英文:

python: make websocket connection to elevenlabs with websocket-client

问题 {#heading}

在 elevenlabs 中，您现在可以使用流功能，同时还可以使用输入流。但是，必须建立套接字连接。

不幸的是，我必须自己实现 WebSocket 连接，而不是使用 elevenlabs 的 Python 客户端，因为我必须使用 WebSocket 客户端库，并且 elevenlabs 客户端使用不同的 WebSocket 库。

问题在于以下代码：

data = json.loads(ws.recv())

它会永远阻塞，并且似乎永远不会收到任何数据。我做错了什么？

以下是我的代码（BOS 和 EOS 与 elevenlabs Python 客户端文档中的代码完全相同，其余部分几乎也是相同的代码）：

with closing(create_connection(url, header={"xi-api-key": "mykey"})) as ws:
                 
    # 发送流的开始
    ws.send(BOS)

    # 流式传输文本块并接收音频
    for text_chunk in text_chunker(text):
        data = dict(text=text_chunk, try_trigger_generation=True)
        ws.send(json.dumps(data))
        try:
            data = json.loads(ws.recv())  # 在这里阻塞，永远无法继续
            if data["audio"]:
                yield base64.b64decode(data["audio"])
        except TimeoutError:
            pass

    # 发送流的结束
    ws.send(EOS)

    # 接收剩余的音频
    while True:
        try:
            data = json.loads(ws.recv())
            if data["audio"]:
                yield base64.b64decode(data["audio"])
        except ws.exceptions.ConnectionClosed:
            break

使用 elevenlabs 客户端的原始代码，它使用了 websockets.sync.client 库，可以按预期工作。 英文:

In elevenlabs you can now use the stream feature also with input stream. However, a socket connection must be established for this.

Unfortunately I have to implement the websocket connection myself instead of using the elevenlabs python client, because i have to use the websocket-client library and the elevenlabs client is using a different websocket library.

The problem is, however, that

data = json.loads(ws.recv())

blocks forever and nothing ever seems to arrive. what did I do wrong?

here is my code(BOS and EOS are exactly the same as in the documentations code of elevenlabs python client and rest of it is also nearly the same code):

with closing(create_connection(url, header={"xi-api-key":"mykey")) as ws:

                # Send beginning of stream
                ws.send(BOS)

                # Stream text chunks and receive audio
                for text_chunk in text_chunker(text):
                    data = dict(text=text_chunk, try_trigger_generation=True)
                    ws.send(json.dumps(data))
                    try:
                        data = json.loads(ws.recv()) -> blocking here, never gets over it
                        if data["audio"]:
                            yield base64.b64decode(data["audio"])  # type: ignore
                    except TimeoutError:
                        pass

                # Send end of stream
                ws.send(EOS)

                # Receive remaining audio
                while True:
                    try:
                        data = json.loads(ws.recv())
                        if data["audio"]:
                            yield base64.b64decode(data["audio"])  # type: ignore
                    except ws.exceptions.ConnectionClosed:
                        break  

with the original code from the elevenlabs client , which uses the websockets.sync.client library, it works as expected

答案1 {#1}

得分: 0

使用WebSocketApp在一个线程中找到了解决方案。 英文:

found a solution by using WebSocketApp in a thread.