webscout 8.3.4__py3-none-any.whl → 8.3.5__py3-none-any.whl

webscout/AIutel.py

@@ -1,1022 +1,58 @@
-import codecs
-import json
-import re
-from typing import (
-    Any,
-    AsyncGenerator,
-    AsyncIterable,
-    Callable,
-    Dict,
-    Generator,
-    Iterable,
-    List,
-    Literal,
-    Optional,
-    Pattern,
-    Union,
-)
-
-# Expanded encoding types
-EncodingType = Literal['utf-8', 'utf-16', 'utf-32', 'ascii', 'latin1', 'cp1252', 'iso-8859-1',
-                        'iso-8859-2', 'windows-1250', 'windows-1251', 'windows-1252', 'gbk', 'big5',
-                        'shift_jis', 'euc-jp', 'euc-kr']
-
-def _compile_regexes(patterns: Optional[List[Union[str, Pattern[str]]]]) -> Optional[List[Pattern[str]]]:
-    """
-    Compile regex patterns from strings or return compiled patterns as-is.
-    
-    Args:
-        patterns: List of regex patterns as strings or compiled Pattern objects.
-        
-    Returns:
-        List of compiled Pattern objects, or None if input is None.
-        
-    Raises:
-        ValueError: If any pattern is invalid.
-    """
-    if not patterns:
-        return None
-    
-    compiled_patterns = []
-    for i, pattern in enumerate(patterns):
-        try:
-            if isinstance(pattern, str):
-                compiled_patterns.append(re.compile(pattern))
-            elif isinstance(pattern, Pattern):
-                compiled_patterns.append(pattern)
-            else:
-                raise ValueError(f"Pattern at index {i} must be a string or compiled regex pattern, got {type(pattern)}")
-        except re.error as e:
-            raise ValueError(f"Invalid regex pattern at index {i}: '{pattern}' - {str(e)}")
-    
-    return compiled_patterns
-
-def _process_chunk(
-    chunk: str,
-    intro_value: str,
-    to_json: bool,
-    skip_markers: List[str],
-    strip_chars: Optional[str],
-    yield_raw_on_error: bool,
-    error_handler: Optional[Callable[[Exception, str], Optional[Any]]] = None,
-    skip_regexes: Optional[List[Pattern[str]]] = None,
-    extract_regexes: Optional[List[Pattern[str]]] = None,
-) -> Union[str, Dict[str, Any], None]:
-    """
-    Sanitizes and potentially parses a single chunk of text.
-
-    This function performs several operations on the input chunk:
-    - Removes a specified prefix (`intro_value`).
-    - Strips leading/trailing characters (`strip_chars`).
-    - Skips chunks matching specific markers (`skip_markers`).
-    - Skips chunks matching regex patterns (`skip_regexes`).
-    - Extracts content using regex capturing groups (`extract_regexes`).
-    - Optionally parses the chunk as JSON (`to_json`).
-    - Handles JSON parsing errors with an optional callback (`error_handler`).
-
-    Args:
-        chunk (str): The chunk of text to process.
-        intro_value (str): The prefix to remove from the chunk.
-        to_json (bool): If True, attempts to parse the chunk as JSON.
-        skip_markers (List[str]): A list of markers; chunks matching these are skipped.
-        strip_chars (Optional[str]): Characters to strip from the beginning and end of the chunk.
-        yield_raw_on_error (bool): If True, returns the raw chunk when JSON parsing fails; otherwise, returns None.
-        error_handler (Optional[Callable[[Exception, str], Optional[Any]]]): An optional callback function that is called when JSON parsing fails.
-            It receives the exception and the sanitized chunk as arguments.  It should return a value to yield instead of the raw chunk, or None to ignore.
-        skip_regexes (Optional[List[Pattern[str]]]): A list of compiled regex patterns; chunks matching any of these are skipped.
-        extract_regexes (Optional[List[Pattern[str]]]): A list of compiled regex patterns for extracting content using capturing groups.
-
-    """
-    if not isinstance(chunk, str):
-        return None
-
-    # Fast path for empty chunks
-    if not chunk:
-        return None
-
-    # Use slicing for prefix removal (faster than startswith+slicing)
-    sanitized_chunk = chunk
-    if intro_value and len(chunk) >= len(intro_value) and chunk[:len(intro_value)] == intro_value:
-        sanitized_chunk = chunk[len(intro_value):]
-
-    # Optimize string stripping operations
-    if strip_chars is not None:
-        sanitized_chunk = sanitized_chunk.strip(strip_chars)
-    else:
-        # lstrip() is faster than strip() when we only need leading whitespace removed
-        sanitized_chunk = sanitized_chunk.lstrip()
-
-    # Skip empty chunks and markers
-    if not sanitized_chunk or any(marker == sanitized_chunk for marker in skip_markers):
-        return None
-
-    # Apply regex-based extraction first (if provided)
-    if extract_regexes:
-        for regex in extract_regexes:
-            match = regex.search(sanitized_chunk)
-            if match:
-                # If there are capturing groups, return the first group or all groups as a tuple
-                if match.groups():
-                    if len(match.groups()) == 1:
-                        sanitized_chunk = match.group(1)
-                    else:
-                        # Multiple groups - return as tuple converted to string for JSON compatibility
-                        sanitized_chunk = str(match.groups())
-                else:
-                    # No capturing groups, return the full match
-                    sanitized_chunk = match.group(0)
-                break  # Use first matching extraction regex
-
-    # Apply regex-based skipping (after extraction)
-    if skip_regexes:
-        if any(regex.search(sanitized_chunk) for regex in skip_regexes):
-            return None
-
-    # JSON parsing with optimized error handling
-    if to_json:
-        try:
-            # Only strip before JSON parsing if needed
-            if sanitized_chunk[0] not in '{[' or sanitized_chunk[-1] not in '}]':
-                sanitized_chunk = sanitized_chunk.strip()
-            return json.loads(sanitized_chunk)
-        except (json.JSONDecodeError, Exception) as e:
-            if error_handler:
-                try:
-                    handled = error_handler(e, sanitized_chunk)
-                    if handled is not None:
-                        return handled
-                except Exception:
-                    pass
-            return sanitized_chunk if yield_raw_on_error else None
-
-    return sanitized_chunk
-
-def _decode_byte_stream(
-    byte_iterator: Iterable[bytes],
-    encoding: EncodingType = 'utf-8',
-    errors: str = 'replace',
-    buffer_size: int = 8192
-) -> Generator[str, None, None]:
-    """
-    Decodes a byte stream in realtime with flexible encoding support.
-
-    This function takes an iterator of bytes and decodes it into a stream of strings
-    using the specified character encoding. It handles encoding errors gracefully
-    and can be tuned for performance with the `buffer_size` parameter.
-
-    Args:
-        byte_iterator (Iterable[bytes]): An iterator that yields chunks of bytes.
-        encoding (EncodingType): The character encoding to use for decoding.
-            Defaults to 'utf-8'.  Supports a wide range of encodings, including:
-            'utf-8', 'utf-16', 'utf-32', 'ascii', 'latin1', 'cp1252', 'iso-8859-1',
-            'iso-8859-2', 'windows-1250', 'windows-1251', 'windows-1252', 'gbk', 'big5',
-            'shift_jis', 'euc-jp', 'euc-kr'.
-        errors (str): Specifies how encoding errors should be handled.
-            Options are 'strict' (raises an error), 'ignore' (skips the error), and
-            'replace' (replaces the erroneous byte with a replacement character).
-            Defaults to 'replace'.
-        buffer_size (int): The size of the internal buffer used for decoding.
-
-    """
-    # Initialize decoder with the specified encoding
-    try:
-        decoder = codecs.getincrementaldecoder(encoding)(errors=errors)
-    except LookupError:
-        # Fallback to utf-8 if the encoding is not supported
-        decoder = codecs.getincrementaldecoder('utf-8')(errors=errors)
-
-    # Process byte stream in realtime
-    buffer = bytearray(buffer_size)
-    buffer_view = memoryview(buffer)
-
-    for chunk_bytes in byte_iterator:
-        if not chunk_bytes:
-            continue
-
-        try:
-            # Use buffer for processing if chunk size is appropriate
-            if len(chunk_bytes) <= buffer_size:
-                buffer[:len(chunk_bytes)] = chunk_bytes
-                text = decoder.decode(buffer_view[:len(chunk_bytes)], final=False)
-            else:
-                text = decoder.decode(chunk_bytes, final=False)
-
-            if text:
-                yield text
-        except UnicodeDecodeError:
-            yield f"[Encoding Error: Could not decode bytes with {encoding}]\n"
-
-    # Final flush
-    try:
-        final_text = decoder.decode(b'', final=True)
-        if final_text:
-            yield final_text
-    except UnicodeDecodeError:
-        yield f"[Encoding Error: Could not decode final bytes with {encoding}]\n"
-
-async def _decode_byte_stream_async(
-    byte_iterator: Iterable[bytes],
-    encoding: EncodingType = 'utf-8',
-    errors: str = 'replace',
-    buffer_size: int = 8192
-) -> AsyncGenerator[str, None]:
-    """
-    Asynchronously decodes a byte stream with flexible encoding support.
-
-    This function is the asynchronous counterpart to `_decode_byte_stream`. It takes
-    an asynchronous iterator of bytes and decodes it into a stream of strings using
-    the specified character encoding. It handles encoding errors gracefully and can
-    be tuned for performance with the `buffer_size` parameter.
-
-    Args:
-        byte_iterator (Iterable[bytes]): An asynchronous iterator that yields chunks of bytes.
-        encoding (EncodingType): The character encoding to use for decoding.
-            Defaults to 'utf-8'.  Supports a wide range of encodings, including:
-            'utf-8', 'utf-16', 'utf-32', 'ascii', 'latin1', 'cp1252', 'iso-8859-1',
-            'iso-8859-2', 'windows-1250', 'windows-1251', 'windows-1252', 'gbk', 'big5',
-            'shift_jis', 'euc-jp', 'euc-kr'.
-        errors (str): Specifies how encoding errors should be handled.
-            Options are 'strict' (raises an error), 'ignore' (skips the error), and
-            'replace' (replaces the erroneous byte with a replacement character).
-            Defaults to 'replace'.
-        buffer_size (int): The size of the internal buffer used for decoding.
-    """
-    try:
-        decoder = codecs.getincrementaldecoder(encoding)(errors=errors)
-    except LookupError:
-        decoder = codecs.getincrementaldecoder('utf-8')(errors=errors)
-
-    buffer = bytearray(buffer_size)
-    buffer_view = memoryview(buffer)
-
-    async for chunk_bytes in byte_iterator:
-        if not chunk_bytes:
-            continue
-        try:
-            if len(chunk_bytes) <= buffer_size:
-                buffer[:len(chunk_bytes)] = chunk_bytes
-                text = decoder.decode(buffer_view[:len(chunk_bytes)], final=False)
-            else:
-                text = decoder.decode(chunk_bytes, final=False)
-            if text:
-                yield text
-        except UnicodeDecodeError:
-            yield f"[Encoding Error: Could not decode bytes with {encoding}]\n"
-
-    try:
-        final_text = decoder.decode(b'', final=True)
-        if final_text:
-            yield final_text
-    except UnicodeDecodeError:
-        yield f"[Encoding Error: Could not decode final bytes with {encoding}]\n"
-
-def _sanitize_stream_sync(
-    data: Union[str, Iterable[str], Iterable[bytes]],
-    intro_value: str = "data:",
-    to_json: bool = True,
-    skip_markers: Optional[List[str]] = None,
-    strip_chars: Optional[str] = None,
-    start_marker: Optional[str] = None,
-    end_marker: Optional[str] = None,
-    content_extractor: Optional[Callable[[Union[str, Dict[str, Any]]], Optional[Any]]] = None,
-    yield_raw_on_error: bool = True,
-    encoding: EncodingType = 'utf-8',
-    encoding_errors: str = 'replace',
-    buffer_size: int = 8192,
-    line_delimiter: Optional[str] = None,
-    error_handler: Optional[Callable[[Exception, str], Optional[Any]]] = None,
-    skip_regexes: Optional[List[Union[str, Pattern[str]]]] = None,
-    extract_regexes: Optional[List[Union[str, Pattern[str]]]] = None,
-) -> Generator[Any, None, None]:
-    """
-    Processes a stream of data (strings or bytes) in real-time, applying various transformations and filtering.
-
-    This function is designed to handle streaming data, allowing for operations such as
-    prefix removal, JSON parsing, skipping lines based on markers, regex-based filtering,
-    and extracting specific content. It also supports custom error handling for JSON parsing failures.
-
-    Args:
-        data: String, iterable of strings, or iterable of bytes to process.
-        intro_value: Prefix indicating the start of meaningful data.
-        to_json: Parse the chunk as JSON if True.
-        skip_markers: Lines containing any of these markers are skipped.
-        strip_chars: Characters to strip from each line.
-        start_marker: Begin processing only after this marker is found.
-        end_marker: Stop processing once this marker is found.
-        content_extractor: Optional callable to transform parsed content before yielding.
-        yield_raw_on_error: Yield raw lines when JSON parsing fails.
-        encoding: Byte stream encoding.
-        encoding_errors: How to handle encoding errors.
-        buffer_size: Buffer size for byte decoding.
-        line_delimiter: Delimiter used to split incoming text into lines. ``None``
-            uses ``str.splitlines()``.
-        error_handler: Callback invoked with ``(Exception, str)`` when JSON
-            parsing fails. If the callback returns a value, it is yielded instead of the raw line.
-        skip_regexes: List of regex patterns (strings or compiled) for skipping lines that match.
-        extract_regexes: List of regex patterns (strings or compiled) for extracting content using capturing groups.
-
-    Yields:
-        Any: Processed data, which can be a string, a dictionary (if `to_json` is True), or the result of `content_extractor`.
-
-    Raises:
-        TypeError: If the input `data` is not a string or an iterable.
-        ValueError: If any regex pattern is invalid.
-    """
-    effective_skip_markers = skip_markers or []
-    # Compile regex patterns
-    compiled_skip_regexes = _compile_regexes(skip_regexes)
-    compiled_extract_regexes = _compile_regexes(extract_regexes)
-    
-    processing_active = start_marker is None
-    buffer = ""
-    found_start = False if start_marker else True
-    line_iterator: Iterable[str]
-
-    if isinstance(data, str):
-        # If data is a string, decide whether to split it into lines
-        # or treat it as an iterable containing a single chunk.
-        temp_lines: List[str]
-        if line_delimiter is None:  # Default: split by newlines if present
-            if '\n' in data or '\r' in data:
-                temp_lines = data.splitlines()
-            else:
-                temp_lines = [data]  # Treat as a single line/chunk
-        elif line_delimiter in data:  # Custom delimiter found in string
-            temp_lines = data.split(line_delimiter)
-        else:  # Custom delimiter not found, or string is effectively a single segment
-            temp_lines = [data]
-        line_iterator = iter(temp_lines)
-    elif hasattr(data, '__iter__'):  # data is an iterable (but not a string)
-        _iter = iter(data)
-        first_item = next(_iter, None)
-
-        if first_item is None:  # Iterable was empty
-            return
-
-        from itertools import chain
-        # Reconstruct the full iterable including the first_item
-        stream_input_iterable = chain([first_item], _iter)
-
-        if isinstance(first_item, bytes):
-            # Ensure stream_input_iterable is typed as Iterable[bytes] for _decode_byte_stream
-            line_iterator = _decode_byte_stream(
-                stream_input_iterable, # type: ignore
-                encoding=encoding,
-                errors=encoding_errors,
-                buffer_size=buffer_size
-            )
-        elif isinstance(first_item, str):
-            # Ensure stream_input_iterable is typed as Iterable[str]
-            line_iterator = stream_input_iterable # type: ignore
-        else:
-            raise TypeError(f"Iterable must yield strings or bytes, not {type(first_item).__name__}")
-    else:  # Not a string and not an iterable
-        raise TypeError(f"Input must be a string or an iterable, not {type(data).__name__}")
-
-    try:
-        for line in line_iterator:
-            if not line:
-                continue
-            buffer += line
-            while True:
-                # Look for start marker if needed
-                if not found_start and start_marker:
-                    idx = buffer.find(start_marker)
-                    if idx != -1:
-                        found_start = True
-                        buffer = buffer[idx + len(start_marker):]
-                    else:
-                        # Not found, keep buffering
-                        buffer = buffer[-max(len(start_marker), 256):]  # avoid unbounded growth
-                        break
-                # Look for end marker if needed
-                if found_start and end_marker:
-                    idx = buffer.find(end_marker)
-                    if idx != -1:
-                        chunk = buffer[:idx]
-                        buffer = buffer[idx + len(end_marker):]
-                        processing_active = False
-                    else:
-                        chunk = buffer
-                        buffer = ""
-                        processing_active = True
-                    # Process chunk if we are in active region
-                    if chunk and processing_active:
-                        for subline in (chunk.split(line_delimiter) if line_delimiter is not None else chunk.splitlines()):
-                            result = _process_chunk(
-                                subline,
-                                intro_value,
-                                to_json,
-                                effective_skip_markers,
-                                strip_chars,
-                                yield_raw_on_error,
-                                error_handler,
-                                compiled_skip_regexes,
-                                compiled_extract_regexes,
-                            )
-                            if result is None:
-                                continue
-                            if content_extractor:
-                                try:
-                                    final_content = content_extractor(result)
-                                    if final_content is not None:
-                                        yield final_content
-                                except Exception:
-                                    pass
-                            else:
-                                yield result
-                    if not processing_active:
-                        found_start = False
-                    if idx == -1:
-                        break
-                elif found_start:
-                    # No end marker, process all buffered content
-                    chunk = buffer
-                    buffer = ""
-                    if chunk:
-                        for subline in (chunk.split(line_delimiter) if line_delimiter is not None else chunk.splitlines()):
-                            result = _process_chunk(
-                                subline,
-                                intro_value,
-                                to_json,
-                                effective_skip_markers,
-                                strip_chars,
-                                yield_raw_on_error,
-                                error_handler,
-                                compiled_skip_regexes,
-                                compiled_extract_regexes,
-                            )
-                            if result is None:
-                                continue
-                            if content_extractor:
-                                try:
-                                    final_content = content_extractor(result)
-                                    if final_content is not None:
-                                        yield final_content
-                                except Exception:
-                                    pass
-                            else:
-                                yield result
-                    break
-                else:
-                    break
-    except Exception as e:
-        import sys
-        print(f"Stream processing error: {str(e)}", file=sys.stderr)
-
-
-async def _sanitize_stream_async(
-    data: Union[str, Iterable[str], Iterable[bytes]],
-    intro_value: str = "data:",
-    to_json: bool = True,
-    skip_markers: Optional[List[str]] = None,
-    strip_chars: Optional[str] = None,
-    start_marker: Optional[str] = None,
-    end_marker: Optional[str] = None,
-    content_extractor: Optional[Callable[[Union[str, Dict[str, Any]]], Optional[Any]]] = None,
-    yield_raw_on_error: bool = True,
-    encoding: EncodingType = 'utf-8',
-    encoding_errors: str = 'replace',
-    buffer_size: int = 8192,
-    line_delimiter: Optional[str] = None,
-    error_handler: Optional[Callable[[Exception, str], Optional[Any]]] = None,
-    skip_regexes: Optional[List[Union[str, Pattern[str]]]] = None,
-    extract_regexes: Optional[List[Union[str, Pattern[str]]]] = None,
-) -> AsyncGenerator[Any, None]:
-    """
-    Asynchronously processes a stream of data (strings or bytes), applying transformations and filtering.
-
-    This function is the asynchronous counterpart to `_sanitize_stream_sync`. It handles
-    streaming data, allowing for operations such as prefix removal, JSON parsing,
-    skipping lines based on markers, regex-based filtering, and extracting specific content. 
-    It also supports custom error handling for JSON parsing failures.
-
-    Args:
-        data: String, iterable of strings, or iterable of bytes to process.
-        intro_value: Prefix indicating the start of meaningful data.
-        to_json: Parse JSON content if ``True``.
-        skip_markers: Lines containing any of these markers are skipped.
-        strip_chars: Characters to strip from each line.
-        start_marker: Begin processing only after this marker is found.
-        end_marker: Stop processing once this marker is found.
-        content_extractor: Optional callable to transform parsed content before yielding.
-        yield_raw_on_error: Yield raw lines when JSON parsing fails.
-        encoding: Byte stream encoding.
-        encoding_errors: How to handle encoding errors.
-        buffer_size: Buffer size for byte decoding.
-        line_delimiter: Delimiter used to split incoming text into lines. ``None`` uses ``str.splitlines()``.
-        error_handler: Callback invoked with ``(Exception, str)`` when JSON parsing fails. If the callback returns a value, it is yielded in place of the raw line.
-        skip_regexes: List of regex patterns (strings or compiled) for skipping lines that match.
-        extract_regexes: List of regex patterns (strings or compiled) for extracting content using capturing groups.
-    """
-    if isinstance(data, str):
-        for item in _sanitize_stream_sync(
-            data,
-            intro_value=intro_value,
-            to_json=to_json,
-            skip_markers=skip_markers,
-            strip_chars=strip_chars,
-            start_marker=start_marker,
-            end_marker=end_marker,
-            content_extractor=content_extractor,
-            yield_raw_on_error=yield_raw_on_error,
-            encoding=encoding,
-            encoding_errors=encoding_errors,
-            buffer_size=buffer_size,
-            line_delimiter=line_delimiter,
-            error_handler=error_handler,
-            skip_regexes=skip_regexes,
-            extract_regexes=extract_regexes,
-        ):
-            yield item
-        return
-
-    if not hasattr(data, "__aiter__"):
-        # Fallback to synchronous processing if possible
-        for item in _sanitize_stream_sync(
-            data,
-            intro_value=intro_value,
-            to_json=to_json,
-            skip_markers=skip_markers,
-            strip_chars=strip_chars,
-            start_marker=start_marker,
-            end_marker=end_marker,
-            content_extractor=content_extractor,
-            yield_raw_on_error=yield_raw_on_error,
-            encoding=encoding,
-            encoding_errors=encoding_errors,
-            buffer_size=buffer_size,
-            line_delimiter=line_delimiter,
-            error_handler=error_handler,
-            skip_regexes=skip_regexes,
-            extract_regexes=extract_regexes,
-        ):
-            yield item
-        return
-
-    effective_skip_markers = skip_markers or []
-    # Compile regex patterns
-    compiled_skip_regexes = _compile_regexes(skip_regexes)
-    compiled_extract_regexes = _compile_regexes(extract_regexes)
-    
-    processing_active = start_marker is None
-    buffer = ""
-    found_start = False if start_marker else True
-
-    iterator = data.__aiter__()
-    first_item = None
-    async for first_item in iterator:
-        break
-    if first_item is None:
-        return
-    async def _chain(first, it):
-        yield first
-        async for x in it:
-            yield x
+from .sanitize import * # noqa: E402,F401
+from .conversation import Conversation  # noqa: E402,F401
+from .Extra.autocoder import AutoCoder  # noqa: E402,F401
+from .optimizers import Optimizers  # noqa: E402,F401
+from .prompt_manager import AwesomePrompts  # noqa: E402,F401
 
-    stream = _chain(first_item, iterator)
+# --- Utility Decorators ---
+from typing import Callable
+import time
+import functools
 
-    if isinstance(first_item, bytes):
-        line_iterator = _decode_byte_stream_async(
-            stream,
-            encoding=encoding,
-            errors=encoding_errors,
-            buffer_size=buffer_size,
-        )
-    elif isinstance(first_item, str):
-        line_iterator = stream
+def timeIt(func: Callable):
+    """
+    Decorator to measure execution time of a function (sync or async).
+    Prints: - Execution time for '{func.__name__}' : {elapsed:.6f} Seconds.  
+    """
+    import asyncio
+    GREEN_BOLD = "\033[1;92m"
+    RESET = "\033[0m"
+    @functools.wraps(func)
+    def sync_wrapper(*args, **kwargs):
+        start_time = time.time()
+        result = func(*args, **kwargs)
+        end_time = time.time()
+        print(f"{GREEN_BOLD}- Execution time for '{func.__name__}' : {end_time - start_time:.6f} Seconds.  {RESET}\n")
+        return result
+
+    @functools.wraps(func)
+    async def async_wrapper(*args, **kwargs):
+        start_time = time.time()
+        result = await func(*args, **kwargs)
+        end_time = time.time()
+        print(f"{GREEN_BOLD}- Execution time for '{func.__name__}' : {end_time - start_time:.6f} Seconds.  {RESET}\n")
+        return result
+
+    if asyncio.iscoroutinefunction(func):
+        return async_wrapper
     else:
-        raise TypeError(
-            f"Stream must yield strings or bytes, not {type(first_item).__name__}"
-        )
-
-    async for line in line_iterator:
-        if not line:
-            continue
-        buffer += line
-        while True:
-            if not found_start and start_marker:
-                idx = buffer.find(start_marker)
-                if idx != -1:
-                    found_start = True
-                    buffer = buffer[idx + len(start_marker) :]
-                else:
-                    buffer = buffer[-max(len(start_marker), 256) :]
-                    break
-            if found_start and end_marker:
-                idx = buffer.find(end_marker)
-                if idx != -1:
-                    chunk = buffer[:idx]
-                    buffer = buffer[idx + len(end_marker) :]
-                    processing_active = False
-                else:
-                    chunk = buffer
-                    buffer = ""
-                    processing_active = True
-                if chunk and processing_active:
-                    for subline in (
-                        chunk.split(line_delimiter)
-                        if line_delimiter is not None
-                        else chunk.splitlines()
-                    ):
-                        result = _process_chunk(
-                            subline,
-                            intro_value,
-                            to_json,
-                            effective_skip_markers,
-                            strip_chars,
-                            yield_raw_on_error,
-                            error_handler,
-                            compiled_skip_regexes,
-                            compiled_extract_regexes,
-                        )
-                        if result is None:
-                            continue
-                        if content_extractor:
-                            try:
-                                final_content = content_extractor(result)
-                                if final_content is not None:
-                                    yield final_content
-                            except Exception:
-                                pass
-                        else:
-                            yield result
-                if not processing_active:
-                    found_start = False
-                if idx == -1:
-                    break
-            elif found_start:
-                chunk = buffer
-                buffer = ""
-                if chunk:
-                    for subline in (
-                        chunk.split(line_delimiter)
-                        if line_delimiter is not None
-                        else chunk.splitlines()
-                    ):
-                        result = _process_chunk(
-                            subline,
-                            intro_value,
-                            to_json,
-                            effective_skip_markers,
-                            strip_chars,
-                            yield_raw_on_error,
-                            error_handler,
-                            compiled_skip_regexes,
-                            compiled_extract_regexes,
-                        )
-                        if result is None:
-                            continue
-                        if content_extractor:
-                            try:
-                                final_content = content_extractor(result)
-                                if final_content is not None:
-                                    yield final_content
-                            except Exception:
-                                pass
-                        else:
-                            yield result
-                break
-            else:
-                break
-
+        return sync_wrapper
 
-def sanitize_stream(
-    data: Union[
-        str,
-        bytes,
-        Iterable[str],
-        Iterable[bytes],
-        AsyncIterable[str],
-        AsyncIterable[bytes],
-        dict,
-        list,
-        int,
-        float,
-        bool,
-        None,
-    ],
-    intro_value: str = "data:",
-    to_json: bool = True,
-    skip_markers: Optional[List[str]] = None,
-    strip_chars: Optional[str] = None,
-    start_marker: Optional[str] = None,
-    end_marker: Optional[str] = None,
-    content_extractor: Optional[Callable[[Union[str, Dict[str, Any]]], Optional[Any]]] = None,
-    yield_raw_on_error: bool = True,
-    encoding: EncodingType = "utf-8",
-    encoding_errors: str = "replace",
-    buffer_size: int = 8192,
-    line_delimiter: Optional[str] = None,
-    error_handler: Optional[Callable[[Exception, str], Optional[Any]]] = None,
-    skip_regexes: Optional[List[Union[str, Pattern[str]]]] = None,
-    extract_regexes: Optional[List[Union[str, Pattern[str]]]] = None,
-    object_mode: Literal["as_is", "json", "str"] = "json",
-    raw: bool = False,
-) -> Union[Generator[Any, None, None], AsyncGenerator[Any, None]]:
+def retry(retries: int = 3, delay: float = 1) -> Callable:
     """
-    Processes streaming data (strings or bytes) in either synchronous or asynchronous mode.
-    Now supports non-iterable and miscellaneous input types (dict, list, int, float, bool, None).
-    Includes regex-based content filtering and extraction capabilities.
-
-    Args:
-        data: The data to be processed. Can be a string, bytes, a synchronous iterable of strings or bytes,
-            an asynchronous iterable of strings or bytes, or a single object (dict, list, int, float, bool, None).
-        intro_value (str): Prefix indicating the start of meaningful data. Defaults to "data:".
-        to_json (bool): Parse JSON content if ``True``. Defaults to True.
-        skip_markers (Optional[List[str]]): Lines containing any of these markers are skipped. Defaults to None.
-        strip_chars (Optional[str]): Characters to strip from each line. Defaults to None.
-        start_marker (Optional[str]): Begin processing only after this marker is found. Defaults to None.
-        end_marker (Optional[str]): Stop processing once this marker is found. Defaults to None.
-        content_extractor (Optional[Callable[[Union[str, Dict[str, Any]]], Optional[Any]]]):
-            Optional callable to transform parsed content before yielding. Defaults to None.
-        yield_raw_on_error (bool): Yield raw lines when JSON parsing fails. Defaults to True.
-        encoding (EncodingType): Byte stream encoding. Defaults to "utf-8".
-        encoding_errors (str): How to handle encoding errors. Defaults to "replace".
-        buffer_size (int): Buffer size for byte decoding. Defaults to 8192.
-        line_delimiter (Optional[str]): Delimiter used to split incoming text into lines.
-            ``None`` uses ``str.splitlines()``. Defaults to None.
-        error_handler (Optional[Callable[[Exception, str], Optional[Any]]]):
-            Callback invoked with ``(Exception, str)`` when JSON parsing fails.
-            If the callback returns a value, it is yielded in place of the raw line. Defaults to None.
-        skip_regexes (Optional[List[Union[str, Pattern[str]]]]): List of regex patterns (strings or compiled) 
-            for skipping lines that match any pattern. Defaults to None.
-        extract_regexes (Optional[List[Union[str, Pattern[str]]]]): List of regex patterns (strings or compiled) 
-            for extracting content using capturing groups. If multiple groups are captured, they are returned as a tuple string. Defaults to None.
-        object_mode (Literal["as_is", "json", "str"]): How to handle non-string, non-iterable objects.
-            "json" (default) yields as JSON string, "str" yields as str(obj), "as_is" yields the object as-is.
-        raw (bool): If True, yields the raw response as returned by the API, chunk by chunk (no splitting or joining).
-
-    Returns:
-        Union[Generator[Any, None, None], AsyncGenerator[Any, None]]:
-            A generator or an asynchronous generator yielding the processed data, or raw data if raw=True.
-            
-    Raises:
-        ValueError: If any regex pattern is invalid.
-    """    # --- RAW MODE: yield each chunk exactly as returned by the API ---
-    if raw:
-        def _raw_passthrough_sync(source_iter):
-            for chunk in source_iter:
-                if isinstance(chunk, (bytes, bytearray)):
-                    # Decode bytes preserving all whitespace and newlines
-                    yield chunk.decode(encoding, encoding_errors)
-                elif chunk is not None:
-                    # Yield string chunks as-is, preserving all formatting
-                    yield chunk
-                # Skip None chunks entirely
-        async def _raw_passthrough_async(source_aiter):
-            async for chunk in source_aiter:
-                if isinstance(chunk, (bytes, bytearray)):
-                    # Decode bytes preserving all whitespace and newlines
-                    yield chunk.decode(encoding, encoding_errors)
-                elif chunk is not None:
-                    # Yield string chunks as-is, preserving all formatting
-                    yield chunk
-                # Skip None chunks entirely
-        # Sync iterable (but not str/bytes)
-        if hasattr(data, "__iter__") and not isinstance(data, (str, bytes)):
-            return _raw_passthrough_sync(data)
-        # Async iterable
-        if hasattr(data, "__aiter__"):
-            return _raw_passthrough_async(data)
-        # Single string or bytes
-        if isinstance(data, (bytes, bytearray)):
-            def _yield_single():
-                yield data.decode(encoding, encoding_errors)
-            return _yield_single()
-        else:
-            def _yield_single():
-                if data is not None:
-                    yield data
-            return _yield_single()
-    # --- END RAW MODE ---
-
-    text_attr = getattr(data, "text", None)
-    content_attr = getattr(data, "content", None)
-
-    # Handle None
-    if data is None:
-        def _empty_gen():
-            if False:
-                yield None
-        return _empty_gen()
-
-    # Handle bytes directly
-    if isinstance(data, bytes):
-        try:
-            payload = data.decode(encoding, encoding_errors)
-        except Exception:
-            payload = str(data)
-        return _sanitize_stream_sync(
-            payload, intro_value, to_json, skip_markers, strip_chars,
-            start_marker, end_marker, content_extractor, yield_raw_on_error,
-            encoding, encoding_errors, buffer_size, line_delimiter, error_handler,
-            skip_regexes, extract_regexes,
-        )
-
-    # Handle string directly
-    if isinstance(data, str):
-        return _sanitize_stream_sync(
-            data, intro_value, to_json, skip_markers, strip_chars,
-            start_marker, end_marker, content_extractor, yield_raw_on_error,
-            encoding, encoding_errors, buffer_size, line_delimiter, error_handler,
-            skip_regexes, extract_regexes,
-        )
-
-    # Handle dict, list, int, float, bool (non-iterable, non-string/bytes)
-    if isinstance(data, (dict, list, int, float, bool)):
-        if object_mode == "as_is":
-            def _as_is_gen():
-                yield data
-            return _as_is_gen()
-        elif object_mode == "str":
-            return _sanitize_stream_sync(
-                str(data), intro_value, to_json, skip_markers, strip_chars,
-                start_marker, end_marker, content_extractor, yield_raw_on_error,
-                encoding, encoding_errors, buffer_size, line_delimiter, error_handler,
-                skip_regexes, extract_regexes,
-            )
-        else:  # "json"
-            try:
-                json_str = json.dumps(data)
-            except Exception:
-                json_str = str(data)
-            return _sanitize_stream_sync(
-                json_str, intro_value, to_json, skip_markers, strip_chars,
-                start_marker, end_marker, content_extractor, yield_raw_on_error,
-                encoding, encoding_errors, buffer_size, line_delimiter, error_handler,
-                skip_regexes, extract_regexes,
-            )
-
-    # Handle file-like objects (optional, treat as string if .read exists)
-    if hasattr(data, "read") and callable(data.read):
-        try:
-            file_content = data.read()
-            if isinstance(file_content, bytes):
-                file_content = file_content.decode(encoding, encoding_errors)
-            return _sanitize_stream_sync(
-                file_content, intro_value, to_json, skip_markers, strip_chars,
-                start_marker, end_marker, content_extractor, yield_raw_on_error,
-                encoding, encoding_errors, buffer_size, line_delimiter, error_handler,
-                skip_regexes, extract_regexes,
-            )
-        except Exception:
-            pass  # fallback to next
-
-    # Handle .text or .content attributes
-    if isinstance(text_attr, str):
-        payload = text_attr
-        return _sanitize_stream_sync(
-            payload, intro_value, to_json, skip_markers, strip_chars,
-            start_marker, end_marker, content_extractor, yield_raw_on_error,
-            encoding, encoding_errors, buffer_size, line_delimiter, error_handler,
-            skip_regexes, extract_regexes,
-        )
-    elif isinstance(content_attr, bytes):
-        try:
-            payload = content_attr.decode(encoding, encoding_errors)
-        except Exception:
-            payload = str(content_attr)
-        return _sanitize_stream_sync(
-            payload, intro_value, to_json, skip_markers, strip_chars,
-            start_marker, end_marker, content_extractor, yield_raw_on_error,
-            encoding, encoding_errors, buffer_size, line_delimiter, error_handler,
-            skip_regexes, extract_regexes,
-        )
-
-    # Handle async iterables
-    if hasattr(data, "__aiter__"):
-        return _sanitize_stream_async(
-            data, intro_value, to_json, skip_markers, strip_chars,
-            start_marker, end_marker, content_extractor, yield_raw_on_error,
-            encoding, encoding_errors, buffer_size, line_delimiter, error_handler,
-            skip_regexes, extract_regexes,
-        )
-    # Handle sync iterables (but not strings/bytes)
-    if hasattr(data, "__iter__"):
-        return _sanitize_stream_sync(
-            data, intro_value, to_json, skip_markers, strip_chars,
-            start_marker, end_marker, content_extractor, yield_raw_on_error,
-            encoding, encoding_errors, buffer_size, line_delimiter, error_handler,
-            skip_regexes, extract_regexes,
-        )
-    # Fallback: treat as string
-    return _sanitize_stream_sync(
-        str(data), intro_value, to_json, skip_markers, strip_chars,
-        start_marker, end_marker, content_extractor, yield_raw_on_error,
-        encoding, encoding_errors, buffer_size, line_delimiter, error_handler,
-        skip_regexes, extract_regexes,
-    )
-
-# --- Decorator version of sanitize_stream ---
-import functools
-from typing import overload
-
-def _sanitize_stream_decorator(
-    _func=None,
-    *,
-    intro_value: str = "data:",
-    to_json: bool = True,
-    skip_markers: Optional[List[str]] = None,
-    strip_chars: Optional[str] = None,
-    start_marker: Optional[str] = None,
-    end_marker: Optional[str] = None,
-    content_extractor: Optional[Callable[[Union[str, Dict[str, Any]]], Optional[Any]]] = None,
-    yield_raw_on_error: bool = True,
-    encoding: EncodingType = "utf-8",
-    encoding_errors: str = "replace",
-    buffer_size: int = 8192,
-    line_delimiter: Optional[str] = None,
-    error_handler: Optional[Callable[[Exception, str], Optional[Any]]] = None,
-    skip_regexes: Optional[List[Union[str, Pattern[str]]]] = None,
-    extract_regexes: Optional[List[Union[str, Pattern[str]]]] = None,
-    object_mode: Literal["as_is", "json", "str"] = "json",
-    raw: bool = False,
-):
+    Decorator to retry a function on exception.
     """
-    Decorator for sanitize_stream. Can be used as @sanitize_stream or @sanitize_stream(...).
-    All arguments are the same as sanitize_stream().
-    """
-    def decorator(func):
-        if asyncio.iscoroutinefunction(func):
-            @functools.wraps(func)
-            async def async_wrapper(*args, **kwargs):
-                result = await func(*args, **kwargs)
-                return sanitize_stream(
-                    result,
-                    intro_value=intro_value,
-                    to_json=to_json,
-                    skip_markers=skip_markers,
-                    strip_chars=strip_chars,
-                    start_marker=start_marker,
-                    end_marker=end_marker,
-                    content_extractor=content_extractor,
-                    yield_raw_on_error=yield_raw_on_error,
-                    encoding=encoding,
-                    encoding_errors=encoding_errors,
-                    buffer_size=buffer_size,
-                    line_delimiter=line_delimiter,
-                    error_handler=error_handler,
-                    skip_regexes=skip_regexes,
-                    extract_regexes=extract_regexes,
-                    object_mode=object_mode,
-                    raw=raw,
-                )
-            return async_wrapper
-        else:
-            @functools.wraps(func)
-            def sync_wrapper(*args, **kwargs):
-                result = func(*args, **kwargs)
-                return sanitize_stream(
-                    result,
-                    intro_value=intro_value,
-                    to_json=to_json,
-                    skip_markers=skip_markers,
-                    strip_chars=strip_chars,
-                    start_marker=start_marker,
-                    end_marker=end_marker,
-                    content_extractor=content_extractor,
-                    yield_raw_on_error=yield_raw_on_error,
-                    encoding=encoding,
-                    encoding_errors=encoding_errors,
-                    buffer_size=buffer_size,
-                    line_delimiter=line_delimiter,
-                    error_handler=error_handler,
-                    skip_regexes=skip_regexes,
-                    extract_regexes=extract_regexes,
-                    object_mode=object_mode,
-                    raw=raw,
-                )
-            return sync_wrapper
-    if _func is None:
-        return decorator
-    else:
-        return decorator(_func)
-
-# Alias for decorator usage
-LITSTREAM = sanitize_stream
-
-# Decorator aliases
-sanitize_stream_decorator = _sanitize_stream_decorator
-lit_streamer = _sanitize_stream_decorator
-
-# Allow @sanitize_stream and @lit_streamer as decorators
-import asyncio
-sanitize_stream.__decorator__ = _sanitize_stream_decorator
-LITSTREAM.__decorator__ = _sanitize_stream_decorator
-lit_streamer.__decorator__ = _sanitize_stream_decorator
-
-def __getattr__(name):
-    if name == 'sanitize_stream':
-        return sanitize_stream
-    if name == 'LITSTREAM':
-        return LITSTREAM
-    if name == 'sanitize_stream_decorator':
-        return _sanitize_stream_decorator
-    if name == 'lit_streamer':
-        return _sanitize_stream_decorator
-    raise AttributeError(f"module {__name__} has no attribute {name}")
-from .conversation import Conversation  # noqa: E402,F401
-from .Extra.autocoder import AutoCoder  # noqa: E402,F401
-from .optimizers import Optimizers  # noqa: E402,F401
-from .prompt_manager import AwesomePrompts  # noqa: E402,F401
+    def decorator(func: Callable):
+        @functools.wraps(func)
+        def wrapper(*args, **kwargs):
+            last_exc = None
+            for attempt in range(retries):
+                try:
+                    return func(*args, **kwargs)
+                except Exception as exc:
+                    last_exc = exc
+                    print(f"Attempt {attempt + 1} failed: {exc}. Retrying in {delay} seconds...")
+                    time.sleep(delay)
+            raise last_exc
+        return wrapper
+    return decorator