程序包 io.agentscope.core.hook

类 TTSHook

java.lang.Object

io.agentscope.core.hook.TTSHook

所有已实现的接口:

Hook

public class TTSHook
extends Object
implements Hook

Hook for real-time Text-to-Speech synthesis during agent execution.

This hook implements "speak as you generate" by listening to streaming reasoning events and synthesizing speech in real-time.

Two Usage Modes:

• Local Playback (CLI/Desktop): Use audioPlayer for direct playback
• Server Mode (Web/SSE): Use audioCallback to return audio to frontend

Example 1: Local Playback (CLI/Testing)

 AudioPlayer player = AudioPlayer.builder().sampleRate(24000).build();

 TTSHook ttsHook = TTSHook.builder()
     .ttsModel(ttsModel)
     .audioPlayer(player)  // Local playback
     .build();
 

Example 2: Server Mode (Return to Frontend via SSE)

 TTSHook ttsHook = TTSHook.builder()
     .ttsModel(ttsModel)
     .audioCallback(audio -> {
         // Send via SSE/WebSocket to frontend
         sseEmitter.send(audio);
     })
     .build();
 

Example 3: Get Audio Stream (Reactive)

 TTSHook ttsHook = TTSHook.builder()
     .ttsModel(ttsModel)
     .build();

 // Subscribe to audio stream
 ttsHook.getAudioStream()
     .subscribe(audio -> sendToClient(audio));
 

嵌套类概要

嵌套类

修饰符和类型	类	说明
static class	TTSHook.Builder	Builder for constructing TTSHook instances.

方法概要

修饰符和类型	方法	说明
static TTSHook.Builder	builder()	Creates a new builder for TTSHook.
reactor.core.publisher.Flux<AudioBlock>	getAudioStream()	Gets the reactive audio stream.
<T extends HookEvent> reactor.core.publisher.Mono<T>	onEvent(T event)	Handle a hook event.
void	stop()	Stop the audio player and clean up resources.

从类继承的方法 java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

从接口继承的方法 io.agentscope.core.hook.Hook

priority

方法详细资料

getAudioStream

public reactor.core.publisher.Flux<AudioBlock> getAudioStream()

Gets the reactive audio stream.

Use this to subscribe to audio blocks as they are generated. This is useful for SSE/WebSocket streaming to frontend.

Example:

 ttsHook.getAudioStream()
     .map(audio -> ((Base64Source) audio.getSource()).getData())
     .subscribe(base64 -> sseEmitter.send(base64));
 

返回:

Flux of AudioBlock that emits audio as it's synthesized

onEvent

public <T extends HookEvent>
reactor.core.publisher.Mono<T> onEvent(T event)

从接口复制的说明: Hook

Handle a hook event.

This method is called for all agent execution events. Use pattern matching to handle specific event types.

Modifiable Events: For events with setters, you can modify the context and the changes will affect agent execution:

• PreReasoningEvent - Modify messages before LLM reasoning
• PostReasoningEvent - Modify reasoning results
• PreActingEvent - Modify tool parameters before execution
• PostActingEvent - Modify tool results
• PreCallEvent - Modify messages before agent starts
• PostCallEvent - Modify final agent response

Notification Events: Events without setters are read-only:

• ReasoningChunkEvent - Streaming reasoning chunks
• ActingChunkEvent - Streaming tool execution chunks
• ErrorEvent - Errors during execution

指定者:

onEvent 在接口中 Hook

类型参数:

T - The concrete event type

参数:

event - The hook event

返回:

Mono containing the potentially modified event

stop

public void stop()

Stop the audio player and clean up resources.

builder

public static TTSHook.Builder builder()

Creates a new builder for TTSHook.

返回:

a new Builder instance