sunholo 0.138.1__tar.gz → 0.139.0__tar.gz

{sunholo-0.138.1/src/sunholo.egg-info → sunholo-0.139.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sunholo
-Version: 0.138.1
+Version: 0.139.0
 Summary: AI DevOps - a package to help deploy GenAI to the Cloud.
 Author-email: Holosun ApS <multivac@sunholo.com>
 License: Apache License, Version 2.0

{sunholo-0.138.1 → sunholo-0.139.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "sunholo"
-version = "0.138.1"
+version = "0.139.0"
 description = "AI DevOps - a package to help deploy GenAI to the Cloud."
 readme = "README.md"
 requires-python = ">=3.10"

sunholo-0.139.0/src/sunholo/gcs/download_gcs_text.py

@@ -0,0 +1,132 @@
+import os
+import json
+
+from ..custom_logging import log
+from ..utils.mime import get_mime_type_gemini
+from .download_url import get_bytes_from_gcs
+
+def download_gcs_source_to_string(source:str) -> str:
+    """
+    source: str The Google Cloud Storage URI of the file to download (e.g., 'gs://bucket_name/file_name').
+    """
+    mime_type = get_mime_type_gemini(source)
+    if mime_type == "":
+        log.warning(f"Can not download to string file source {source}")
+        return ""
+    """
+        mime_types = {
+
+        # Images
+        'png': 'image/png',
+        'jpg': 'image/jpeg',
+        'jpeg': 'image/jpeg',
+        'gif': 'image/gif',
+        'webp': 'image/webp',
+
+        # Document formats
+        'pdf': 'application/pdf',
+        
+        # Programming languages
+        'js': 'text/javascript',
+        'py': 'text/x-python',
+        
+        # Web formats
+        'html': 'text/html',
+        'htm': 'text/html',
+        'css': 'text/css',
+        
+        # Text formats
+        'txt': 'text/plain',
+        'md': 'text/md',
+        'csv': 'text/csv',
+        'xml': 'text/xml',
+        'rtf': 'text/rtf',
+        
+        # Special case: JSON files are treated as plain text
+        'json': 'text/plain'
+    }
+    """
+    if mime_type.startswith("image/") or mime_type == "application/pdf":
+        log.warning(f"Can not download to string file source {source} of type {mime_type}")
+        return ""
+
+    try:
+        log.info(f"Extracting text for {source}")
+        bytes = get_bytes_from_gcs(source)
+        string = bytes.decode('utf-8', errors='replace')
+        log.info(f"Extracted {len(string)} characters from {source}: {string[:100]}")
+
+    except Exception as err:
+        log.error(f"Could not extract string text for {source}: {str(err)}")
+
+        return ""
+
+    if not string:
+        raise ValueError(f"No string text for {source}")
+
+    file_ext = os.path.splitext(source)[1].lower().lstrip('.')
+    if file_ext == "json":
+        try:
+            extracted_data = json.loads(string)
+            log.debug("Turning json text into markdown format so as not to confuse structured output", log_struct=extracted_data)
+            string = json_data_to_markdown(extracted_data)
+        except json.JSONDecodeError:
+            log.warning(f"Could not get valid json from .json file: {source}")
+    
+    return string
+
+def json_data_to_markdown(data, indent_level: int = 0) -> str:
+    """
+    Recursively converts a Python object (from parsed JSON) into a Markdown string.
+    """
+    indent = "  " * indent_level  # Use 2 spaces for indentation
+    markdown_parts = []
+
+    if isinstance(data, dict):
+        if not data:
+            return f"{indent}(empty object)"
+        for key, value in data.items():
+            # Process the value recursively
+            value_md = json_data_to_markdown(value, indent_level + 1)
+            # Determine if the rendered value is complex (multi-line or was list/dict)
+            is_complex_render = "\n" in value_md.strip() or (isinstance(value, (dict, list)) and value)
+
+            if is_complex_render:
+                markdown_parts.append(f"{indent}**{key}**:")
+                markdown_parts.append(value_md)
+            else:
+                # Simple value rendering, strip its own indent before adding key
+                markdown_parts.append(f"{indent}**{key}**: {value_md.strip()}")
+        return "\n".join(markdown_parts)
+
+    elif isinstance(data, list):
+        if not data:
+            return f"{indent}(empty list)"
+        for item in data:
+            # Process item recursively
+            item_md = json_data_to_markdown(item, indent_level + 1)
+            # Remove leading indent from the recursive call before processing lines
+            lines = item_md.lstrip(' ').split('\n')
+            # Add bullet point to the first line
+            first_line = f"{indent}- {lines[0]}"
+            # Ensure subsequent lines are indented correctly relative to the bullet
+            rest_lines = [f"{indent}  {line}" for line in lines[1:]]
+            markdown_parts.append(first_line)
+            markdown_parts.extend(rest_lines)
+        return "\n".join(markdown_parts)
+
+    elif isinstance(data, str):
+        # Handle multi-line strings: indent subsequent lines
+        lines = data.split('\n')
+        if len(lines) <= 1:
+            return f"{indent}{data}" # Single line string
+        else:
+            indented_lines = [f"{indent}{lines[0]}"] + [f"{indent}  {line}" for line in lines[1:]]
+            return "\n".join(indented_lines)
+
+    elif data is None:
+        return f"{indent}*null*" # Represent None distinctly
+    elif isinstance(data, bool):
+        return f"{indent}{str(data).lower()}" # true / false
+    else:  # Numbers (int, float)
+        return f"{indent}{str(data)}"

{sunholo-0.138.1 → sunholo-0.139.0}/src/sunholo/utils/mime.py

@@ -66,3 +66,53 @@ def guess_mime_type(file_path: str) -> str:
 
     return mime
 
+
+def get_mime_type_gemini(file_path:str) -> str:
+    """
+    Determine the MIME type based on file extension.
+    Only returns valid Gemini formats, or None if they are not supported.
+    
+    Args:
+        file_path (str): Path to the file
+        
+    Returns:
+        str: The appropriate MIME type for the file
+    """
+    # Extract the file extension (lowercase)
+    ext = os.path.splitext(file_path)[1].lower().lstrip('.')
+    
+    # Define the mapping of extensions to MIME types
+    mime_types = {
+
+        # Images
+        'png': 'image/png',
+        'jpg': 'image/jpeg',
+        'jpeg': 'image/jpeg',
+        'gif': 'image/gif',
+        'webp': 'image/webp',
+
+        # Document formats
+        'pdf': 'application/pdf',
+        
+        # Programming languages
+        'js': 'text/javascript',
+        'py': 'text/x-python',
+        
+        # Web formats
+        'html': 'text/html',
+        'htm': 'text/html',
+        'css': 'text/css',
+        
+        # Text formats
+        'txt': 'text/plain',
+        'md': 'text/md',
+        'csv': 'text/csv',
+        'xml': 'text/xml',
+        'rtf': 'text/rtf',
+        
+        # Special case: JSON files are treated as plain text
+        'json': 'text/plain'
+    }
+    
+    # Return the appropriate MIME type, defaulting to None if unknown
+    return mime_types.get(ext, "")

{sunholo-0.138.1 → sunholo-0.139.0/src/sunholo.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sunholo
-Version: 0.138.1
+Version: 0.139.0
 Summary: AI DevOps - a package to help deploy GenAI to the Cloud.
 Author-email: Holosun ApS <multivac@sunholo.com>
 License: Apache License, Version 2.0

{sunholo-0.138.1 → sunholo-0.139.0}/src/sunholo.egg-info/SOURCES.txt

@@ -95,6 +95,7 @@ src/sunholo/excel/plugin.py
 src/sunholo/gcs/__init__.py
 src/sunholo/gcs/add_file.py
 src/sunholo/gcs/download_folder.py
+src/sunholo/gcs/download_gcs_text.py
 src/sunholo/gcs/download_url.py
 src/sunholo/gcs/extract_and_sign.py
 src/sunholo/gcs/metadata.py