bug fix for multimodal-dialog agent and add parameter speaker_id in asr result

[songguocola]

No description provided.

[gemini-code-assist]

Code Review

This pull request introduces robust connection management, graceful shutdown, and resource cleanup. Key changes include adding a speaker ID to sentences, implementing a graceful stop with a 30-second timeout and an idempotent close in MultiModalDialog, and enhancing OkHttpWebSocketClient to handle cancellations and retries safely. Additionally, OkHttpWebSocketClientForAudio is refactored to reduce code duplication. The reviewer recommended clearing DialogBuffer during close() in MultiModalDialog to prevent memory leaks and release buffered objects immediately.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

[gemini-code-assist]

When close() is called, the dialog session is being terminated. Any pending audio or text frames currently buffered in DialogBuffer will never be sent. To prevent memory leaks and release references to potentially large ByteBuffer objects immediately, we should clear DialogBuffer inside the synchronized block of close().

Suggested change

	// Nullify emitter to prevent further data sending
	synchronized (MultiModalDialog.this) {
	conversationEmitter = null;
	}
	// Nullify emitter and clear buffer to prevent further data sending and release memory
	synchronized (MultiModalDialog.this) {
	conversationEmitter = null;
	DialogBuffer.clear();
	}