public static void start(IModHelper h)
{
Helper = h;
currentText = "";
tmppath = Path.Combine(Path.Combine(Environment.CurrentDirectory, "Content"), "TTS");
ensureFolderStructureExists(Path.Combine(tmppath, "speech.mp3"));
pc = AWSHandler.getPollyClient();
currentVoice = VoiceId.Amy;
lastText = "";
lastDialog = "";
lastHud = "";
speak = false;
runSpeech = true;
currentCulture = CultureInfo.CreateSpecificCulture("en-us");
culturelang = "en";
installedVoices = new List <string>();
setupVoices();
setVoice("default");
speechThread = new Thread(t2sOut);
speechThread.Start();
GameEvents.QuarterSecondTick += GameEvents_QuarterSecondTick;
MenuEvents.MenuClosed += MenuEvents_MenuClosed;
//ControlEvents.KeyPressed += ControlEvents_KeyPressed;
}
public byte[] SynthesizeSpeech(VoiceId voiceId, string text, OutputFormat format = null, Engine engine = null)
{
if (format == null)
{
format = OutputFormat.Mp3;
}
var request = new SynthesizeSpeechRequest();
request.OutputFormat = format;
request.VoiceId = voiceId;
// Prefer neural voices.
if (engine != null)
{
request.Engine = engine;
}
request.Text = text;
request.TextType = TextType.Ssml;
var ms = new MemoryStream();
var response = Client.SynthesizeSpeech(request);
var bufferSize = 2 * 1024;
var buffer = new byte[bufferSize];
int readBytes;
var inputStream = response.AudioStream;
while ((readBytes = inputStream.Read(buffer, 0, bufferSize)) > 0)
{
ms.Write(buffer, 0, readBytes);
}
var bytes = ms.ToArray();
ms.Dispose();
return(bytes);
}
public static void setVoice(string name, bool female = true)
{
speakerName = name;
string t = PelicanTTSMod.i18n.Get(name);
if (PelicanTTSMod.i18n.LocaleEnum == LocalizedContentManager.LanguageCode.en && PelicanTTSMod.config.Voices.ContainsKey(name))
{
t = PelicanTTSMod.config.Voices[name].Voice;
}
if (t.ToString() == "")
{
t = PelicanTTSMod.i18n.Get("default_" + (female ? "female" : "male"));
}
if (VoiceId.FindValue(t) is VoiceId vId1)
{
currentVoice = vId1;
}
else if (VoiceId.FindValue(PelicanTTSMod.i18n.Get("default")) is VoiceId vId2)
{
currentVoice = vId2;
}
else
{
speakerName = "default";
currentVoice = VoiceId.Salli;
}
}
public async Task getVoice(CommandContext ctx)
{
var ActualVoice = VoiceId.FindValue(Settings.PSettings.PollyVoice);
await ctx.RespondAsync("Current voice: " + ActualVoice);
await ctx.Message.DeleteAsync();
}
public override int GetHashCode()
{
int hash = 1;
if (HasVoiceId)
{
hash ^= VoiceId.GetHashCode();
}
if (HasToken)
{
hash ^= Token.GetHashCode();
}
if (HasUrl)
{
hash ^= Url.GetHashCode();
}
if (HasJoinType)
{
hash ^= JoinType.GetHashCode();
}
if (HasMuteReason)
{
hash ^= MuteReason.GetHashCode();
}
if (_unknownFields != null)
{
hash ^= _unknownFields.GetHashCode();
}
return(hash);
}
public async Task UtterInStream(string text, VoiceId voice, Action <byte[], int> dataAvailable)
{
var req = new SynthesizeSpeechRequest();
req.VoiceId = voice;
req.TextType = TextType.Text;
req.Text = text;
req.OutputFormat = OutputFormat.Pcm;
req.SampleRate = "16000";
var response = await polly.SynthesizeSpeechAsync(req);
using (MemoryStream ms = new MemoryStream())
{
response.AudioStream.CopyTo(ms);
ms.Flush();
ms.Position = 0;
var buffer = new byte[1 * 640];
int bytesRead;
while ((bytesRead = await ms.ReadAsync(buffer, 0, buffer.Length)) > 0)
{
dataAvailable(buffer, bytesRead);
}
;
}
}
public static void setVoice(string name, bool female = true)
{
speakerName = name;
string t = PelicanTTSMod.i18n.Get(name);
if (t.ToString() == "")
{
t = PelicanTTSMod.i18n.Get("default_" + (female ? "female" : "male"));
}
if (VoiceId.FindValue(t) is VoiceId vId1)
{
currentVoice = vId1;
}
else if (VoiceId.FindValue(PelicanTTSMod.i18n.Get("default")) is VoiceId vId2)
{
currentVoice = vId2;
}
else
{
speakerName = "default";
currentVoice = VoiceId.Salli;
}
}
public async Task <string> Utter(string text, string dir, VoiceId voice)
{
string speaker = "Chatbot";
ConsoleLogger.WriteLine(speaker, $"{text}");
string fileName = (text + "-" + voice).GetMd5Hash() + ".mp3";
string recordBasePath = $"wwwroot{Path.DirectorySeparatorChar}voice";
string filePath = $"{dir}{Path.DirectorySeparatorChar}{recordBasePath}{Path.DirectorySeparatorChar}{fileName}";
if (File.Exists(filePath))
{
return($"/voice/{fileName}");
}
SynthesizeSpeechRequest sreq = new SynthesizeSpeechRequest();
sreq.Text = text;
sreq.OutputFormat = OutputFormat.Mp3;
sreq.VoiceId = voice;
SynthesizeSpeechResponse sres = await polly.SynthesizeSpeechAsync(sreq);
using (FileStream fileStream = File.Create(filePath))
{
sres.AudioStream.CopyTo(fileStream);
fileStream.Flush();
fileStream.Close();
}
return($"/voice/{fileName}");
}
public async Task JokeDump(CommandContext ctx)
{
var client = new RestClient("https://icanhazdadjoke.com");
var request = new RestRequest();
request.Method = Method.GET;
request.AddHeader("Accept", "text/plain");
var response = client.Execute(request);
var ActualVoice = VoiceId.FindValue(Settings.PSettings.PollyVoice);
// resolve a track from youtube
//var myTrack = await audioService.GetTrackAsync("The noob song", SearchMode.YouTube);
var SpeechResponse = await Program.Polly.SynthesizeSpeechAsync(new SynthesizeSpeechRequest
{
Engine = Engine.Neural,
LanguageCode = LanguageCode.EnUS,
OutputFormat = OutputFormat.Mp3,
SampleRate = "24000",
TextType = TextType.Text,
Text = response.Content,
VoiceId = ActualVoice
});
var g = Guid.NewGuid();
string path = $@"{g}.Mp3";
//FileStream f = new IsolatedStorageFileStream(path, FileMode.CreateNew, FileAccess.ReadWrite);
//await SpeechResponse.AudioStream.CopyToAsync(f);
await ctx.RespondWithFileAsync(DateTime.UtcNow.ToString("F") + ".Mp3", SpeechResponse.AudioStream);
await ctx.Message.DeleteAsync();
//await ctx.Message.CreateReactionAsync(DiscordEmoji.FromName(ctx.Client, ":white_check_mark:"));
}
private void RadioButton_Checked(object sender, RoutedEventArgs e)
{
var radioButton = (RadioButton)sender;
var name = radioButton.Name;
if (name == "Joanna")
{
VoiceName = VoiceId.Joanna;
}
if (name == "Salli")
{
VoiceName = VoiceId.Salli;
}
if (name == "Matthew")
{
VoiceName = VoiceId.Matthew;
}
if (name == "Justin")
{
VoiceName = VoiceId.Justin;
}
}
public string GetCurrentFilePath(string root, string ext = "wav")
{
string dirname = Path.Combine(root, FolderName);
if (!Directory.Exists(dirname))
{
return(null);
}
string path = Path.Combine(dirname, $"{VoiceId}.{ext}");
if (File.Exists(path))
{
return(path);
}
var files = Directory.GetFiles(dirname, $"{VoiceId}_*.{ext}");
if (files.Length == 1)
{
return(files[0]);
}
string voiceIdZen = VoiceId.ConvertNumHanToZen();
path = Path.Combine(dirname, $"{voiceIdZen}.{ext}");
if (File.Exists(path))
{
return(path);
}
files = Directory.GetFiles(dirname, $"{voiceIdZen}_*.{ext}");
if (files.Length == 1)
{
return(files[0]);
}
return(null);
}
public static void start(IModHelper h)
{
Helper = h;
currentText = "";
tmppath = Path.Combine(Helper.DirectoryPath, "TTS");
if (!Directory.Exists(tmppath))
{
Directory.CreateDirectory(tmppath);
}
if (pc == null)
{
pc = AWSHandler.getPollyClient();
}
currentVoice = VoiceId.Salli;
lastText = "";
lastDialog = "";
lastHud = "";
speak = false;
runSpeech = true;
setVoice("default");
speechThread = new Thread(t2sOut);
speechThread.Start();
h.Events.GameLoop.UpdateTicked += OnUpdateTicked;
h.Events.Display.MenuChanged += OnMenuChanged;
}
/// <summary>
/// Get voice id from string name.
/// List of voices available: https://docs.aws.amazon.com/polly/latest/dg/voicelist.html
/// </summary>
/// <param name="voice">Voice. Default is Joanna in English, US.</param>
/// <returns></returns>
static private VoiceId VoiceIdFromString(string voice = "")
{
VoiceId voiceId = VoiceId.Joanna; // Default
switch (voice)
{
// English, US
case "Salli": voiceId = VoiceId.Salli; break; // Salli, Female
case "Kimberly": voiceId = VoiceId.Kimberly; break; // Kimberly, Female
case "Kendra": voiceId = VoiceId.Kendra; break; // Kendra, Female
// case "Joanna": voiceId = VoiceId.Joanna; break; // Joanna, Female
case "Ivy": voiceId = VoiceId.Ivy; break; // Ivy, Female
case "Matthew": voiceId = VoiceId.Matthew; break; // Matthew, Male
case "Justin": voiceId = VoiceId.Justin; break; // Justin, Male
case "Joey": voiceId = VoiceId.Joey; break; // Joey, Male
// English, British
case "Emma": voiceId = VoiceId.Emma; break; // Emma, Female
case "Amy": voiceId = VoiceId.Amy; break; // Amy, Female
case "Brian": voiceId = VoiceId.Brian; break; // Brian, Female
}
return(voiceId);
}
public PollySpeechSynthesizer(IAmazonPolly amazonPolly, IOptions <PollySpeechSynthesizerConfiguration> configOptions)
{
if (configOptions != null && configOptions.Value != null)
{
voiceId = VoiceId.FindValue(configOptions.Value.VoiceId);
}
_amazonPolly = amazonPolly;
}
private void Speak(string message, VoiceId voiceId, AmazonPollyClient pollyClient)
{
var audioConfiguration = configurationManager.LoadConfiguration <AudioConfiguration>();
audioConfiguration.InitializeConfiguration();
const int volume = 100;
if (pollyClient == null)
{
pollyClient = GetPollyClient();
}
if (voiceId == null)
{
voiceId = pollyClient.DescribeVoices(new Amazon.Polly.Model.DescribeVoicesRequest {
LanguageCode = "de-DE"
}).Voices.FirstOrDefault()?.Id;
if (voiceId == null)
{
return;
}
}
string ttsPath = Path.Combine(soundPathProvider.Path, "tts");
if (!Directory.Exists(ttsPath))
{
Directory.CreateDirectory(ttsPath);
}
var synthesizeResponse = pollyClient.SynthesizeSpeech(new Amazon.Polly.Model.SynthesizeSpeechRequest()
{
OutputFormat = OutputFormat.Mp3, Text = message, VoiceId = voiceId
});
string ttsFilePath = Path.Combine(ttsPath, "tts.mp3");
using var fileStream = new FileStream(ttsFilePath, FileMode.Create);
synthesizeResponse.AudioStream.CopyTo(fileStream);
synthesizeResponse.AudioStream.Close();
synthesizeResponse.AudioStream.Dispose();
using var reader = new MediaFoundationReader(ttsFilePath);
using var volumeStream = new WaveChannel32(reader);
using var outputStream = new WasapiOut(audioConfiguration.SelectedSoundCommandDevice, NAudio.CoreAudioApi.AudioClientShareMode.Shared, false, 10);
volumeStream.Volume = NormalizeVolume(volume);
outputStream.Init(volumeStream);
outputStream.Play();
Thread.Sleep(reader.TotalTime.Add(TimeSpan.FromMilliseconds(100)));
outputStream.Stop();
}
protected internal static void BuildModel(ModelBuilder modelBuilder)
{
modelBuilder.Entity <AllowedConversationVoice>(entity =>
{
entity.HasKey(nameof(ChannelId), nameof(VoiceId));
entity.Property(e => e.VoiceId).HasConversion(
v => v.ToString(),
v => VoiceId.FindValue(v)
);
});
}
public async Task speakssml(CommandContext ctx, [RemainingText] string textToSpeak)
{
try
{
var player = Program.audioService.GetPlayer <QueuedLavalinkPlayerV2>(ctx.Guild.Id)
?? await Program.audioService.JoinAsync <QueuedLavalinkPlayerV2>(ctx.Guild.Id,
ctx.Member.VoiceState.Channel.Id);
var ActualVoice = VoiceId.FindValue(Settings.PSettings.PollyVoice);
// resolve a track from youtube
//var myTrack = await audioService.GetTrackAsync("The noob song", SearchMode.YouTube);
foreach (var user in ctx.Message.MentionedUsers)
{
Console.WriteLine(user.Mention.ToString());
var DisMem = await ctx.Guild.GetMemberAsync(user.Id);
var callout = DisMem.Nickname.IsNullOrEmpty() ? DisMem.DisplayName : DisMem.Nickname;
textToSpeak = textToSpeak.Replace(user.Mention.ToString(), callout);
}
var SpeechResponse = await Program.Polly.SynthesizeSpeechAsync(new SynthesizeSpeechRequest
{
Engine = Engine.Standard,
LanguageCode = LanguageCode.EnUS,
OutputFormat = OutputFormat.Mp3,
SampleRate = "24000",
TextType = TextType.Ssml,
Text = textToSpeak,
VoiceId = ActualVoice
});
var g = Guid.NewGuid();
string path = $@"C:\temp\{g}.Mp3";
FileStream f = new FileStream(path, FileMode.CreateNew, FileAccess.ReadWrite);
await SpeechResponse.AudioStream.CopyToAsync(f);
f.Flush();
f.Close();
var track = await Program.audioService.GetTrackAsync(HttpUtility.UrlEncode(path));
// play track
await player.PlayAsync(track);
await ctx.Message.DeleteAsync();
//await ctx.Message.CreateReactionAsync(DiscordEmoji.FromName(ctx.Client, ":white_check_mark:"));
}
catch (Exception e)
{
Console.WriteLine(textToSpeak);
}
}
public async Task <Stream> SynthesizeTextToStreamAsync(IVoice voice, string text)
{
var pollyVoice = (AmazonPollyVoice)voice;
var request = new SynthesizeSpeechRequest()
{
Text = text,
VoiceId = VoiceId.FindValue(pollyVoice.VoiceId),
OutputFormat = OutputFormat.Mp3
};
var response = await _client.SynthesizeSpeechAsync(request);
return(response.AudioStream);
}
public static void setVoiceById(string id)
{
if (VoiceId.FindValue(id) is VoiceId vId1)
{
currentVoice = vId1;
}
else if (VoiceId.FindValue(PelicanTTSMod.i18n.Get("default")) is VoiceId vId2)
{
currentVoice = vId2;
}
else
{
speakerName = "default";
currentVoice = VoiceId.Salli;
}
}
public async Task Say(Engine engine, VoiceId voice, int sampleRate, float volume, IList <string> lexicons, TextSource source, string text)
{
var req = new SynthesizeSpeechRequest
{
Text = text,
VoiceId = voice,
Engine = engine,
OutputFormat = OutputFormat.Mp3,
SampleRate = sampleRate.ToString(),
LexiconNames = lexicons.Where(l => !string.IsNullOrEmpty(l)).ToList(),
TextType = TextType.Ssml,
};
SynthesizeSpeechResponse res;
try
{
res = await this.client.SynthesizeSpeechAsync(req);
}
catch (LexiconNotFoundException e)
{
if (lexicons.Any())
{
PluginLog.LogError(e, "A lexicon could not be found, retrying without any lexicons... Additional data: {@Data}", e.Data);
await Say(engine, voice, sampleRate, volume, Array.Empty <string>(), source, text);
}
else
{
PluginLog.LogError(e, "A lexicon could not be found... but we aren't using any lexicons? Additional data: {@Data}", e.Data);
}
return;
}
catch (Exception e)
{
PluginLog.LogError(e, "Synthesis request failed in {0}.", nameof(PollyClient));
return;
}
var responseStream = new MemoryStream();
await res.AudioStream.CopyToAsync(responseStream);
responseStream.Seek(0, SeekOrigin.Begin);
this.soundQueue.EnqueueSound(responseStream, source, volume);
}
protected internal static void BuildModel(ModelBuilder modelBuilder)
{
modelBuilder.Entity <Reward>(entity =>
{
entity.Property(e => e.IsConversation).HasDefaultValue(true);
entity.Property(e => e.IsSubOnly).HasDefaultValue(false);
entity.Property(e => e.Cooldown).HasDefaultValue(0);
entity.Property(e => e.DefaultPlaybackSpeed).HasDefaultValue(1.0f);
entity.Property(e => e.VoiceId).HasConversion(
v => v.ToString(),
v => VoiceId.FindValue(v)
);
entity.Property(e => e.VoiceEngine).HasConversion(
e => e.ToString(),
e => Engine.FindValue(e)
);
});
}
public async Task speakdump(CommandContext ctx, [RemainingText] string textToSpeak)
{
try
{
var ActualVoice = VoiceId.FindValue(Settings.PSettings.PollyVoice);
// resolve a track from youtube
//var myTrack = await audioService.GetTrackAsync("The noob song", SearchMode.YouTube);
foreach (var user in ctx.Message.MentionedUsers)
{
Console.WriteLine(user.Mention.ToString());
var DisMem = await ctx.Guild.GetMemberAsync(user.Id);
var callout = DisMem.Nickname.IsNullOrEmpty() ? DisMem.DisplayName : DisMem.Nickname;
textToSpeak = textToSpeak.Replace(user.Mention.ToString(), callout);
}
var SpeechResponse = await Program.Polly.SynthesizeSpeechAsync(new SynthesizeSpeechRequest
{
Engine = Engine.Neural,
LanguageCode = LanguageCode.EnUS,
OutputFormat = OutputFormat.Mp3,
SampleRate = "24000",
TextType = TextType.Text,
Text = textToSpeak,
VoiceId = ActualVoice
});
var g = Guid.NewGuid();
string path = $@"{g}.Mp3";
//FileStream f = new IsolatedStorageFileStream(path, FileMode.CreateNew, FileAccess.ReadWrite);
//await SpeechResponse.AudioStream.CopyToAsync(f);
await ctx.RespondWithFileAsync(DateTime.UtcNow.ToString("F") + ".Mp3", SpeechResponse.AudioStream);
await ctx.Message.DeleteAsync();
//await ctx.Message.CreateReactionAsync(DiscordEmoji.FromName(ctx.Client, ":white_check_mark:"));
}
catch (Exception e)
{
Console.WriteLine(e);
throw;
}
}
public void Voice(VoiceId v)
{
if (v == VoiceId.Get)
{
audioSource.clip = getVoice;
audioSource.Play();
}
if (v == VoiceId.Jump)
{
audioSource.clip = jumpVoice;
audioSource.Play();
}
if (v == VoiceId.Damage)
{
int n;
n = Random.Range(0, 4); // Random.Range(0, damageVoice.Length) と書くと安全
audioSource.clip = damageVoice[n];
audioSource.Play();
}
}
public async Task Joke(CommandContext ctx)
{
var client = new RestClient("https://icanhazdadjoke.com");
var request = new RestRequest();
request.Method = Method.GET;
request.AddHeader("Accept", "text/plain");
var response = client.Execute(request);
var player = Program.audioService.GetPlayer <QueuedLavalinkPlayerV2>(ctx.Guild.Id)
?? await Program.audioService.JoinAsync <QueuedLavalinkPlayerV2>(ctx.Guild.Id, ctx.Member.VoiceState.Channel.Id);
var ActualVoice = VoiceId.FindValue(Settings.PSettings.PollyVoice);
// resolve a track from youtube
//var myTrack = await audioService.GetTrackAsync("The noob song", SearchMode.YouTube);
var SpeechResponse = await Program.Polly.SynthesizeSpeechAsync(new SynthesizeSpeechRequest
{
Engine = Engine.Neural,
LanguageCode = LanguageCode.EnUS,
OutputFormat = OutputFormat.Mp3,
SampleRate = "24000",
TextType = TextType.Text,
Text = response.Content,
VoiceId = ActualVoice
});
var g = Guid.NewGuid();
string path = $@"C:\temp\{g}.Mp3";
FileStream f = new FileStream(path, FileMode.CreateNew, FileAccess.ReadWrite);
await SpeechResponse.AudioStream.CopyToAsync(f);
f.Flush();
f.Close();
var track = await Program.audioService.GetTrackAsync(HttpUtility.UrlEncode(path));
// play track
await player.PlayAsync(track);
await ctx.Message.DeleteAsync();
//await ctx.Message.CreateReactionAsync(DiscordEmoji.FromName(ctx.Client, ":white_check_mark:"));
}
public async Task <byte[]> Speech(string query, string languageCode) // Получение аудио-файла озвученного слова (через api Amazon Polly)
{
var lang = _languageService.GetLanguage(languageCode);
var client = new AmazonPollyClient(_awsPollyOptions.Value.awsAccessKeyId, _awsPollyOptions.Value.awsSecretAccessKey, RegionEndpoint.USEast2);
var synthesizeSpeechRequest = new SynthesizeSpeechRequest()
{
OutputFormat = OutputFormat.Mp3,
LanguageCode = lang.LanguageCode,
VoiceId = VoiceId.FindValue(lang.VoiceId),
Text = query
};
var synthesizeSpeechResponse = await client.SynthesizeSpeechAsync(synthesizeSpeechRequest);
var inputStream = synthesizeSpeechResponse.AudioStream;
MemoryStream mem = new MemoryStream();
inputStream.CopyTo(mem);
return(mem.ToArray());
}
public async Task <SynthesizeSpeechResponse> Synthesize(string text, VoiceId voiceId, Engine engine, TextType textType = null)
{
if (text.Trim().ToLowerInvariant().StartsWith("ssml: "))
{
text = text.Trim()[6..];
public static void setVoice(string name)
{
speakerName = name;
NPC speaker = Game1.getCharacterFromName(name);
if (speaker != null && speaker.Gender == 0)
{
if (speaker.Age == 0)
{
currentVoice = VoiceId.Joey;
}
else
{
currentVoice = VoiceId.Joey;
}
}
if (speaker != null && speaker.Gender == 1)
{
if (speaker.Age == 0)
{
currentVoice = VoiceId.Kendra;
}
else
{
currentVoice = VoiceId.Salli;
}
}
switch (name)
{
case "Elliot": currentVoice = VoiceId.Geraint; break;
case "Sam": currentVoice = VoiceId.Russell; break;
case "Emily": currentVoice = VoiceId.Emma; break;
case "Haley": currentVoice = VoiceId.Emma; break;
case "Harvey": currentVoice = VoiceId.Matthew; break;
case "George": currentVoice = VoiceId.Brian; break;
case "Linus": currentVoice = VoiceId.Brian; break;
case "Lewis": currentVoice = VoiceId.Brian; break;
case "Governor": currentVoice = VoiceId.Brian; break;
case "Grandpa": currentVoice = VoiceId.Brian; break;
case "Clint": currentVoice = VoiceId.Brian; break;
case "Willy": currentVoice = VoiceId.Brian; break;
case "Wizard": currentVoice = VoiceId.Geraint; break;
case "Pierre": currentVoice = VoiceId.Matthew; break;
case "Gunther": currentVoice = VoiceId.Brian; break;
case "Govenor": currentVoice = VoiceId.Brian; break;
case "Marlon": currentVoice = VoiceId.Brian; break;
case "Morris": currentVoice = VoiceId.Geraint; break;
case "Mister Qi": currentVoice = VoiceId.Geraint; break;
case "Gil": currentVoice = VoiceId.Brian; break;
case "Penny": currentVoice = VoiceId.Amy; break;
case "Evelyn": currentVoice = VoiceId.Amy; break;
case "Jas": currentVoice = VoiceId.Ivy; break;
case "Jodi": currentVoice = VoiceId.Nicole; break;
case "Marnie": currentVoice = VoiceId.Kimberly; break;
case "Pam": currentVoice = VoiceId.Kimberly; break;
case "Sandy": currentVoice = VoiceId.Raveena; break;
case "Vincent": currentVoice = VoiceId.Justin; break;
case "default": currentVoice = VoiceId.Salli; break;
default: break;
}
}
private void femaleVoice_CheckedChanged(object sender, EventArgs e)
{
_voiceId = VoiceId.Tatyana;
}
/// <summary>
/// PollyNotes-DictateFunction
///
/// This lambda function is integrated with the following API methods:
/// /notes/{id}/POST
///
/// This function does the following:
///
/// 1. Takes a JSON payload from API gateway and converts it into a DictateRequest
/// 2. Queries DynamoDB for the note from the request to fetch the note text
/// 3. Calls the Polly synthensize_speech API to convert text to speech
/// 4. Stores the resulting audio in an MP3 file
/// 5. Uploads the MP3 file to S3
/// 6. Creates a pre-signed URL for the MP3 file
/// 7. Returns the pre-signed URL to API Gateway
/// </summary>
/// <param name="request">DictateRequest containing the voiceId and the note to create an mp3 file from</param>
/// <param name="context">Lambda context</param>
/// <returns>string of the URL for the pre-signed mp3 file from S3</returns>
public string FunctionHandler(DictateRequest request, ILambdaContext context)
{
// The note object contains the voiceId, userId and noteId from the /notes/{id}/POST
// {
// "voiceId": "...",
// "note": {
// "userId": "...",
// "noteId": "..."
// }
// }
Console.WriteLine("Initiating PollyNotes-DictateFunction...");
Console.WriteLine("DictateRequest received: " + JsonConvert.SerializeObject(request));
// Get the name of the bucketName from the environment variable MP3_BUCKET_NAME
string bucketName = Environment.GetEnvironmentVariable("MP3_BUCKET_NAME");
// Create the DynamoDB client and the Context for Object Persistence Model
AmazonDynamoDBClient client = new AmazonDynamoDBClient();
DynamoDBContext ddbcontext = new DynamoDBContext(client);
// Use the LoadAsync method to fetch all of the attributes of the note from the request from DynamoDB and wait
// This is really to get the note attribute from the userId and noteId of the request
var ddbTask = ddbcontext.LoadAsync(request.note);
ddbTask.Wait();
// The result will be stored in note
Note note;
// If there are no Exceptions
if (ddbTask.Exception == null)
{
Console.WriteLine("Successfully executed LoadAsync with userId: " + request.note.userId + " and noteId: " + request.note.noteId);
// Set the note variable to the result of the LoadAsync from DynamoDB
note = ddbTask.Result;
}
else
{
// There was an exception, log the entry data and the exception
Console.WriteLine("Unable to LoadAsync note with userId: " + request.note.userId + " and noteId: " + request.note.noteId);
Console.WriteLine(ddbTask.Exception);
// Return an empty string
return("");
}
// Invoke Polly API, which will transform text into audio using the note we fetched from DynamoDB and the voiceId from the request
var polly = new AmazonPollyClient();
SynthesizeSpeechRequest speechRequest = new SynthesizeSpeechRequest
{
OutputFormat = OutputFormat.Mp3,
Text = note.note,
VoiceId = VoiceId.FindValue(request.voiceId)
};
var pollyTask = polly.SynthesizeSpeechAsync(speechRequest);
pollyTask.Wait();
Console.WriteLine("Successfully synthesized the Note text");
// Save the audio stream returned by Amazon Polly on Lambda's temp directory '/tmp'
string path = Path.Combine(
Path.GetTempPath(),
bucketName,
request.note.userId);
string filename = Path.Combine(path, request.note.noteId + ".mp3");
Directory.CreateDirectory(path);
using (FileStream file = new FileStream(filename, FileMode.Create, System.IO.FileAccess.Write))
{
pollyTask.Result.AudioStream.CopyTo(file);
}
Console.WriteLine("Successfully saved the Polly AudioStream to " + filename);
// Upload our local file to S3
var s3 = new AmazonS3Client();
var s3Task = s3.PutObjectAsync(new PutObjectRequest
{
BucketName = bucketName,
Key = String.Format("{0}/{1}.mp3", request.note.userId, request.note.noteId),
FilePath = filename
});
s3Task.Wait();
Console.WriteLine("Successfully uploaded the MP3 file to S3");
// Generate a pre-signed URL so that we can securely access our MP3
string url = s3.GetPreSignedURL(new GetPreSignedUrlRequest
{
BucketName = bucketName,
Key = String.Format("{0}/{1}.mp3", request.note.userId, request.note.noteId),
Expires = DateTime.Now + TimeSpan.FromHours(1)
});
Console.WriteLine("Successfully generated a pre-signed URL for the MP3 file: " + url);
// Return the presigned URL to API Gateway
return(url);
}
internal static void configSay(string name, string voice, string text, int rate = -1, float pitch = -1, float volume = -1)
{
Task.Run(() =>
{
currentVoice = VoiceId.FindValue(voice);
tmppath = Path.Combine(PelicanTTSMod._helper.DirectoryPath, "TTS");
if (pc == null)
{
pc = AWSHandler.getPollyClient();
}
bool mumbling = PelicanTTSMod.config.MumbleDialogues;
string language1 = "<lang xml:lang=\"" + getLanguageCode() + "\">";
string language2 = "</lang>";
text = text.Replace("0g", "0 gold").Replace("< ", " ").Replace("` ", " ").Replace("> ", " ").Replace('^', ' ').Replace(Environment.NewLine, " ").Replace("$s", "").Replace("$h", "").Replace("$g", "").Replace("$e", "").Replace("$u", "").Replace("$b", "").Replace("$8", "").Replace("$l", "").Replace("$q", "").Replace("$9", "").Replace("$a", "").Replace("$7", "").Replace("<", "").Replace("$r", "").Replace("[", "<").Replace("]", ">");
text = language1 + text + language2;
bool neural = shouldUseNeuralEngine(voice, out string v);
if (!neural && voice != v)
{
currentVoice = VoiceId.FindValue(v);
}
bool useNeuralEngine = !mumbling && neural;
var amzeffectIn = mumbling ? "<amazon:effect phonation='soft'><amazon:effect vocal-tract-length='-20%'>" : "<amazon:auto-breaths><amazon:effect phonation='soft'>";
var amzeffectOut = mumbling ? "</amazon:effect></amazon:effect>" : "</amazon:effect></amazon:auto-breaths>";
if (mumbling)
{
text = @"<speak>" + (useNeuralEngine ? "" : amzeffectIn) + Dialogue.convertToDwarvish(text) + (useNeuralEngine ? "" : amzeffectOut) + @"</speak>";
}
else
{
text = @"<speak>" + (useNeuralEngine ? "" : amzeffectIn) + "<prosody rate='" + (rate == -1 ? PelicanTTSMod.config.Rate : rate) + "%'>" + text + @"</prosody>" + (useNeuralEngine ? "" : amzeffectOut) + "</speak>";
}
int hash = (text + (useNeuralEngine ? "-neural" : "")).GetHashCode();
if (!Directory.Exists(Path.Combine(tmppath, name)))
{
Directory.CreateDirectory(Path.Combine(tmppath, name));
}
string file = Path.Combine(Path.Combine(tmppath, name), "speech_" + currentVoice.Value + (mumbling ? "_mumble_" : "_") + hash + ".wav");
SoundEffect nextSpeech = null;
if (!File.Exists(file))
{
SynthesizeSpeechRequest sreq = new SynthesizeSpeechRequest();
sreq.Text = text;
sreq.TextType = TextType.Ssml;
sreq.OutputFormat = OutputFormat.Ogg_vorbis;
sreq.Engine = useNeuralEngine ? Engine.Neural : Engine.Standard;
sreq.VoiceId = currentVoice;
SynthesizeSpeechResponse sres = pc.SynthesizeSpeech(sreq);
using (var memStream = new MemoryStream())
{
sres.AudioStream.CopyTo(memStream);
nextSpeech = Convert(memStream, file);
}
}
using (FileStream stream = new FileStream(file, FileMode.Open))
nextSpeech = SoundEffect.FromStream(stream);
if (currentSpeech != null)
{
currentSpeech.Stop();
}
currentSpeech = nextSpeech.CreateInstance();
speak = false;
currentSpeech.Pitch = (mumbling ? 0.5f : pitch == -1 ? PelicanTTSMod.config.Voices[name].Pitch : pitch);
currentSpeech.Volume = volume == -1 ? PelicanTTSMod.config.Volume : volume;
currentSpeech.Play();
});
}
