> ## Documentation Index
> Fetch the complete documentation index at: https://www.getmaxim.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# StreamManager

> Stream_Manager utilities for anthropic ai model integration and logging utilities.

[View module source on GitHub](https://github.com/maximhq/maxim-py/blob/main/maxim/logger/anthropic/stream_manager.py)

## [TextStreamWrapper](/sdk/python/references/logger/anthropic/stream_manager)

```python theme={null}
class TextStreamWrapper()
```

Wrapper for text stream events with callback support.

This class wraps a text stream to provide callback functionality
for processing text events. It implements the iterator protocol
to allow for seamless iteration over text content.

**Attributes**:

| Name        | Description                                 |
| ----------- | ------------------------------------------- |
| `_wrapper`  | The underlying stream wrapper.              |
| `_callback` | Callback function to process stream events. |

#### \_\_init\_\_

```python theme={null}
def __init__(wrapper, callback)
```

Initialize the text stream wrapper.

**Arguments**:

| Name       | Description                                    |
| ---------- | ---------------------------------------------- |
| `wrapper`  | The underlying stream wrapper to delegate to.  |
| `callback` | Callback function to be called for each event. |

#### \_\_iter\_\_

```python theme={null}
def __iter__()
```

Return the iterator object.

**Returns**:

| Name                                                                          | Description           |
| ----------------------------------------------------------------------------- | --------------------- |
| `[TextStreamWrapper](/sdk/python/references/logger/anthropic/stream_manager)` | Self as the iterator. |

#### \_\_next\_\_

```python theme={null}
def __next__()
```

Get the next text content from the stream.

This method processes the next event from the stream, calls the
callback function, and returns the text content if available.

**Returns**:

| Name  | Description                                                  |
| ----- | ------------------------------------------------------------ |
| `str` | The text content from the event, or empty string if no text. |

**Raises**:

* `StopIteration` - When the stream is exhausted.

## [StreamWrapper](/sdk/python/references/logger/anthropic/stream_manager)

```python theme={null}
class StreamWrapper(MessageStreamManager)
```

Wrapper for Anthropic MessageStreamManager with callback support.

This class wraps the Anthropic MessageStreamManager to provide callback
functionality for processing stream events. It maintains compatibility
with the original API while adding the ability to execute custom callbacks
for each stream event.

The wrapper implements the context manager protocol and iterator protocol
to provide seamless integration with existing code.

**Attributes**:

| Name        | Type                   | Description                              |
| ----------- | ---------------------- | ---------------------------------------- |
| `_mgr`      | *MessageStreamManager* | The underlying stream manager.           |
| `_callback` | *Callable*             | Callback function for processing events. |
| `_stream`   | *MessageStream*        | The active stream instance.              |

#### \_\_init\_\_

```python theme={null}
def __init__(mgr: MessageStreamManager,
             callback: Callable[[MessageStreamEvent], None]) -> None
```

Initialize the stream wrapper.

**Arguments**:

| Name       | Type                                     | Description                            |
| ---------- | ---------------------------------------- | -------------------------------------- |
| `mgr`      | *MessageStreamManager*                   | The underlying stream manager to wrap. |
| `callback` | *Callable\[\[MessageStreamEvent], None]* | Callback function                      |

to be called for each stream event.

**Notes**:

Does not call super().**init**() since we're wrapping another manager.

#### \_\_enter\_\_

```python theme={null}
def __enter__() -> MessageStream
```

Enter the stream context and return a callback-enabled stream.

This method creates a custom MessageStream that wraps the original
stream to add callback functionality for each event.

**Returns**:

| Name            | Description                              |
| --------------- | ---------------------------------------- |
| `MessageStream` | A stream instance with callback support. |

#### \_\_exit\_\_

```python theme={null}
def __exit__(exc_type: Union[type[BaseException], None],
             exc: Union[BaseException, None], exc_tb: Union[TracebackType,
                                                            None]) -> None
```

Exit the stream context.

This method delegates the context exit to the underlying stream manager
to ensure proper cleanup and resource management.

**Arguments**:

| Name       | Description                                      |
| ---------- | ------------------------------------------------ |
| `exc_type` | The exception type if an exception occurred.     |
| `exc`      | The exception instance if an exception occurred. |
| `exc_tb`   | The traceback if an exception occurred.          |

#### \_\_iter\_\_

```python theme={null}
def __iter__()
```

Return the iterator object.

If no stream is active, this method will automatically enter
the context to create one.

**Returns**:

| Name                                                                      | Description           |
| ------------------------------------------------------------------------- | --------------------- |
| `[StreamWrapper](/sdk/python/references/logger/anthropic/stream_manager)` | Self as the iterator. |

#### \_\_next\_\_

```python theme={null}
def __next__()
```

Get the next event from the stream.

If no stream is active, this method will automatically enter
the context to create one.

**Returns**:

| Name                 | Description                     |
| -------------------- | ------------------------------- |
| `MessageStreamEvent` | The next event from the stream. |

**Raises**:

* `StopIteration` - When the stream is exhausted.

#### \_\_getattr\_\_

```python theme={null}
def __getattr__(name: str)
```

Delegate attribute access to the wrapped manager.

This method ensures that any attributes or methods not explicitly
defined in this wrapper are delegated to the underlying stream manager.

**Arguments**:

| Name   | Type  | Description                   |
| ------ | ----- | ----------------------------- |
| `name` | *str* | The attribute name to access. |

**Returns**:

| Name  | Description                                      |
| ----- | ------------------------------------------------ |
| `Any` | The attribute value from the underlying manager. |