ara-cli 0.1.9.81__tar.gz → 0.1.9.84__tar.gz

{ara_cli-0.1.9.81 → ara_cli-0.1.9.84}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ara_cli
-Version: 0.1.9.81
+Version: 0.1.9.84
 Summary: Powerful, open source command-line tool for managing, structuring and automating software development artifacts in line with Business-Driven Development (BDD) and AI-assisted processes
 Description-Content-Type: text/markdown
 Requires-Dist: litellm

{ara_cli-0.1.9.81 → ara_cli-0.1.9.84}/ara_cli/chat.py

@@ -1,12 +1,20 @@
 import os
-import cmd2
 import argparse
+import cmd2
 from ara_cli.prompt_handler import send_prompt
 
+from ara_cli.file_loaders.document_file_loader import DocumentFileLoader
+from ara_cli.file_loaders.binary_file_loader import BinaryFileLoader
+from ara_cli.file_loaders.text_file_loader import TextFileLoader
+
 
 extract_parser = argparse.ArgumentParser()
 extract_parser.add_argument('-s', '--skip-queries', action='store_true', help='Force extraction')
 
+load_parser = argparse.ArgumentParser()
+load_parser.add_argument('file_name', nargs='?', default='', help='File to load')
+load_parser.add_argument('--load-images', action='store_true', help='Extract and describe images from documents')
+
 
 class Chat(cmd2.Cmd):
     CATEGORY_CHAT_CONTROL = "Chat control commands"
@@ -353,75 +361,67 @@ Start chatting (type 'HELP'/'h' for available commands, 'QUIT'/'q' to exit chat
             return None
         return file_path
 
-    @file_exists_check
-    def load_text_file(self, file_path, prefix: str = "", suffix: str = "", block_delimiter: str = ""):
-        with open(file_path, 'r', encoding='utf-8', errors="replace") as file:
-            file_content = file.read()
-            if block_delimiter:
-                file_content = f"{block_delimiter}\n{file_content}\n{block_delimiter}"
-        write_content = f"{prefix}{file_content}{suffix}\n"
-
-        with open(self.chat_name, 'a', encoding='utf-8') as chat_file:
-            chat_file.write(write_content)
-            return True
+    # @file_exists_check
+    def load_text_file(
+        self,
+        file_path,
+        prefix: str = "",
+        suffix: str = "",
+        block_delimiter: str = "",
+        extract_images: bool = False
+    ):
+        loader = TextFileLoader(self)
+        return loader.load(
+            file_path,
+            prefix=prefix,
+            suffix=suffix,
+            block_delimiter=block_delimiter,
+            extract_images=extract_images
+        )
 
-    @file_exists_check
+    # @file_exists_check
     def load_binary_file(self, file_path, mime_type: str, prefix: str = "", suffix: str = ""):
-        import base64
-
-        with open(file_path, 'rb') as file:
-            file_content = file.read()
-        base64_image = base64.b64encode(file_content).decode("utf-8")
-
-        write_content = f"{prefix}![{os.path.basename(file_path)}](data:{mime_type};base64,{base64_image}){suffix}\n"
-
-        with open(self.chat_name, 'a', encoding='utf-8') as chat_file:
-            chat_file.write(write_content)
-            return True
-    
-    def read_docx(self, file_path):
-        import docx
-        doc = docx.Document(file_path)
-        return '\n'.join(para.text for para in doc.paragraphs)
-
-    def read_pdf(self, file_path):
-        import pymupdf4llm
-        return pymupdf4llm.to_markdown(file_path, write_images=False)
-
-    def read_odt(self, file_path):
-        import pymupdf4llm
-        return pymupdf4llm.to_markdown(file_path, write_images=False)
-
-    @file_exists_check
-    def load_document_file(self, file_path: str, prefix: str = "", suffix: str = "", block_delimiter: str = "```"):
-        import os
-
-        _, ext = os.path.splitext(file_path)
-        ext = ext.lower()
-
-        text_content = ""
-        match ext:
-            case ".docx":
-                text_content = self.read_docx(file_path)
-            case ".pdf":
-                text_content = self.read_pdf(file_path)
-            case ".odt":
-                text_content = self.read_odt(file_path)
-            # Add more cases if needed.
-            case _:
-                print("Unsupported document type.")
-                return False
+        loader = BinaryFileLoader(self)
+        return loader.load(
+            file_path,
+            mime_type=mime_type,
+            prefix=prefix,
+            suffix=suffix
+        )
 
-        if block_delimiter:
-            text_content = f"{block_delimiter}\n{text_content}\n{block_delimiter}"
+    def read_markdown(self, file_path: str, extract_images: bool = False) -> str:
+        """Read markdown file and optionally extract/describe images"""
+        from ara_cli.file_loaders.text_file_loader import MarkdownReader
 
-        write_content = f"{prefix}{text_content}{suffix}\n"
+        reader = MarkdownReader(file_path)
+        return reader.read(extract_images=extract_images)
 
-        with open(self.chat_name, 'a', encoding='utf-8') as chat_file:
-            chat_file.write(write_content)
-        return True
+    # @file_exists_check
+    def load_document_file(
+        self,
+        file_path: str,
+        prefix: str = "",
+        suffix: str = "",
+        block_delimiter: str = "```",
+        extract_images: bool = False
+    ):
+        loader = DocumentFileLoader(self)
+        return loader.load(
+            file_path,
+            prefix=prefix,
+            suffix=suffix,
+            block_delimiter=block_delimiter,
+            extract_images=extract_images
+        )
 
-    def load_file(self, file_name: str, prefix: str = "", suffix: str = "", block_delimiter: str = ""):
+    def load_file(
+        self,
+        file_name: str,
+        prefix: str = "",
+        suffix: str = "",
+        block_delimiter: str = "",
+        extract_images: bool = False
+    ):
         binary_type_mapping = Chat.BINARY_TYPE_MAPPING
         document_type_extensions = Chat.DOCUMENT_TYPE_EXTENSIONS
 
@@ -432,28 +432,31 @@ Start chatting (type 'HELP'/'h' for available commands, 'QUIT'/'q' to exit chat
                 file_type = mime_type
                 break
 
-        is_file_document = any(file_name_lower.endswith(ext) for ext in document_type_extensions)
+        is_file_document = any(file_name_lower.endswith(ext)
+                               for ext in document_type_extensions)
 
         if is_file_document:
             return self.load_document_file(
-                file_name=file_name,
+                file_path=file_name,
                 prefix=prefix,
                 suffix=suffix,
-                block_delimiter=block_delimiter
+                block_delimiter=block_delimiter,
+                extract_images=extract_images
             )
         elif file_type:
             return self.load_binary_file(
-                file_name=file_name,
+                file_path=file_name,
                 mime_type=file_type,
                 prefix=prefix,
                 suffix=suffix
             )
         else:
             return self.load_text_file(
-                file_name=file_name,
+                file_path=file_name,
                 prefix=prefix,
                 suffix=suffix,
-                block_delimiter=block_delimiter
+                block_delimiter=block_delimiter,
+                extract_images=extract_images
             )
 
     def choose_file_to_load(self, files: list[str], pattern: str):
@@ -496,8 +499,14 @@ Start chatting (type 'HELP'/'h' for available commands, 'QUIT'/'q' to exit chat
         self.message_buffer.append(self.full_input)
 
     @cmd2.with_category(CATEGORY_CHAT_CONTROL)
-    def do_LOAD(self, file_name):
-        """Load a file and append its contents to chat file. Can be given the file name in-line. Will attempt to find the file relative to chat file first, then treat the given path as absolute"""
+    @cmd2.with_argparser(load_parser)
+    def do_LOAD(self, args):
+        """Load a file and append its contents to chat file. Can be given the file name in-line. Will attempt to find the file relative to chat file first, then treat the given path as absolute. Use --load-images flag to extract and describe images from documents."""
+        from ara_cli.commands.load_command import LoadCommand
+
+        file_name = args.file_name
+        load_images = args.load_images
+        
         matching_files = self.find_matching_files_to_load(file_name)
         if not matching_files:
             return
@@ -506,12 +515,20 @@ Start chatting (type 'HELP'/'h' for available commands, 'QUIT'/'q' to exit chat
             block_delimiter = "```"
             prefix = f"\nFile: {file_path}\n"
             self.add_prompt_tag_if_needed(self.chat_name)
-            if not os.path.isdir(file_path) and self.load_file(file_path, prefix=prefix, block_delimiter=block_delimiter):
-                print(f"Loaded contents of file {file_path}")
+            
+            if not os.path.isdir(file_path):
+                command = LoadCommand(
+                    chat_instance=self,
+                    file_path=file_path,
+                    prefix=prefix,
+                    block_delimiter=block_delimiter,
+                    extract_images=load_images,
+                    output=self.poutput
+                )
+                command.execute()
 
     def complete_LOAD(self, text, line, begidx, endidx):
         import glob
-
         return [x for x in glob.glob(text + '*')]
 
     def _retrieve_ara_config(self):
@@ -547,7 +564,7 @@ Start chatting (type 'HELP'/'h' for available commands, 'QUIT'/'q' to exit chat
 
         if file_type:
             return self.load_binary_file(
-                file_name=file_name,
+                file_path=file_name,
                 mime_type=file_type,
                 prefix=prefix,
                 suffix=suffix
@@ -562,22 +579,12 @@ Start chatting (type 'HELP'/'h' for available commands, 'QUIT'/'q' to exit chat
             return False
         return True
 
-    @cmd2.with_category(CATEGORY_CHAT_CONTROL)
-    def do_LOAD_DOCUMENT(self, file_name):
-        """Load a document file (PDF, DOCX, DOC, ODT) and append its text content to chat file. Can be given the file name in-line. Will attempt to find the file relative to chat file first, then treat the given path as absolute"""
-        matching_files = self.find_matching_files_to_load(file_name)
-        if not matching_files:
-            return
-
-        for file_path in matching_files:
-            prefix = f"\nFile: {file_path}\n"
-            self.add_prompt_tag_if_needed(self.chat_name)
-            if not os.path.isdir(file_path) and self.load_document_file(file_path, prefix=prefix):
-                print(f"Loaded document file {file_path}")
 
     @cmd2.with_category(CATEGORY_CHAT_CONTROL)
     def do_LOAD_IMAGE(self, file_name):
         """Load an image file and append it to chat file. Can be given the file name in-line. Will attempt to find the file relative to chat file first, then treat the given path as absolute"""
+        from ara_cli.commands.load_image_command import LoadImageCommand
+        
         matching_files = self.find_matching_files_to_load(file_name)
         if not matching_files:
             return
@@ -585,8 +592,27 @@ Start chatting (type 'HELP'/'h' for available commands, 'QUIT'/'q' to exit chat
         for file_path in matching_files:
             prefix = f"\nFile: {file_path}\n"
             self.add_prompt_tag_if_needed(self.chat_name)
-            if not os.path.isdir(file_path) and self.load_image(file_path, prefix=prefix):
-                print(f"Loaded image file {file_path}")
+            
+            if not os.path.isdir(file_path):
+                # Determine mime type
+                file_type = None
+                file_path_lower = file_path.lower()
+                for extension, mime_type in Chat.BINARY_TYPE_MAPPING.items():
+                    if file_path_lower.endswith(extension):
+                        file_type = mime_type
+                        break
+                
+                if file_type:
+                    command = LoadImageCommand(
+                        chat_instance=self,
+                        file_path=file_path,
+                        mime_type=file_type,
+                        prefix=prefix,
+                        output=self.poutput
+                    )
+                    command.execute()
+                else:
+                    self.perror(f"File {file_path} not recognized as image, could not load")
 
     @cmd2.with_category(CATEGORY_LLM_CONTROL)
     def do_LIST_MODELS(self, _):

ara_cli-0.1.9.84/ara_cli/commands/load_command.py

@@ -0,0 +1,65 @@
+from ara_cli.commands.command import Command
+from ara_cli.file_loaders.file_loader import FileLoaderFactory
+from ara_cli.file_loaders.binary_file_loader import BinaryFileLoader
+
+
+class LoadCommand(Command):
+    def __init__(
+        self,
+        chat_instance,
+        file_path: str,
+        prefix: str = "",
+        suffix: str = "",
+        block_delimiter: str = "",
+        extract_images: bool = False,
+        output=None
+    ):
+        self.chat = chat_instance
+        self.file_path = file_path
+        self.prefix = prefix
+        self.suffix = suffix
+        self.block_delimiter = block_delimiter
+        self.extract_images = extract_images
+        self.output = output or print
+
+    def execute(self) -> bool:
+        loader = FileLoaderFactory.create_loader(self.file_path, self.chat)
+
+        if isinstance(loader, BinaryFileLoader):
+            # Determine mime type for binary files
+            file_name_lower = self.file_path.lower()
+            mime_type = None
+            for extension, mt in FileLoaderFactory.BINARY_TYPE_MAPPING.items():
+                if file_name_lower.endswith(extension):
+                    mime_type = mt
+                    break
+
+            if not mime_type:
+                self.output(
+                    f"Could not determine mime type for {self.file_path}")
+                return False
+
+            success = loader.load(
+                self.file_path,
+                mime_type=mime_type,
+                prefix=self.prefix,
+                suffix=self.suffix
+            )
+        elif hasattr(loader, 'load'):
+            success = loader.load(
+                self.file_path,
+                prefix=self.prefix,
+                suffix=self.suffix,
+                block_delimiter=self.block_delimiter,
+                extract_images=self.extract_images
+            )
+        else:
+            return False
+
+        if success:
+            if self.extract_images and not isinstance(loader, BinaryFileLoader):
+                self.output(
+                    f"Loaded contents of file {self.file_path} with images extracted")
+            else:
+                self.output(f"Loaded contents of file {self.file_path}")
+        return success

ara_cli-0.1.9.84/ara_cli/commands/load_image_command.py

@@ -0,0 +1,34 @@
+from ara_cli.commands.command import Command
+from ara_cli.file_loaders.binary_file_loader import BinaryFileLoader
+
+
+class LoadImageCommand(Command):
+    def __init__(
+        self,
+        chat_instance,
+        file_path: str,
+        mime_type: str,
+        prefix: str = "",
+        suffix: str = "",
+        output=None
+    ):
+        self.chat = chat_instance
+        self.file_path = file_path
+        self.mime_type = mime_type
+        self.prefix = prefix
+        self.suffix = suffix
+        self.output = output or print
+
+    def execute(self) -> bool:
+        loader = BinaryFileLoader(self.chat)
+        success = loader.load(
+            self.file_path,
+            mime_type=self.mime_type,
+            prefix=self.prefix,
+            suffix=self.suffix
+        )
+
+        if success:
+            self.output(f"Loaded image file {self.file_path}")
+
+        return success

{ara_cli-0.1.9.81 → ara_cli-0.1.9.84}/ara_cli/prompt_handler.py

@@ -15,6 +15,8 @@ import glob
 import logging
 
 
+
+
 class LLMSingleton:
     _instance = None
     _model = None
@@ -105,6 +107,59 @@ def send_prompt(prompt):
         yield chunk
 
 
+def describe_image(image_path: str) -> str:
+    """
+    Send an image to the LLM and get a text description.
+    
+    Args:
+        image_path: Path to the image file
+        
+    Returns:
+        Text description of the image
+    """
+    import base64
+    
+    # Read and encode the image
+    with open(image_path, 'rb') as image_file:
+        base64_image = base64.b64encode(image_file.read()).decode('utf-8')
+    
+    # Determine image type
+    image_extension = os.path.splitext(image_path)[1].lower()
+    mime_type = {
+        '.png': 'image/png',
+        '.jpg': 'image/jpeg',
+        '.jpeg': 'image/jpeg',
+        '.gif': 'image/gif',
+        '.bmp': 'image/bmp'
+    }.get(image_extension, 'image/png')
+    
+    # Create message with image
+    message = {
+        "role": "user",
+        "content": [
+            {
+                "type": "text",
+                "text": "Please describe this image in detail. If it contains text, transcribe it exactly. If it's a diagram or chart, explain its structure and content. If it's a photo or illustration, describe what you see."
+            },
+            {
+                "type": "image_url",
+                "image_url": {
+                    "url": f"data:{mime_type};base64,{base64_image}"
+                }
+            }
+        ]
+    }
+    
+    # Get response from LLM
+    response_text = ""
+    for chunk in send_prompt([message]):
+        chunk_content = chunk.choices[0].delta.content
+        if chunk_content:
+            response_text += chunk_content
+    
+    return response_text.strip()
+
+
 def append_headings(classifier, param, heading_name):
     sub_directory = Classifier.get_sub_directory(classifier)
 

{ara_cli-0.1.9.81 → ara_cli-0.1.9.84}/ara_cli/version.py

@@ -1,2 +1,2 @@
 # version.py
-__version__ = "0.1.9.81"   # fith parameter like .0 for local install test purposes only. official numbers should be 4 digit numbers
+__version__ = "0.1.9.84"   # fith parameter like .0 for local install test purposes only. official numbers should be 4 digit numbers

{ara_cli-0.1.9.81 → ara_cli-0.1.9.84}/ara_cli.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ara_cli
-Version: 0.1.9.81
+Version: 0.1.9.84
 Summary: Powerful, open source command-line tool for managing, structuring and automating software development artifacts in line with Business-Driven Development (BDD) and AI-assisted processes
 Description-Content-Type: text/markdown
 Requires-Dist: litellm

{ara_cli-0.1.9.81 → ara_cli-0.1.9.84}/ara_cli.egg-info/SOURCES.txt

@@ -60,6 +60,8 @@ ara_cli/artefact_models/vision_artefact_model.py
 ara_cli/commands/__init__.py
 ara_cli/commands/command.py
 ara_cli/commands/extract_command.py
+ara_cli/commands/load_command.py
+ara_cli/commands/load_image_command.py
 ara_cli/templates/agile.artefacts
 ara_cli/templates/template.businessgoal.prompt_log.md
 ara_cli/templates/template.capability.prompt_log.md