sunholo 0.107.0__tar.gz → 0.109.1__tar.gz

{sunholo-0.107.0 → sunholo-0.109.1}/PKG-INFO

@@ -1,9 +1,9 @@
 Metadata-Version: 2.1
 Name: sunholo
-Version: 0.107.0
+Version: 0.109.1
 Summary: Large Language Model DevOps - a package to help deploy LLMs to the Cloud.
 Home-page: https://github.com/sunholo-data/sunholo-py
-Download-URL: https://github.com/sunholo-data/sunholo-py/archive/refs/tags/v0.107.0.tar.gz
+Download-URL: https://github.com/sunholo-data/sunholo-py/archive/refs/tags/v0.109.1.tar.gz
 Author: Holosun ApS
 Author-email: multivac@sunholo.com
 License: Apache License, Version 2.0

{sunholo-0.107.0 → sunholo-0.109.1}/setup.py

@@ -1,6 +1,6 @@
 from setuptools import setup, find_packages
 
-version = '0.107.0'
+version = '0.109.1'
 
 setup(
     name='sunholo',

{sunholo-0.107.0 → sunholo-0.109.1}/sunholo/database/alloydb.py

@@ -275,17 +275,19 @@ def _list_sources_from_docstore(sources, vector_name, search_type="OR"):
     if sources:
         conditions = and_or_ilike(sources, search_type=search_type)
         query = f"""
-            SELECT DISTINCT langchain_metadata->>'objectId' AS objectId
+            SELECT source AS objectId
             FROM {table_name}
             WHERE {conditions}
-            ORDER BY langchain_metadata->>'objectId' ASC
+            GROUP BY source
+            ORDER BY source ASC
             LIMIT 500;
         """
     else:
         query = f"""
-            SELECT DISTINCT langchain_metadata->>'objectId' AS objectId
+            SELECT source AS objectId
             FROM {table_name}
-            ORDER BY langchain_metadata->>'objectId' ASC
+            GROUP BY source
+            ORDER BY source ASC
             LIMIT 500;
         """
 

sunholo-0.109.1/sunholo/genai/__init__.py

@@ -0,0 +1,4 @@
+from .process_funcs_cls import GenAIFunctionProcessor
+from .safety import genai_safety
+from .init import init_genai
+from .file_handling import download_gcs_upload_genai, construct_file_content

sunholo-0.109.1/sunholo/genai/file_handling.py

@@ -0,0 +1,186 @@
+from ..custom_logging import log
+from ..gcs import get_bytes_from_gcs
+
+import mimetypes
+import asyncio
+import tempfile
+import re
+import traceback
+try:
+    import google.generativeai as genai
+    from google.generativeai.types import file_types
+except ImportError:
+    genai = None
+    file_types = None
+    
+DOCUMENT_MIMES = [
+    'application/pdf',
+    'application/x-javascript',
+    'text/javascript',
+    'application/x-python',
+    'text/x-python',
+    'text/plain',
+    'text/html',
+    'text/css',
+    'text/md',
+    'text/csv',
+    'text/xml',
+    'text/rtf'
+    ]
+
+IMAGE_MIMES = [
+    'image/png',
+    'image/jpeg',
+    'image/webp',
+    'image/heic',
+    'image/heif',
+]
+
+VIDEO_MIMES = [
+    'video/mp4',
+    'video/mpeg',
+    'video/mov',
+    'video/avi',
+    'video/x-flv',
+    'video/mpg',
+    'video/webm',
+    'video/wmv',
+    'video/3gpp'
+]
+
+AUDIO_MIMES = [
+    'audio/wav',
+    'audio/mp3',
+    'audio/aiff',
+    'audio/aac',
+    'audio/ogg',
+    'audio/flac',    
+]
+
+ALLOWED_MIME_TYPES = set(AUDIO_MIMES + VIDEO_MIMES + IMAGE_MIMES + DOCUMENT_MIMES)
+
+# 'documents': 
+# [
+#   {'storagePath': 'users/UQcKi4u7s...dsd.png', 
+#    'url': 'https://firebasestorage.googleapis.com/v0/b/multi...', 
+#    'contentType': 'image/png', 
+#    'type': 'image', 
+#    'name': 'multivac-data-architecture.png'}, 
+#   {'storagePath': 'users/UQc...3dc59e1.jpg', 
+#   'type': 'image', 
+#   'name': 'holosun-circle.jpg', 
+#   'url': 'https://firebasestorage.googleapis.com/v0/b/multiv...', 
+#   'contentType': 'image/jpeg'}
+# ]
+async def construct_file_content(gs_list, bucket:str):
+    """
+    Args:
+    - gs_list: a list of dicts representing files in a bucket
+       - contentType: The content type of the file on GCS
+       - storagePath: The path in the bucket
+    - bucket: The bucket the files are in
+
+    """
+    
+    file_list = []
+    for element in gs_list:
+
+        the_mime_type = element.get('contentType')
+        if the_mime_type is None:
+            continue
+        if element.get('storagePath') is None:
+            continue
+        if the_mime_type in ALLOWED_MIME_TYPES:
+            file_list.append(element)
+    
+    if not file_list:
+        return {"role": "user", "parts": [{"text": "No eligible contentTypes were found"}]}
+
+    content = []
+    
+    # Loop through the valid files and process them
+    tasks = []
+    for file_info in file_list:
+        img_url = f"gs://{bucket}/{file_info['storagePath']}"
+        mime_type = file_info['contentType']
+        # Append the async download task to the task list
+        tasks.append(download_gcs_upload_genai(img_url, mime_type))
+
+    # Run all tasks in parallel
+    content = await asyncio.gather(*tasks)
+
+    return content
+
+# Helper function to handle each file download with error handling
+async def download_file_with_error_handling(img_url, mime_type):
+    try:
+        return await download_gcs_upload_genai(img_url, mime_type)
+    except Exception as err:
+        msg= f"Error processing file from {img_url}: {str(err)}"
+        log.error(msg)
+        return {"role": "user", "parts": [{"text": msg}]}
+
+async def download_gcs_upload_genai(img_url, mime_type, retries=3, delay=2):
+    import aiofiles
+    """
+    Downloads and uploads a file with retries in case of failure.
+    
+    Args:
+      - img_url: str The URL of the file to download.
+      - mime_type: str The MIME type of the file.
+      - retries: int Number of retry attempts before failing.
+      - delay: int Initial delay between retries, exponentially increasing.
+      
+    Returns:
+      - downloaded_content: The result of the file upload if successful.
+    """
+    for attempt in range(retries):
+        try:
+            log.info(f"Upload {attempt} for {img_url=}")
+            # Download the file bytes asynchronously
+            file_bytes = await asyncio.to_thread(get_bytes_from_gcs, img_url)
+            if not file_bytes:
+                msg = f"Failed to download file from {img_url}: got None"
+                log.warning(msg)
+                return {"role": "user", "parts": [{"text": msg}]}
+            
+            # Log the size of the file bytes
+            file_size = len(file_bytes)
+            log.info(f"Downloaded file size for {img_url}: {file_size} bytes")
+
+            if file_size > 19434343:
+                log.warning(f"File size for {img_url}: {file_size} is too big.")
+                msg =  f"The file for {img_url} is too large ({file_size} bytes) to be used directly.  Use RAG instead."
+                return {"role": "user", "parts": [{"text": msg}]}
+            
+            extension = mimetypes.guess_extension(mime_type)
+            
+            # Use aiofiles for asynchronous file operations
+            temp_file = tempfile.NamedTemporaryFile(delete=False, suffix=extension)
+            downloaded_file = temp_file.name
+
+            sanitized_file = re.sub(r'[^\w\-.]', '_', downloaded_file)
+
+            log.info(f"Writing file {sanitized_file}")
+            async with aiofiles.open(sanitized_file, 'wb') as f:
+                await f.write(file_bytes)
+
+            # Upload the file and get its content reference
+            try:
+                downloaded_content: file_types.File = await asyncio.to_thread(genai.upload_file, sanitized_file )
+                return {"role": "user", "parts": [{"file_data": downloaded_content}]}
+            except Exception as err:
+                msg = f"Could not upload {sanitized_file} to genai.upload_file: {str(err)} {traceback.format_exc()}"
+                log.error(msg)
+                return {"role": "user", "parts": [{"text": msg}]}
+        
+        except Exception as err:
+            log.error(f"Error processing file {img_url} on attempt {attempt + 1}/{retries}: {str(err)}")
+
+            if attempt < retries - 1:
+                log.info(f"Retrying in {delay} seconds...")
+                await asyncio.sleep(delay)
+                delay *= 2  # Exponential backoff
+            else:
+                raise err  # Raise the error after max retries
+

{sunholo-0.107.0 → sunholo-0.109.1}/sunholo/senses/stream_voice.py

@@ -23,6 +23,8 @@ import queue
 import threading
 import time
 from concurrent.futures import ThreadPoolExecutor
+import io
+import wave
 
 import argparse
 import json
@@ -131,6 +133,43 @@ class StreamingTTS:
         # Convert audio bytes to numpy array for playback
         audio_np = np.frombuffer(response.audio_content, dtype=np.int16)
         return audio_np
+
+    def generate_audio_stream(self, text):
+        """
+        Generate a stream of audio data from a text chunk.
+        Returns audio in WAV format for streaming.
+        
+        Args:
+            text (str): Text to convert to speech
+            
+        Yields:
+            bytes: WAV-formatted audio data
+        """
+        try:
+            # Convert text to audio using existing method
+            audio_chunk = self.text_to_audio(text)
+            
+            # Process audio chunk with fading
+            processed_chunk = self._apply_fade(
+                audio_chunk,
+                fade_duration=self.file_fade_duration,
+                fade_in=True,
+                fade_out=True
+            )
+            
+            # Convert to WAV format
+            wav_buffer = io.BytesIO()
+            with wave.open(wav_buffer, 'wb') as wav_file:
+                wav_file.setnchannels(1)
+                wav_file.setsampwidth(2)
+                wav_file.setframerate(self.sample_rate)
+                wav_file.writeframes(processed_chunk.tobytes())
+            
+            yield wav_buffer.getvalue()
+            
+        except Exception as e:
+            log.error(f"Error generating audio stream: {e}")
+            yield b''
     
     def _initialize_audio_device(self):
         """Initialize audio device with proper settings."""

{sunholo-0.107.0 → sunholo-0.109.1}/sunholo.egg-info/PKG-INFO

@@ -1,9 +1,9 @@
 Metadata-Version: 2.1
 Name: sunholo
-Version: 0.107.0
+Version: 0.109.1
 Summary: Large Language Model DevOps - a package to help deploy LLMs to the Cloud.
 Home-page: https://github.com/sunholo-data/sunholo-py
-Download-URL: https://github.com/sunholo-data/sunholo-py/archive/refs/tags/v0.107.0.tar.gz
+Download-URL: https://github.com/sunholo-data/sunholo-py/archive/refs/tags/v0.109.1.tar.gz
 Author: Holosun ApS
 Author-email: multivac@sunholo.com
 License: Apache License, Version 2.0

{sunholo-0.107.0 → sunholo-0.109.1}/sunholo.egg-info/SOURCES.txt

@@ -97,6 +97,7 @@ sunholo/gcs/download_url.py
 sunholo/gcs/extract_and_sign.py
 sunholo/gcs/metadata.py
 sunholo/genai/__init__.py
+sunholo/genai/file_handling.py
 sunholo/genai/images.py
 sunholo/genai/init.py
 sunholo/genai/process_funcs_cls.py

sunholo-0.107.0/sunholo/genai/__init__.py

@@ -1,3 +0,0 @@
-from .process_funcs_cls import GenAIFunctionProcessor
-from .safety import genai_safety
-from .init import init_genai