP3Wrap 1</code><br><br>
|
||
<code>set MyTTS tts <speak>Mary had a little lamb.</speak></code>
|
||
<br>
|
||
<code>define MyTTS Text2Speech hw=0.0</code><br>
|
||
<code>set MyTTS tts Die Alarmanlage ist bereit.</code><br>
|
||
<code>set MyTTS tts :beep.mp3:</code><br>
|
||
<code>set MyTTS tts :mytemplates/alarm.mp3:Die Alarmanlage ist bereit.:ring.mp3:</code>
|
||
<br>
|
||
Beispiel MaryTTS und SSML: <br>
|
||
<code>
|
||
define T2S Text2Speech default
|
||
attr T2S TTS_MplayerCall /usr/bin/mplayer
|
||
attr T2S TTS_Ressource maryTTS
|
||
attr T2S TTS_User host=192.168.100.1 port=59125 lang=de_DE voice=de_DE/thorsten_low ssml=1
|
||
set T2S tts '<voice name="de_DE/m-ailabs_low#rebecca_braunert_plunkett">Das ist ein Test in deutsch </voice><voice name="en_US/vctk_low#p236">and this is an test in english.</voice>'
|
||
</code>
|
||
</ul>
|
||
|
||
=end html_DE
|
||
=cut
|