ara-cli 0.1.10.1__py3-none-any.whl → 0.1.11.0__py3-none-any.whl

ara_cli/chat_agent/agent_workspace_manager.py

@@ -0,0 +1,76 @@
+from ara_cli.error_handler import AraError
+
+import os
+
+
+class AgentWorkspaceManager:
+    """Manages workspace directories for agent execution."""
+
+    @staticmethod
+    def determine_agent_workspace(chat_instance, artefact_classifier=None, artefact_name=None):
+        """
+        Determines the appropriate workspace directory for agent logs and data.
+
+        Args:
+            chat_instance: The Chat instance
+            artefact_classifier: Optional artefact classifier (e.g., 'feature')
+            artefact_name: Optional artefact name
+
+        Returns:
+            str: Absolute path to the agent workspace directory (for logs)
+        """
+        base_directory = os.path.dirname(chat_instance.chat_name)
+
+        if artefact_classifier and artefact_name:
+            # Use artefact-specific workspace for logs
+            from ara_cli.classifier import Classifier
+            classifier_dir = Classifier.get_sub_directory(artefact_classifier)
+            if not classifier_dir:
+                raise AraError(f"Unknown classifier: {artefact_classifier}")
+
+            workspace_dir = os.path.join(
+                base_directory,
+                "ara",
+                classifier_dir,
+                f"{artefact_name}.data"
+            )
+        else:
+            # Use chat-specific workspace for logs
+            chat_name_without_ext = os.path.splitext(
+                os.path.basename(chat_instance.chat_name))[0]
+            workspace_dir = os.path.join(
+                base_directory,
+                "ara",
+                f"{chat_name_without_ext}"
+            )
+
+        # Ensure workspace directory exists
+        os.makedirs(workspace_dir, exist_ok=True)
+        return os.path.abspath(workspace_dir)
+
+    @staticmethod
+    def determine_base_work_dir(chat_instance):
+        """
+        Determines the base working directory (project root).
+
+        Args:
+            chat_instance: The Chat instance
+
+        Returns:
+            str: Absolute path to the project root directory
+        """
+
+        return os.path.dirname("./")
+        # The bwd should be the project root, not inside ara/
+        chat_dir = os.path.dirname(chat_instance.chat_name)
+
+        # Find project root by looking for 'ara' directory
+        current_dir = chat_dir
+        while True:
+            if os.path.isdir(os.path.join(current_dir, "ara")):
+                return os.path.abspath(current_dir)
+            parent_dir = os.path.dirname(current_dir)
+            if parent_dir == current_dir:  # Reached filesystem root
+                # Fallback to chat directory if no 'ara' folder found
+                return os.path.abspath(chat_dir)
+            current_dir = parent_dir

ara_cli/directory_navigator.py

@@ -1,4 +1,5 @@
 import os
+import sys
 from os.path import join, exists, isdir, dirname, basename
 # from ara_cli.directory_searcher import DirectorySearcher
 
@@ -23,7 +24,8 @@ class DirectoryNavigator:
             return original_directory
 
         current_directory = original_directory
-        while current_directory != dirname(current_directory):  # Ensure loop breaks at root
+        # Ensure loop breaks at root
+        while current_directory != dirname(current_directory):
             potential_path = join(current_directory, self.target_directory)
             if self.exists(potential_path):
                 os.chdir(potential_path)
@@ -31,7 +33,8 @@ class DirectoryNavigator:
             current_directory = dirname(current_directory)
 
         # If the loop completes, the target directory was not found
-        user_input = input(f"Unable to locate the '{self.target_directory}' directory. Do you want to create an 'ara' folder in the working directory? (y/N): ").strip().lower()
+        user_input = input(
+            f"Unable to locate the '{self.target_directory}' directory. Do you want to create an 'ara' folder in the working directory? (y/N): ").strip().lower()
 
         if user_input == '' or user_input == 'y':
             ara_folder_path = join(original_directory, 'ara')
@@ -40,7 +43,8 @@ class DirectoryNavigator:
             os.chdir(ara_folder_path)
             return original_directory
         else:
-            print(f"Unable to locate the '{self.target_directory}' directory and user declined to create 'ara' folder.")
+            print(
+                f"Unable to locate the '{self.target_directory}' directory and user declined to create 'ara' folder.")
             sys.exit(0)
 
     def navigate_to_relative(self, relative_path):
@@ -56,7 +60,36 @@ class DirectoryNavigator:
         if self.exists(path):
             os.chdir(path)
         else:
-            raise Exception(f"Unable to navigate to '{relative_path}' relative to the target directory.")
+            raise Exception(
+                f"Unable to navigate to '{relative_path}' relative to the target directory.")
+
+    @staticmethod
+    def find_ara_directory_root():
+        """Find the root ara directory by traversing up the directory tree."""
+        current_dir = os.getcwd()
+
+        # Check if we're already inside an ara directory structure
+        path_parts = current_dir.split(os.sep)
+
+        # Look for 'ara' in the path parts
+        if 'ara' in path_parts:
+            ara_index = path_parts.index('ara')
+            # Reconstruct path up to and including 'ara'
+            ara_root_parts = path_parts[:ara_index + 1]
+            potential_ara_root = os.sep.join(ara_root_parts)
+            if os.path.exists(potential_ara_root) and os.path.isdir(potential_ara_root):
+                return potential_ara_root
+
+        # If not inside ara directory, check current directory and parents
+        check_dir = current_dir
+        # Stop at filesystem root
+        while check_dir != os.path.dirname(check_dir):
+            ara_path = os.path.join(check_dir, 'ara')
+            if os.path.exists(ara_path) and os.path.isdir(ara_path):
+                return ara_path
+            check_dir = os.path.dirname(check_dir)
+
+        return None
 
     # debug version
     # def get_ara_directory(self):

ara_cli/file_loaders/text_file_loader.py

@@ -19,14 +19,14 @@ class TextFileLoader(FileLoader):
 
         if is_md_file and extract_images:
             reader = MarkdownReader(file_path)
-            file_content = reader.read(extract_images=True)
+            file_content = reader.read(extract_images=True).replace('\r\n', '\n')
         else:
             # Use charset-normalizer to detect encoding
             encoded_content = from_path(file_path).best()
             if not encoded_content:
                 print(f"Failed to detect encoding for {file_path}")
                 return False
-            file_content = str(encoded_content)
+            file_content = str(encoded_content).replace('\r\n', '\n')
 
         if block_delimiter:
             file_content = f"{block_delimiter}\n{file_content}\n{block_delimiter}"

ara_cli/global_file_lister.py

@@ -2,22 +2,19 @@ import os
 import fnmatch
 from typing import List, Dict, Any
 
-# Ağaç yapımız için bir tip tanımı yapalım
 DirTree = Dict[str, Any]
 
 def _build_tree(root_path: str, patterns: List[str]) -> DirTree:
-    """Belirtilen yoldaki dizin yapısını temsil eden iç içe bir sözlük oluşturur."""
+    """Creates a nested dictionary representing the directory structure in the specified path."""
     tree: DirTree = {'files': [], 'dirs': {}}
     try:
         for item in os.listdir(root_path):
             item_path = os.path.join(root_path, item)
             if os.path.isdir(item_path):
                 subtree = _build_tree(item_path, patterns)
-                # Sadece içinde dosya olan veya dosyası olan alt klasörleri ekle
                 if subtree['files'] or subtree['dirs']:
                     tree['dirs'][item] = subtree
             elif os.path.isfile(item_path):
-                # Dosyanın verilen desenlerden herhangi biriyle eşleşip eşleşmediğini kontrol et
                 if any(fnmatch.fnmatch(item, pattern) for pattern in patterns):
                     tree['files'].append(item)
     except OSError as e:
@@ -25,23 +22,17 @@ def _build_tree(root_path: str, patterns: List[str]) -> DirTree:
     return tree
 
 def _write_tree_to_markdown(md_file, tree: DirTree, level: int):
-    """Ağaç veri yapısını markdown formatında dosyaya yazar."""
-    # Dosyaları girintili olarak yaz
+    """Writes the tree data structure to the file in markdown format."""
     indent = '    ' * level
     for filename in sorted(tree['files']):
         md_file.write(f"{indent}- [] {filename}\n")
     
-    # Alt dizinler için başlık oluştur ve recursive olarak devam et
     for dirname, subtree in sorted(tree['dirs'].items()):
-        # Alt başlıklar için girinti yok, sadece başlık seviyesi artıyor
         md_file.write(f"{'    ' * (level -1)}{'#' * (level + 1)} {dirname}\n")
         _write_tree_to_markdown(md_file, subtree, level + 1)
 
 def generate_global_markdown_listing(directories: List[str], file_patterns: List[str], output_file: str):
-    """
-    Global dizinler için hiyerarşik bir markdown dosya listesi oluşturur.
-    En üst başlık olarak mutlak yolu kullanır, alt öğeler için göreceli isimler kullanır.
-    """
+    """Creates a hierarchical list of markdown files for global directories. Uses the absolute path as the top heading and relative names for children."""
     with open(output_file, 'w', encoding='utf-8') as md_file:
         for directory in directories:
             abs_dir = os.path.abspath(directory)
@@ -49,12 +40,11 @@ def generate_global_markdown_listing(directories: List[str], file_patterns: List
             if not os.path.isdir(abs_dir):
                 print(f"Warning: Global directory not found: {abs_dir}")
                 md_file.write(f"# {directory}\n")
-                md_file.write(f"    - !! UYARI: Dizin bulunamadı: {abs_dir}\n\n")
+                md_file.write(f"    - !! Warning: Global directory not found: {abs_dir}\n\n")
                 continue
 
             tree = _build_tree(abs_dir, file_patterns)
-            
-            # Sadece ağaç boş değilse yaz
+
             if tree['files'] or tree['dirs']:
                 md_file.write(f"# {abs_dir}\n")
                 _write_tree_to_markdown(md_file, tree, 1)

ara_cli/prompt_extractor.py

@@ -9,90 +9,199 @@ from ara_cli.directory_navigator import DirectoryNavigator
 from ara_cli.artefact_models.artefact_mapping import title_prefix_to_artefact_class
 
 
-def extract_code_blocks_md(markdown_text):
-    md = MarkdownIt()
-    tokens = md.parse(markdown_text)
-    code_blocks = [token.content for token in tokens if token.type == 'fence']
-    return code_blocks
+def _find_extract_token(tokens):
+    """Find the first token that needs to be processed."""
+    for token in tokens:
+        if token.type == 'fence' and token.content.strip().startswith("# [x] extract"):
+            return token
+    return None
+
+
+def _extract_file_path(content_lines):
+    """Extract file path from content lines."""
+    if not content_lines:
+        return None
+    file_path_search = re.search(r"# filename: (.+)", content_lines[0])
+    return file_path_search.group(1).strip() if file_path_search else None
+
+
+def _find_artefact_class(content_lines):
+    """Find the appropriate artefact class from content lines."""
+    for line in content_lines[:2]:
+        words = line.strip().split(' ')
+        if not words:
+            continue
+        first_word = words[0]
+        if first_word in title_prefix_to_artefact_class:
+            return title_prefix_to_artefact_class[first_word]
+    return None
 
 
-def extract_responses(document_path, relative_to_ara_root=False, force=False, write=False):
-    print(f"Starting extraction from '{document_path}'")
-    block_extraction_counter = 0
+def _process_file_extraction(file_path, code_content, force, write):
+    """Process file extraction logic."""
+    print(f"Filename extracted: {file_path}")
+    handle_existing_file(file_path, code_content, force, write)
 
-    with open(document_path, 'r', encoding='utf-8', errors='replace') as file:
-        content = file.read()
-
-    cwd = os.getcwd()
-    if relative_to_ara_root:
-        navigator = DirectoryNavigator()
-        navigator.navigate_to_target()
-        os.chdir('..')
 
-    code_blocks_found = extract_code_blocks_md(content)
-    updated_content = content
+def _process_artefact_extraction(artefact_class, content_lines, force, write):
+    """Process artefact extraction logic."""
+    artefact = artefact_class.deserialize('\n'.join(content_lines))
+    serialized_artefact = artefact.serialize()
 
-    for block in code_blocks_found:
-        block_lines = block.split('\n')
+    original_directory = os.getcwd()
+    directory_navigator = DirectoryNavigator()
+    directory_navigator.navigate_to_target()
 
-        if "# [x] extract" not in block_lines[0]:
-            continue
-        print("Block found and processed.")
+    artefact_path = artefact.file_path
+    directory = os.path.dirname(artefact_path)
+    os.makedirs(directory, exist_ok=True)
+    handle_existing_file(artefact_path, serialized_artefact, force, write)
 
-        block_lines = block_lines[1:]
+    os.chdir(original_directory)
 
-        file_path_search = re.search(r"# filename: (.+)", block_lines[0])
 
-        if file_path_search:
-            file_path = file_path_search.group(1).strip()
-            print(f"Filename extracted: {file_path}")
+def _perform_extraction_for_block(source_lines, block_start, block_end, force, write):
+    """Helper function to process a single, identified block."""
+    original_block_text = '\n'.join(source_lines[block_start:block_end + 1])
+    block_content_lines = source_lines[block_start + 1:block_end]
+    block_content = '\n'.join(block_content_lines)
 
-            block_lines = block_lines[1:]  # Remove first line again after removing filename line
-            block = '\n'.join(block_lines)
+    block_lines = block_content.split('\n')
+    content_lines_after_extract = block_lines[1:]
 
-            handle_existing_file(file_path, block, force, write)
-            block_extraction_counter += 1
+    file_path = _extract_file_path(content_lines_after_extract)
 
-            # Update the markdown content
-            updated_content = update_markdown(content, block, file_path)
+    if file_path:
+        code_content = '\n'.join(content_lines_after_extract[1:])
+        _process_file_extraction(file_path, code_content, force, write)
+    else:
+        artefact_class = _find_artefact_class(content_lines_after_extract)
+        if artefact_class:
+            _process_artefact_extraction(artefact_class, content_lines_after_extract, force, write)
         else:
-            # Extract artefact
-            artefact_class = None
-            for line in block_lines[:2]:
-                words = line.strip().split(' ')
-                if not words:
-                    continue
-                first_word = words[0]
-                if first_word not in title_prefix_to_artefact_class:
-                    continue
-                artefact_class = title_prefix_to_artefact_class[first_word]
-            if not artefact_class:
-                print("No filename found, skipping this block.")
+            print("No filename or valid artefact found, skipping processing for this block.")
+            return None, None
+
+    modified_block_text = original_block_text.replace("# [x] extract", "# [v] extract", 1)
+    return original_block_text, modified_block_text
+
+
+class FenceDetector:
+    """Helper class to detect and match fence blocks."""
+    
+    def __init__(self, source_lines):
+        self.source_lines = source_lines
+    
+    def is_extract_fence(self, line_num):
+        """Check if line is a fence with extract marker."""
+        line = self.source_lines[line_num]
+        stripped_line = line.strip()
+        
+        is_fence = stripped_line.startswith('```') or stripped_line.startswith('~~~')
+        if not is_fence:
+            return False
+            
+        if not (line_num + 1 < len(self.source_lines)):
+            return False
+            
+        return self.source_lines[line_num + 1].strip().startswith("# [x] extract")
+    
+    def find_matching_fence_end(self, start_line):
+        """Find the matching end fence for a given start fence."""
+        fence_line = self.source_lines[start_line]
+        indentation = len(fence_line) - len(fence_line.lstrip())
+        stripped_fence_line = fence_line.strip()
+        fence_char = stripped_fence_line[0]
+        fence_length = len(stripped_fence_line) - len(stripped_fence_line.lstrip(fence_char))
+
+        for i in range(start_line + 1, len(self.source_lines)):
+            scan_line = self.source_lines[i]
+            stripped_scan_line = scan_line.strip()
+
+            if not stripped_scan_line or stripped_scan_line[0] != fence_char:
+                continue
+            
+            if not all(c == fence_char for c in stripped_scan_line):
                 continue
-            artefact = artefact_class.deserialize('\n'.join(block_lines))
-            serialized_artefact = artefact.serialize()
 
-            original_directory = os.getcwd()
-            directory_navigator = DirectoryNavigator()
-            directory_navigator.navigate_to_target()
+            candidate_indentation = len(scan_line) - len(scan_line.lstrip())
+            candidate_length = len(stripped_scan_line)
+            
+            if candidate_length == fence_length and candidate_indentation == indentation:
+                return i
+        
+        return -1
+
+
+def _process_document_blocks(source_lines, force, write):
+    """Process all extract blocks in the document."""
+    fence_detector = FenceDetector(source_lines)
+    replacements = []
+    line_num = 0
+    
+    while line_num < len(source_lines):
+        if not fence_detector.is_extract_fence(line_num):
+            line_num += 1
+            continue
 
-            artefact_path = artefact.file_path
-            directory = os.path.dirname(artefact_path)
-            os.makedirs(directory, exist_ok=True)
-            handle_existing_file(artefact_path, serialized_artefact, force, write)
+        block_start_line = line_num
+        block_end_line = fence_detector.find_matching_fence_end(block_start_line)
+        
+        if block_end_line != -1:
+            print(f"Block found and processed starting on line {block_start_line + 1}.")
+            original, modified = _perform_extraction_for_block(
+                source_lines, block_start_line, block_end_line, force, write
+            )
+            if original and modified:
+                replacements.append((original, modified))
+            line_num = block_end_line + 1
+        else:
+            line_num += 1
+    
+    return replacements
 
-            os.chdir(original_directory)
 
-            # TODO: make update_markdown work block by block instead of updating the whole document at once
-            block_extraction_counter += 1
-            updated_content = update_markdown(content, block, None)
+def _apply_replacements(content, replacements):
+    """Apply all replacements to the content."""
+    updated_content = content
+    for original, modified in replacements:
+        updated_content = updated_content.replace(original, modified, 1)
+    return updated_content
+
+
+def _setup_working_directory(relative_to_ara_root):
+    """Setup working directory and return original cwd."""
+    cwd = os.getcwd()
+    if relative_to_ara_root:
+        navigator = DirectoryNavigator()
+        navigator.navigate_to_target()
+        os.chdir('..')
+    return cwd
+
+
+def extract_responses(document_path, relative_to_ara_root=False, force=False, write=False):
+    print(f"Starting extraction from '{document_path}'")
+    
+    try:
+        with open(document_path, 'r', encoding='utf-8', errors='replace') as file:
+            content = file.read()
+    except FileNotFoundError:
+        print(f"Error: File not found at '{document_path}'. Skipping extraction.")
+        return
+
+    cwd = _setup_working_directory(relative_to_ara_root)
+    
+    source_lines = content.split('\n')
+    replacements = _process_document_blocks(source_lines, force, write)
+    
+    updated_content = _apply_replacements(content, replacements)
 
     os.chdir(cwd)
-    # Save the updated markdown content
     with open(document_path, 'w', encoding='utf-8') as file:
         file.write(updated_content)
 
-    print(f"End of extraction. Found {block_extraction_counter} blocks.")
+    if replacements:
+        print(f"End of extraction. Found and processed {len(replacements)} blocks in '{os.path.basename(document_path)}'.")
 
 
 def modify_and_save_file(response, file_path):
@@ -138,7 +247,9 @@ def create_file_if_not_exist(filename, content, skip_query=False):
         if not os.path.exists(filename):
             if determine_should_create(skip_query):
                 # Ensure the directory exists
-                os.makedirs(os.path.dirname(filename), exist_ok=True)
+                dir_name = os.path.dirname(filename)
+                if dir_name:
+                    os.makedirs(dir_name, exist_ok=True)
 
                 with open(filename, 'w', encoding='utf-8') as file:
                     file.write(content)
@@ -185,7 +296,12 @@ def create_prompt_for_file_modification(content_str, filename):
 def handle_existing_file(filename, block_content, skip_query=False, write=False):
     if not os.path.isfile(filename):
         print(f"File {filename} does not exist, attempting to create")
+        # Ensure directory exists before writing
+        directory = os.path.dirname(filename)
+        if directory:
+            os.makedirs(directory, exist_ok=True)
         create_file_if_not_exist(filename, block_content, skip_query)
+
     elif write:
         print(f"File {filename} exists. Overwriting without LLM merge as requested.")
         try:
@@ -219,12 +335,4 @@ def extract_and_save_prompt_results(classifier, param, write=False):
     prompt_log_file = f"ara/{sub_directory}/{param}.data/{classifier}.prompt_log.md"
     print(f"Extract marked sections from: {prompt_log_file}")
 
-    extract_responses(prompt_log_file, write=write)
-
-
-def update_markdown(original_content, block_content, filename):
-    """
-    Update the markdown content by changing the extract block from "# [x] extract" to "# [v] extract"
-    """
-    updated_content = original_content.replace("# [x] extract", "# [v] extract")
-    return updated_content
+    extract_responses(prompt_log_file, write=write)