markitect-main/markitect/plugins/builtin/markdown_commands.py

"""
Markdown commands plugin for MarkiTect.

This plugin provides the core markdown file operations with md- prefixes,
using the new explode-implode variant system for enhanced functionality.
"""

import click
import json
import os
import re
import tempfile
import unicodedata
from pathlib import Path
from typing import Dict, Any

from markitect.plugins.base import CommandPlugin, PluginMetadata, PluginType
from markitect.plugins.decorators import register_plugin
# DocumentManager removed - using CleanDocumentManager directly
from markitect.serializer import ASTSerializer

# Try to load themes from the new modular system
try:
    from markitect.themes import get_layered_themes
    _MODULAR_THEMES = get_layered_themes()
except ImportError:
    _MODULAR_THEMES = {}

# Simple helper function - avoiding circular imports
def get_default_format(available_formats=['table', 'json', 'yaml', 'simple'], fallback='simple'):
    """Get the default output format - simplified version for plugin."""
    return fallback


# Layered theme system - themes can be combined across different scopes
LAYERED_THEMES = {
    # Mode Themes - Light/dark color schemes
    'light': {
        'scope': 'mode',
        'properties': {
            'body_background': '#ffffff',
            'body_color': '#333333',
            'heading_color': '#24292f',
            'code_background': '#f6f8fa',
            'code_color': '#24292e',
            'border_color': '#d0d7de',
            'blockquote_border': '#dfe2e5',
            'blockquote_color': '#6a737d',
            'table_border': '#d0d7de',
            'table_header_bg': '#f6f8fa',
            'link_color': '#0969da',
            'link_hover_color': '#0550ae'
        }
    },
    'dark': {
        'scope': 'mode',
        'properties': {
            'body_background': '#0d1117',
            'body_color': '#e1e4e8',
            'heading_color': '#58a6ff',
            'code_background': '#161b22',
            'code_color': '#e1e4e8',
            'border_color': '#30363d',
            'blockquote_border': '#58a6ff',
            'blockquote_color': '#8b949e',
            'table_border': '#30363d',
            'table_header_bg': '#161b22',
            'link_color': '#79c0ff',
            'link_hover_color': '#a5d6ff'
        }
    },

    # UI Themes - Editor interface elements (floating panels, buttons, editing frames)
    'standard': {
        'scope': 'ui',
        'properties': {
            'editor_panel_bg': '#f8f9fa',
            'editor_panel_border': '#dee2e6',
            'editor_button_bg': '#ffffff',
            'editor_button_hover': '#e9ecef',
            'editor_button_active': '#dee2e6',
            'editor_text_color': '#212529',
            'editor_focus_color': '#0066cc',
            'editor_shadow': 'rgba(0,0,0,0.1)',
            'editor_danger_button': '#dc3545',
            'editor_danger_button_hover': '#c82333',
            'editor_secondary_button': '#6c757d',
            'editor_secondary_button_hover': '#545b62',
            'editor_warning_bg': '#fff3cd',
            'editor_warning_border': '#ffeaa7',
            'editor_warning_text': '#856404'
        }
    },
    'greyscale': {
        'scope': 'ui',
        'properties': {
            'editor_panel_bg': '#f5f5f5',
            'editor_panel_border': '#d0d0d0',
            'editor_button_bg': '#ffffff',
            'editor_button_hover': '#e8e8e8',
            'editor_button_active': '#d4d4d4',
            'editor_text_color': '#333333',
            'editor_focus_color': '#666666',
            'editor_shadow': 'rgba(0,0,0,0.15)',
            'editor_accept_bg': '#888888',
            'editor_accept_hover': '#777777',
            'editor_cancel_bg': '#999999',
            'editor_cancel_hover': '#808080',
            'editor_danger_button': '#8b0000',
            'editor_danger_button_hover': '#700000',
            'editor_secondary_button': '#666666',
            'editor_secondary_button_hover': '#555555',
            'editor_warning_bg': '#f0f0f0',
            'editor_warning_border': '#cccccc',
            'editor_warning_text': '#555555'
        }
    },
    'electric': {
        'scope': 'ui',
        'properties': {
            'editor_panel_bg': '#001122',
            'editor_panel_border': '#00ffff',
            'editor_button_bg': '#003366',
            'editor_button_hover': '#0066cc',
            'editor_button_active': '#0099ff',
            'editor_text_color': '#00ffff',
            'editor_focus_color': '#ffff00',
            'editor_shadow': '0 0 20px rgba(0,255,255,0.5), 0 0 40px rgba(255,255,0,0.2)',
            'editor_danger_button': '#ff3366',
            'editor_danger_button_hover': '#ff0033',
            'editor_secondary_button': '#006699',
            'editor_secondary_button_hover': '#004d73',
            'editor_warning_bg': '#003366',
            'editor_warning_border': '#00ffff',
            'editor_warning_text': '#ffff00'
        }
    },
    'psychedelic': {
        'scope': 'ui',
        'properties': {
            'editor_panel_bg': 'linear-gradient(45deg, #ff6b35, #f7931e, #ffd23f, #06ffa5)',
            'editor_panel_border': '#ff1493',
            'editor_button_bg': 'rgba(255,255,255,0.2)',
            'editor_button_hover': 'rgba(255,20,147,0.3)',
            'editor_button_active': 'rgba(255,20,147,0.5)',
            'editor_text_color': '#ffffff',
            'editor_focus_color': '#ff1493',
            'editor_shadow': 'rgba(255,20,147,0.4)',
            'editor_danger_button': 'linear-gradient(45deg, #ff0066, #cc0044)',
            'editor_danger_button_hover': 'linear-gradient(45deg, #ff3388, #dd1155)',
            'editor_secondary_button': 'linear-gradient(45deg, #8a2be2, #4b0082)',
            'editor_secondary_button_hover': 'linear-gradient(45deg, #9932cc, #6a1a9a)',
            'editor_warning_bg': 'linear-gradient(45deg, #ffa500, #ff8c00)',
            'editor_warning_border': '#ff1493',
            'editor_warning_text': '#ffffff'
        }
    },

    # Document Themes - Typography and layout
    'basic': {
        'scope': 'document',
        'properties': {
            'font_family': '-apple-system, BlinkMacSystemFont, Segoe UI, Helvetica, Arial, sans-serif',
            'max_width': '800px',
            'heading_style': 'simple',
            'text_align': 'left'
        }
    },
    'github': {
        'scope': 'document',
        'properties': {
            'font_family': '-apple-system, BlinkMacSystemFont, Segoe UI, Roboto, Helvetica Neue, Arial, sans-serif',
            'max_width': '900px',
            'heading_style': 'underlined',
            'text_align': 'left'
        }
    },
    'academic': {
        'scope': 'document',
        'properties': {
            'font_family': 'Georgia, Times New Roman, serif',
            'max_width': '650px',
            'heading_style': 'centered',
            'text_align': 'justify',
            'link_color': '#777777',
            'link_hover_color': '#999999'
        }
    },
    'substack': {
        'scope': 'document',
        'properties': {
            'font_family': 'Spectral, Georgia, "Times New Roman", serif',
            'heading_font_family': 'Lora, -apple-system, BlinkMacSystemFont, sans-serif',
            'max_width': '680px',
            'body_background': '#FAF9F1',
            'body_color': '#333333',
            'heading_color': '#333333',
            'text_align': 'left',
            'line_height': '1.6',
            'heading_style': 'simple',
            'accent_color': '#b08d57',
            'link_color': '#b08d57',
            'link_hover_color': '#8b6c42',
            'code_background': '#f5f4ed',
            'code_color': '#333333',
            'blockquote_border': '#b08d57',
            'blockquote_color': '#666666'
        }
    },
    'chatgpt': {
        'scope': 'document',
        'properties': {
            'font_family': 'Inter, -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, sans-serif',
            'heading_font_family': 'Inter, -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, sans-serif',
            'max_width': '580px',
            'body_background': '#ffffff',
            'body_color': '#1f1f1f',
            'heading_color': '#1f1f1f',
            'text_align': 'left',
            'line_height': '1.5',
            'heading_style': 'minimal',
            'accent_color': '#10a37f',
            'link_color': '#10a37f',
            'link_hover_color': '#0d8c6d',
            'code_background': '#f7f7f7',
            'code_color': '#1f1f1f',
            'code_font_family': '"SF Mono", Monaco, Inconsolata, "Roboto Mono", Consolas, "Courier New", monospace',
            'font_size': '15px',
            'heading_margin': '1.2em 0 0.6em 0',
            'paragraph_margin': '1em 0',
            'border_radius': '8px',
            'blockquote_border': '#10a37f',
            'blockquote_color': '#6b7280'
        }
    },

    # Branding Themes - Company/personal styling
    'corporate': {
        'scope': 'branding',
        'properties': {
            'accent_color': '#0066cc',
            'secondary_color': '#f8f9fa',
            'brand_font': 'inherit'
        }
    },
    'startup': {
        'scope': 'branding',
        'properties': {
            'accent_color': '#ff6b35',
            'secondary_color': '#f4f4f4',
            'brand_font': 'inherit'
        }
    }
}

# Merge modular themes with inline themes
# Modular themes take precedence over inline themes
if _MODULAR_THEMES:
    LAYERED_THEMES.update(_MODULAR_THEMES)

# Legacy compatibility - map old theme names to new layered equivalents
LEGACY_THEME_MAPPING = {
    'basic': ['light', 'standard', 'basic'],
    'github': ['light', 'standard', 'github'],
    'dark': ['dark', 'standard', 'basic'],
    'academic': ['light', 'standard', 'academic'],
    'substack': ['light', 'standard', 'substack'],
    'chatgpt': ['light', 'standard', 'chatgpt']
}

# Keep TEMPLATE_STYLES for backward compatibility in tests
TEMPLATE_STYLES = {
    'basic': {
        'body_color': '#333',
        'font_family': '-apple-system, BlinkMacSystemFont, Segoe UI, Helvetica, Arial, sans-serif',
        'max_width': '800px'
    },
    'github': {
        'body_color': '#24292f',
        'font_family': '-apple-system, BlinkMacSystemFont, Segoe UI, Roboto, Helvetica Neue, Arial, sans-serif',
        'max_width': '900px'
    },
    'dark': {
        'body_color': '#e1e4e8',
        'font_family': '-apple-system, BlinkMacSystemFont, Segoe UI, Helvetica, Arial, sans-serif',
        'max_width': '800px'
    },
    'academic': {
        'body_color': '#333',
        'font_family': 'Georgia, Times New Roman, serif',
        'max_width': '650px'
    },
    'substack': {
        'body_color': '#333333',
        'font_family': 'Spectral, Georgia, "Times New Roman", serif',
        'max_width': '680px'
    },
    'chatgpt': {
        'body_color': '#1f1f1f',
        'font_family': 'Inter, -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, sans-serif',
        'max_width': '580px'
    }
}


def parse_theme_string(theme_string: str) -> list:
    """
    Parse theme string into list of individual themes.

    Supports:
    - Single theme: "dark"
    - Multiple themes: "dark,academic" or "dark, academic"
    - Legacy theme mapping: "basic" -> ["light", "basic"]

    Args:
        theme_string: Comma-separated theme names

    Returns:
        List of theme names in order
    """
    if not theme_string:
        return ['light', 'basic']  # Default themes

    # Split by comma and clean up whitespace
    themes = [theme.strip() for theme in theme_string.split(',')]

    # Expand legacy themes only if they don't exist in the new layered system
    expanded_themes = []
    for theme in themes:
        if theme in LAYERED_THEMES:
            # Theme exists in new system, use as-is
            expanded_themes.append(theme)
        elif theme in LEGACY_THEME_MAPPING:
            # Legacy theme, expand it
            expanded_themes.extend(LEGACY_THEME_MAPPING[theme])
        else:
            # Unknown theme, add as-is (will be warned about later)
            expanded_themes.append(theme)

    return expanded_themes


class ThemeType(click.ParamType):
    """Custom click type for theme validation."""

    name = "theme"

    def convert(self, value, param, ctx):
        if value is None:
            return value

        try:
            validate_theme_string(value)
            return value
        except click.BadParameter as e:
            self.fail(str(e), param, ctx)


def validate_theme_string(theme_string: str) -> None:
    """
    Validate that all themes in a theme string are known themes.

    Args:
        theme_string: Comma-separated theme names

    Raises:
        click.BadParameter: If any theme is unknown
    """
    if not theme_string:
        return  # Allow empty/None themes

    themes = parse_theme_string(theme_string)
    unknown_themes = []

    for theme_name in themes:
        if theme_name not in LAYERED_THEMES and theme_name not in LEGACY_THEME_MAPPING:
            unknown_themes.append(theme_name)

    if unknown_themes:
        available_themes = list(LAYERED_THEMES.keys()) + list(LEGACY_THEME_MAPPING.keys())
        raise click.BadParameter(
            f"Unknown theme(s): {', '.join(unknown_themes)}. "
            f"Available themes: {', '.join(sorted(set(available_themes)))}"
        )


def combine_theme_properties(theme_list: list) -> dict:
    """
    Combine properties from multiple themes, with later themes overriding earlier ones.

    Args:
        theme_list: List of theme names in order of application

    Returns:
        Combined properties dictionary
    """
    combined_properties = {}

    for theme_name in theme_list:
        if theme_name in LAYERED_THEMES:
            theme_data = LAYERED_THEMES[theme_name]
            # Later themes override earlier ones
            combined_properties.update(theme_data['properties'])
        elif theme_name in LEGACY_THEME_MAPPING:
            # Handle legacy themes by expanding them
            expanded_themes = LEGACY_THEME_MAPPING[theme_name]
            for expanded_theme in expanded_themes:
                if expanded_theme in LAYERED_THEMES:
                    theme_data = LAYERED_THEMES[expanded_theme]
                    combined_properties.update(theme_data['properties'])
        else:
            # This should not happen if validation is working
            print(f"Warning: Unknown theme '{theme_name}' - skipping")
    return combined_properties


def generate_html_with_embedded_markdown(markdown_content, title, theme, css_content, template_vars):
    """
    Generate HTML with embedded markdown content for testing.

    This function is used by tests to validate template functionality.
    """
    # Create a temporary document manager for rendering
    from markitect.clean_document_manager import CleanDocumentManager
    doc_manager = CleanDocumentManager(None)

    # Generate HTML template
    html_content = doc_manager._generate_html_template(
        markdown_content=markdown_content,
        title=title,
        css=css_content,
        template=theme
    )

    return html_content


# Publication directory management functions
def get_publication_directory() -> Path:
    """
    Get the publication directory path.

    Returns the path specified by MARKITECT_PUBLICATION_DIR environment variable,
    or defaults to ~/Notes if not set.
    """
    pub_dir = os.environ.get('MARKITECT_PUBLICATION_DIR')
    if pub_dir:
        return Path(pub_dir)
    return Path.home() / "Notes"


def ensure_publication_directory(pub_dir: Path) -> None:
    """
    Ensure the publication directory exists, creating it if necessary.

    Args:
        pub_dir: Path to the publication directory
    """
    pub_dir.mkdir(parents=True, exist_ok=True)


def normalize_publication_path(path_str: str) -> Path:
    """
    Normalize a publication directory path.

    Handles tilde expansion and resolves relative paths to absolute paths.

    Args:
        path_str: String path that may contain ~ or relative components

    Returns:
        Absolute Path object
    """
    path = Path(path_str).expanduser().resolve()
    return path


def get_output_filename(input_file: Path) -> str:
    """
    Get the output filename for a markdown file.

    Args:
        input_file: Path to the input markdown file

    Returns:
        Output filename with .html extension
    """
    return input_file.stem + ".html"


def find_markdown_files(directory: Path) -> list[Path]:
    """
    Find all markdown files in a directory recursively.

    Args:
        directory: Directory to search in

    Returns:
        List of Path objects for found markdown files
    """
    if not directory.exists():
        return []

    markdown_files = []
    for md_file in directory.rglob("*.md"):
        if md_file.is_file():
            markdown_files.append(md_file)

    return sorted(markdown_files)


def get_relative_output_path(source_file: Path, base_dir: Path, pub_dir: Path) -> Path:
    """
    Get the output path for a source file, preserving directory structure.

    Args:
        source_file: Path to the source markdown file
        base_dir: Base directory (to calculate relative path from)
        pub_dir: Publication directory (destination base)

    Returns:
        Full output path in publication directory
    """
    # Get relative path from base directory
    relative_path = source_file.relative_to(base_dir)
    # Change extension to .html
    html_relative = relative_path.with_suffix('.html')
    # Combine with publication directory
    return pub_dir / html_relative


def process_single_file(input_file: Path, use_publication_dir: bool, publication_dir: Path) -> Path:
    """
    Process a single markdown file.

    Args:
        input_file: Path to the input markdown file
        use_publication_dir: Whether to use publication directory
        publication_dir: Publication directory path

    Returns:
        Path to the output HTML file

    Raises:
        FileNotFoundError: If input file doesn't exist
    """
    if not input_file.exists():
        raise FileNotFoundError(f"Input file does not exist: {input_file}")

    # Determine output path
    if use_publication_dir:
        ensure_publication_directory(publication_dir)
        output_file = publication_dir / get_output_filename(input_file)
    else:
        output_file = input_file.with_suffix('.html')

    # Create document manager and render
    from markitect.clean_document_manager import CleanDocumentManager
    doc_manager = CleanDocumentManager(None)
    doc_manager.render_file(str(input_file), str(output_file))

    return output_file


def process_directory(input_dir: Path, use_publication_dir: bool, publication_dir: Path) -> list[Path]:
    """
    Process all markdown files in a directory.

    Args:
        input_dir: Directory containing markdown files
        use_publication_dir: Whether to use publication directory
        publication_dir: Publication directory path

    Returns:
        List of paths to generated HTML files
    """
    markdown_files = find_markdown_files(input_dir)
    output_files = []

    from markitect.clean_document_manager import CleanDocumentManager
    doc_manager = CleanDocumentManager(None)

    for md_file in markdown_files:
        if use_publication_dir:
            ensure_publication_directory(publication_dir)
            output_file = get_relative_output_path(md_file, input_dir, publication_dir)
            # Ensure subdirectories exist
            output_file.parent.mkdir(parents=True, exist_ok=True)
        else:
            output_file = md_file.with_suffix('.html')

        # Render the file
        doc_manager.render_file(str(md_file), str(output_file))
        output_files.append(output_file)

    return output_files


# Index generation functions
def find_html_files(directory: Path, recursive: bool = False) -> list[Path]:
    """
    Find all HTML files in a directory.

    Args:
        directory: Directory to search in
        recursive: Whether to search recursively in subdirectories

    Returns:
        List of Path objects for found HTML files
    """
    if not directory.exists():
        return []

    html_files = []
    if recursive:
        # Search recursively
        for html_file in directory.rglob("*.html"):
            if html_file.is_file():
                html_files.append(html_file)
    else:
        # Search only in current directory
        for html_file in directory.glob("*.html"):
            if html_file.is_file():
                html_files.append(html_file)

    return sorted(html_files)


def extract_html_title(html_file: Path) -> str:
    """
    Extract title from an HTML file.

    Tries to extract the title from <title> tag first, then from <h1> tag,
    and finally falls back to the filename.

    Args:
        html_file: Path to the HTML file

    Returns:
        Extracted title string
    """
    try:
        content = html_file.read_text(encoding='utf-8', errors='ignore')

        # Try to extract from <title> tag
        import re
        title_match = re.search(r'<title[^>]*>(.*?)</title>', content, re.IGNORECASE | re.DOTALL)
        if title_match:
            title = title_match.group(1).strip()
            # Clean up any HTML entities or extra whitespace
            title = re.sub(r'\s+', ' ', title)
            if title:
                return title

        # Try to extract from <h1> tag
        h1_match = re.search(r'<h1[^>]*>(.*?)</h1>', content, re.IGNORECASE | re.DOTALL)
        if h1_match:
            h1_title = h1_match.group(1).strip()
            # Remove any HTML tags within the h1
            h1_title = re.sub(r'<[^>]+>', '', h1_title)
            h1_title = re.sub(r'\s+', ' ', h1_title)
            if h1_title:
                return h1_title

    except Exception:
        # If anything goes wrong reading/parsing the file, fall back to filename
        pass

    # Fallback to filename without extension
    return html_file.stem


def generate_index_html(html_files: list, title: str, theme: str = None) -> str:
    """
    Generate HTML content for an index page.

    Args:
        html_files: List of dictionaries with 'path', 'title', and 'relative_path' keys
        title: Title for the index page
        theme: Theme to use

    Returns:
        HTML content string
    """
    # Get template CSS
    from markitect.clean_document_manager import CleanDocumentManager
    doc_manager = CleanDocumentManager(None)
    template_css = doc_manager._get_template_css(theme)

    # Generate file list HTML
    if not html_files:
        file_list_html = '<p class="no-files">No HTML files found in this directory.</p>'
    else:
        file_items = []
        for file_info in html_files:
            href = file_info['relative_path']
            link_title = file_info['title']
            file_items.append(f'        <li><a href="{href}">{link_title}</a></li>')

        file_list_html = f"""
    <ul>
{chr(10).join(file_items)}
    </ul>"""

    # Generate complete HTML
    html_content = f"""<!DOCTYPE html>
<html lang="en">
<head>
    <meta charset="utf-8">
    <meta name="viewport" content="width=device-width, initial-scale=1.0">
    <title>{title}</title>
    <style>
        {template_css}
        .file-list {{
            list-style: none;
            padding: 0;
            margin: 2rem 0;
        }}
        .file-list li {{
            margin: 0.75rem 0;
            padding: 0.5rem;
            border-left: 3px solid #007acc;
            background: rgba(0, 122, 204, 0.05);
            border-radius: 4px;
        }}
        .file-list a {{
            text-decoration: none;
            color: #007acc;
            font-weight: 500;
            display: block;
        }}
        .file-list a:hover {{
            color: #005999;
            text-decoration: underline;
        }}
        .no-files {{
            color: #666;
            font-style: italic;
            text-align: center;
            margin: 2rem 0;
            padding: 2rem;
            background: #f9f9f9;
            border-radius: 8px;
        }}
        .header {{
            border-bottom: 2px solid #eee;
            padding-bottom: 1rem;
            margin-bottom: 2rem;
        }}
        .header h1 {{
            margin: 0;
            color: #333;
        }}
        .footer {{
            margin-top: 3rem;
            padding-top: 1rem;
            border-top: 1px solid #eee;
            color: #666;
            font-size: 0.9em;
            text-align: center;
        }}
    </style>
</head>
<body>
    <div class="header">
        <h1>{title}</h1>
    </div>

    <main>
        {file_list_html}
    </main>

    <div class="footer">
        <p>Generated by MarkiTect</p>
    </div>
</body>
</html>"""

    return html_content


def process_directory_for_index(directory: Path, index_filename: str = "index.html") -> Path:
    """
    Process a directory and create an index HTML file.

    Args:
        directory: Directory to process
        index_filename: Name of the index file to create

    Returns:
        Path to the created index file

    Raises:
        FileNotFoundError: If directory doesn't exist
    """
    if not directory.exists():
        raise FileNotFoundError(f"Directory does not exist: {directory}")

    # Find all HTML files except the index file itself
    html_files = find_html_files(directory, recursive=False)

    # Create file info list, excluding the index file
    file_info_list = []
    for html_file in html_files:
        if html_file.name != index_filename:
            title = extract_html_title(html_file)
            relative_path = html_file.name  # Since we're not doing recursive, just use filename
            file_info_list.append({
                'path': html_file,
                'title': title,
                'relative_path': relative_path
            })

    # Generate index page title
    index_title = f"Index - {directory.name}"

    # Generate HTML content
    html_content = generate_index_html(file_info_list, index_title)

    # Write index file
    index_path = directory / index_filename
    index_path.write_text(html_content, encoding='utf-8')

    return index_path


# Markdown parsing functions - decoupled utilities
class MarkdownSection:
    """
    Represents a section of markdown content with hierarchical structure.

    This is a simple data class that doesn't depend on any external systems,
    making it easily reusable and testable.
    """
    def __init__(self, level: int, title: str, content: str = "", line_start: int = 0, line_end: int = 0):
        self.level = level
        self.title = title
        self.content = content
        self.line_start = line_start
        self.line_end = line_end
        self.children = []
        self.parent = None

    def add_child(self, child: 'MarkdownSection'):
        """Add a child section with hierarchy validation."""
        # Validate hierarchy - child level should be exactly one level deeper
        if child.level != self.level + 1:
            raise ValueError(f"Invalid heading hierarchy: level {child.level} cannot be child of level {self.level}")

        child.parent = self
        self.children.append(child)

    def __repr__(self):
        return f"MarkdownSection(level={self.level}, title='{self.title}', children={len(self.children)})"


def extract_headings(markdown_content: str) -> list[dict]:
    """
    Extract all headings from markdown content with their positions.

    Decoupled function that only requires markdown text as input.
    Returns a simple list of dictionaries for easy processing.

    Args:
        markdown_content: Raw markdown text

    Returns:
        List of dictionaries with 'level', 'title', and 'line' keys
    """
    import re

    headings = []
    lines = markdown_content.split('\n')

    for line_num, line in enumerate(lines):
        # Match ATX-style headings (### Title)
        heading_match = re.match(r'^(#{1,6})\s+(.+)$', line.strip())
        if heading_match:
            level = len(heading_match.group(1))
            title = heading_match.group(2).strip()
            headings.append({
                'level': level,
                'title': title,
                'line': line_num
            })

    return headings


def extract_section_content(markdown_content: str, headings: list[dict], section_index: int) -> str:
    """
    Extract content for a specific section between headings.

    Decoupled function that operates on simple data structures.

    Args:
        markdown_content: Raw markdown text
        headings: List of heading dictionaries from extract_headings()
        section_index: Index of the heading to extract content for

    Returns:
        Markdown content for the specified section
    """
    if not headings or section_index >= len(headings):
        return ""

    lines = markdown_content.split('\n')
    current_heading = headings[section_index]
    start_line = current_heading['line']

    # Find the end line (next heading at same or higher level)
    end_line = len(lines)
    current_level = current_heading['level']

    for next_heading in headings[section_index + 1:]:
        if next_heading['level'] <= current_level:
            end_line = next_heading['line']
            break

    # Extract the section content
    section_lines = lines[start_line:end_line]
    return '\n'.join(section_lines)


def parse_markdown_structure(file_path: Path) -> tuple[list[MarkdownSection], dict]:
    """
    Parse a markdown file into hierarchical structure with front matter.

    Decoupled function that works with file paths and returns simple objects.

    Args:
        file_path: Path to the markdown file

    Returns:
        Tuple of (list of root MarkdownSection objects, front_matter dict or None)
    """
    import re

    # Read file content
    try:
        content = file_path.read_text(encoding='utf-8')
    except Exception as e:
        raise FileNotFoundError(f"Could not read markdown file: {file_path}") from e

    # Extract front matter if present
    front_matter = None
    markdown_content = content

    # Check for YAML front matter
    front_matter_match = re.match(r'^---\n(.*?)\n---\n(.*)$', content, re.DOTALL)
    if front_matter_match:
        # Return raw YAML string as tests expect
        front_matter = front_matter_match.group(1)
        markdown_content = front_matter_match.group(2)

    # Extract headings
    headings = extract_headings(markdown_content)

    if not headings:
        return [], front_matter

    # Build hierarchical structure
    root_sections = []
    section_stack = []

    for i, heading in enumerate(headings):
        # Extract content for this section
        section_content = extract_section_content(markdown_content, headings, i)

        # Create section object
        section = MarkdownSection(
            level=heading['level'],
            title=heading['title'],
            content=section_content,
            line_start=heading['line']
        )

        # Find the right place in hierarchy
        while section_stack and section_stack[-1].level >= section.level:
            section_stack.pop()

        if section_stack:
            # Add as child to the last section in stack
            # Use direct assignment to handle hierarchy gaps gracefully during parsing
            parent = section_stack[-1]
            section.parent = parent
            parent.children.append(section)
        else:
            # This is a root level section
            root_sections.append(section)

        section_stack.append(section)

    return root_sections, front_matter


def title_to_filesystem_name(title: str) -> str:
    """Convert a markdown heading title to a filesystem-safe name.

    Args:
        title: The markdown heading title

    Returns:
        A filesystem-safe name (lowercase, spaces/punctuation to underscores)
    """
    import re
    # Remove any markdown formatting
    cleaned = re.sub(r'[#*`\[\](){}]', '', title)
    # Convert to lowercase
    cleaned = cleaned.lower()
    # Remove non-alphanumeric chars except spaces, hyphens, periods, colons, slashes
    cleaned = re.sub(r'[^\w\s.-:/]', '', cleaned)
    # Replace dots, spaces, hyphens, colons, and slashes with underscores
    cleaned = re.sub(r'[.\s:/\-]', '_', cleaned)
    # Collapse multiple underscores into single underscore
    cleaned = re.sub(r'_+', '_', cleaned)
    # Remove leading/trailing underscores
    cleaned = cleaned.strip('_')
    return cleaned or 'untitled'


def create_directory_structure(sections: list[MarkdownSection], target_dir: Path) -> list[Path]:
    """Create directory structure from markdown sections.

    Args:
        sections: List of root-level MarkdownSection objects
        target_dir: Target directory to create structure in

    Returns:
        List of created paths (files and directories)
    """
    target_dir = Path(target_dir)
    target_dir.mkdir(parents=True, exist_ok=True)
    created_paths = []
    used_names = set()

    def get_unique_name(base_name: str, is_file: bool = False) -> str:
        """Get a unique name, adding numeric suffix if needed."""
        extension = '.md' if is_file else ''
        name = base_name
        counter = 2
        while name + extension in used_names:
            name = f"{base_name}_{counter}"
            counter += 1
        used_names.add(name + extension)
        return name

    def create_structure_recursive(sections: list[MarkdownSection], parent_dir: Path):
        """Recursively create directory structure."""
        for section in sections:
            safe_name = title_to_filesystem_name(section.title)

            if section.children:
                # Create directory for sections with children
                unique_name = get_unique_name(safe_name)
                section_dir = parent_dir / unique_name
                section_dir.mkdir(exist_ok=True)
                created_paths.append(section_dir)

                # Create README.md for the section content if it exists
                if section.content.strip():
                    readme_path = section_dir / 'README.md'
                    readme_path.write_text(section.content)
                    created_paths.append(readme_path)

                # Recursively create children
                create_structure_recursive(section.children, section_dir)
            else:
                # Create markdown file for leaf sections
                unique_name = get_unique_name(safe_name, is_file=True)
                file_path = parent_dir / f"{unique_name}.md"
                file_path.write_text(section.content)
                created_paths.append(file_path)

    create_structure_recursive(sections, target_dir)
    return created_paths


def explode_markdown_file(input_file: Path, output_dir: Path) -> Path:
    """Explode a markdown file into a directory structure.

    Args:
        input_file: Path to input markdown file
        output_dir: Path to output directory

    Returns:
        Path to the created output directory

    Raises:
        FileNotFoundError: If input file doesn't exist
        PermissionError: If can't create output directory
    """
    input_file = Path(input_file)
    output_dir = Path(output_dir)

    if not input_file.exists():
        raise FileNotFoundError(f"Input file not found: {input_file}")

    try:
        # Parse the markdown file structure
        sections, front_matter = parse_markdown_structure(input_file)

        # Create the directory structure
        created_paths = create_directory_structure(sections, output_dir)

        # Create front matter file if present
        if front_matter:
            front_matter_file = output_dir / '_frontmatter.yml'
            front_matter_file.write_text(front_matter)

        return output_dir

    except PermissionError as e:
        raise PermissionError(f"Cannot create output directory: {e}")


class DirectoryStructureBuilder:
    """Builder class for creating directory structures from markdown sections."""

    def __init__(self, output_dir: Path = None, target_dir: Path = None,
                 max_depth: int = None, file_extension: str = '.md'):
        # Support both output_dir and target_dir for backward compatibility
        self.target_dir = Path(output_dir or target_dir)
        self.output_dir = self.target_dir  # Alias for tests
        self.max_depth = max_depth
        self.file_extension = file_extension
        self.created_paths = []

    def build(self, sections: list[MarkdownSection]) -> list[Path]:
        """Build directory structure from sections."""
        # Apply depth limiting if specified
        if self.max_depth is not None:
            sections = self._limit_depth(sections, self.max_depth)

        self.created_paths = create_directory_structure(sections, self.target_dir)
        return self.created_paths

    def _limit_depth(self, sections: list[MarkdownSection], max_depth: int) -> list[MarkdownSection]:
        """Recursively limit section depth."""
        if max_depth <= 0:
            return []

        limited_sections = []
        for section in sections:
            if section.level <= max_depth:
                # Create a shallow copy and limit children
                limited_section = MarkdownSection(
                    level=section.level,
                    title=section.title,
                    content=section.content,
                    line_start=getattr(section, 'line_start', 0),
                    line_end=getattr(section, 'line_end', 0)
                )
                if section.level < max_depth:
                    limited_section.children = self._limit_depth(section.children, max_depth)
                limited_sections.append(limited_section)

        return limited_sections


def sanitize_heading_text(heading_text: str) -> str:
    """Remove markdown formatting from heading text.

    Args:
        heading_text: Raw heading text with potential markdown formatting

    Returns:
        Clean text with markdown formatting removed
    """
    import re
    # Remove bold and italic formatting
    cleaned = re.sub(r'\*\*([^*]+)\*\*', r'\1', heading_text)  # **bold**
    cleaned = re.sub(r'\*([^*]+)\*', r'\1', cleaned)  # *italic*
    cleaned = re.sub(r'__([^_]+)__', r'\1', cleaned)  # __bold__
    cleaned = re.sub(r'_([^_]+)_', r'\1', cleaned)  # _italic_

    # Remove code formatting
    cleaned = re.sub(r'`([^`]+)`', r'\1', cleaned)  # `code`

    # Remove links but keep text
    cleaned = re.sub(r'\[([^\]]+)\]\([^)]+\)', r'\1', cleaned)  # [text](url)

    # Remove other markdown elements
    cleaned = re.sub(r'[#]+\s*', '', cleaned)  # heading markers
    cleaned = cleaned.strip()

    return cleaned


def generate_safe_filename(heading: str, max_length: int = 100) -> str:
    """Generate a filesystem-safe filename from a heading.

    Args:
        heading: The heading text to convert
        max_length: Maximum length for the filename

    Returns:
        A safe filename suitable for use across platforms
    """
    import re
    import unicodedata

    if not heading or not heading.strip():
        return 'untitled'

    # First sanitize markdown formatting
    cleaned = sanitize_heading_text(heading)

    # Normalize unicode characters (café -> cafe)
    cleaned = unicodedata.normalize('NFKD', cleaned)
    cleaned = ''.join(c for c in cleaned if not unicodedata.combining(c))

    # Convert to lowercase
    cleaned = cleaned.lower()

    # Remove non-alphanumeric chars except spaces, hyphens, periods, colons, slashes
    cleaned = re.sub(r'[^\w\s.-:/\\]', '', cleaned)

    # Replace dots, spaces, hyphens, colons, slashes, backslashes with underscores
    cleaned = re.sub(r'[.\s:/\\\-]', '_', cleaned)

    # Collapse multiple underscores into single underscore
    cleaned = re.sub(r'_+', '_', cleaned)

    # Remove leading/trailing underscores
    cleaned = cleaned.strip('_')

    # Handle empty result
    if not cleaned:
        return 'untitled'

    # Apply length limit, but try to break at word boundaries
    if len(cleaned) > max_length:
        truncated = cleaned[:max_length]
        # Find last underscore before limit
        last_underscore = truncated.rfind('_')
        if last_underscore > max_length // 2:  # Only if it's not too early
            truncated = truncated[:last_underscore]
        cleaned = truncated.rstrip('_')

    return cleaned or 'untitled'


def resolve_filename_conflicts(base_filename: str, existing_files: list[str]) -> str:
    """Resolve filename conflicts by adding numeric suffixes.

    Args:
        base_filename: The desired filename (without extension)
        existing_files: List of already existing filenames (may include extensions)

    Returns:
        A unique filename that doesn't conflict with existing ones
    """
    # Normalize existing files to remove extensions for comparison
    existing_basenames = set()
    for filename in existing_files:
        # Remove common extensions for comparison
        base = filename
        for ext in ['.md', '.txt', '.html']:
            if base.endswith(ext):
                base = base[:-len(ext)]
                break
        existing_basenames.add(base)

    if base_filename not in existing_basenames:
        return base_filename

    # Try adding numeric suffixes
    counter = 2
    while True:
        candidate = f"{base_filename}_{counter}"
        if candidate not in existing_basenames:
            return candidate
        counter += 1


class FilenameGenerator:
    """Generator for creating unique, filesystem-safe filenames from headings."""

    def __init__(self, max_length: int = 100, separator: str = '_',
                 case_style: str = 'lower', preserve_numbers: bool = False):
        self.max_length = max_length
        self.separator = separator
        self.case_style = case_style
        self.preserve_numbers = preserve_numbers
        self.used_filenames = set()

    def generate(self, heading: str) -> str:
        """Generate a unique safe filename from a heading."""
        import re

        # Handle numbered headings if preserve_numbers is enabled
        processed_heading = heading
        if self.preserve_numbers:
            # Look for patterns like "1. Introduction" or "10. Advanced Topics"
            match = re.match(r'^(\d+)\.\s*(.+)$', heading.strip())
            if match:
                number = match.group(1).zfill(2)  # Zero-pad to 2 digits
                title = match.group(2)
                processed_heading = f"{number}. {title}"

        # Use the existing generate_safe_filename function
        base_filename = generate_safe_filename(processed_heading, self.max_length)

        # Apply case style and separator customization
        if self.case_style == 'camel':
            # For camelCase, split on underscores, capitalize each word after first, join without separator
            parts = base_filename.split('_')
            if parts:
                camel_cased = parts[0].lower()
                for part in parts[1:]:
                    if part:
                        camel_cased += part.capitalize()
                base_filename = camel_cased
        else:
            # Apply separator customization for other styles
            if self.separator != '_':
                base_filename = base_filename.replace('_', self.separator)

            # Apply case style
            if self.case_style == 'upper':
                base_filename = base_filename.upper()
            elif self.case_style == 'title':
                base_filename = base_filename.title().replace(self.separator, self.separator.lower())
            # 'lower' is already default

        unique_filename = resolve_filename_conflicts(base_filename, list(self.used_filenames))
        self.used_filenames.add(unique_filename)
        return unique_filename

    def reset(self):
        """Reset the internal state of used filenames."""
        self.used_filenames.clear()


class ImplodeOptions:
    """Options for the implode operation."""

    def __init__(self, input_dir: Path = None, output_file: Path = None,
                 preserve_front_matter: bool = True, section_spacing: int = 2,
                 overwrite: bool = False, dry_run: bool = False, verbose: bool = False,
                 preserve_heading_levels: bool = False, include_readme_files: bool = False):
        self.input_dir = input_dir
        self.output_file = output_file
        self.preserve_front_matter = preserve_front_matter
        self.section_spacing = section_spacing
        self.overwrite = overwrite
        self.dry_run = dry_run
        self.verbose = verbose
        self.preserve_heading_levels = preserve_heading_levels
        self.include_readme_files = include_readme_files


class ValidationResult:
    """Result of validation operation."""
    def __init__(self, is_valid: bool, errors: list = None):
        self.is_valid = is_valid
        self.errors = errors or []


def validate_implode_arguments(options: ImplodeOptions) -> ValidationResult:
    """Validate arguments for the implode operation.

    Args:
        options: Implode options

    Returns:
        ValidationResult with is_valid flag and any errors
    """
    errors = []

    if not options.input_dir:
        errors.append("Input directory is required")
    elif not options.input_dir.exists():
        errors.append(f"Input directory does not exist: {options.input_dir}")
    elif not options.input_dir.is_dir():
        errors.append(f"Input path is not a directory: {options.input_dir}")

    if options.output_file and not options.overwrite:
        try:
            if options.output_file.exists():
                errors.append(f"Output file already exists: {options.output_file}")
        except (PermissionError, OSError) as e:
            errors.append(f"Cannot access output file: {e}")

    return ValidationResult(is_valid=len(errors) == 0, errors=errors)


class ImplodeResult:
    """Result of implode operation."""
    def __init__(self, success: bool, output_file: Path = None, errors: list = None,
                 preview: str = None, processing_info: list = None):
        self.success = success
        self.output_file = output_file
        self.errors = errors or []
        self.preview = preview
        self.processing_info = processing_info or []

    @property
    def error_message(self) -> str:
        """Get the first error message or None."""
        return self.errors[0] if self.errors else None


def cli_implode_directory(input_dir: Path = None, output_file: Path = None,
                         options: ImplodeOptions = None, dry_run: bool = False,
                         verbose: bool = False, overwrite: bool = False, **kwargs) -> ImplodeResult:
    """Implode a directory structure back into a markdown file using variant system.

    Args:
        input_dir: Directory containing markdown files to implode
        options: Options for the implode operation
        output_file: Output file path (alternative to options.output_file)
        dry_run: Preview mode without creating files
        verbose: Provide detailed processing information
        overwrite: Overwrite existing output file
        **kwargs: Additional arguments for compatibility

    Returns:
        ImplodeResult with success flag and output file path (legacy format)
    """
    from markitect.explode_variants import get_variant_factory

    # Handle different calling patterns
    if options is None:
        options = ImplodeOptions(
            output_file=output_file,
            preserve_front_matter=True,
            section_spacing=2,
            dry_run=dry_run
        )
    else:
        # Update options with any provided keyword arguments
        if output_file and not options.output_file:
            options.output_file = output_file
        if dry_run:
            options.dry_run = dry_run

    # Determine input directory
    if input_dir is None:
        return ImplodeResult(success=False, errors=["Input directory is required"])

    input_dir = Path(input_dir)
    if not input_dir.exists() or not input_dir.is_dir():
        return ImplodeResult(success=False, errors=[f"Input directory does not exist: {input_dir}"])

    # Determine output file
    if options.output_file is None:
        options.output_file = input_dir.parent / f"{input_dir.name}_imploded.md"

    processing_info = []
    preview_content = None

    try:
        # Use variant factory to auto-detect and implode
        factory = get_variant_factory()

        # Detect variant from directory structure
        detection_result = factory.detect_variant(input_dir)

        processing_info.append(f"Processing directory: {input_dir}")
        processing_info.append(f"Detected variant: {detection_result.variant.value}")
        processing_info.append(f"Confidence: {detection_result.confidence}")
        processing_info.append(f"Manifest found: {detection_result.manifest_found}")

        # Get the appropriate variant
        variant = factory.create_variant(detection_result.variant)

        # Count files for verbose output
        md_files = list(input_dir.rglob("*.md"))
        # Exclude manifest.md from count
        md_files = [f for f in md_files if f.name != "manifest.md"]
        processing_info.append(f"Found {len(md_files)} markdown files in directory")

        # Handle dry run mode differently
        if dry_run:
            # For dry run, temporarily disable dry_run to generate content
            options.dry_run = False
            variant_result = variant.implode(input_dir, options)

            if not variant_result.success:
                return ImplodeResult(
                    success=False,
                    errors=variant_result.errors,
                    processing_info=processing_info
                )

            # Read the generated content for preview
            if options.output_file.exists():
                preview_content = options.output_file.read_text(encoding='utf-8')
                # Remove the file since this is dry run
                options.output_file.unlink()
            else:
                preview_content = "No content generated"

            return ImplodeResult(
                success=True,
                output_file=options.output_file,
                preview=preview_content,
                processing_info=processing_info
            )

        # Normal mode - perform the implode operation
        variant_result = variant.implode(input_dir, options)

        if not variant_result.success:
            return ImplodeResult(
                success=False,
                errors=variant_result.errors,
                processing_info=processing_info
            )

        # Return successful result in legacy format
        return ImplodeResult(
            success=True,
            output_file=variant_result.output_file,
            processing_info=processing_info
        )

    except Exception as e:
        processing_info.append(f"Error during implode: {e}")
        return ImplodeResult(
            success=False,
            errors=[f"Error during implode: {e}"],
            processing_info=processing_info
        )


def _adjust_heading_levels(content: str, base_level: int) -> str:
    """Adjust heading levels in markdown content.

    Args:
        content: Markdown content
        base_level: Base level to add to existing headings

    Returns:
        Content with adjusted heading levels
    """
    import re

    def adjust_heading(match):
        current_level = len(match.group(1))
        new_level = min(current_level + base_level, 6)  # Max 6 heading levels
        return '#' * new_level + ' ' + match.group(2)

    return re.sub(r'^(#{1,6})\s+(.+)$', adjust_heading, content, flags=re.MULTILINE)


def combine_markdown_files(file_paths: list[Path], section_spacing: int = 2) -> str:
    """Combine multiple markdown files into a single content string.

    Args:
        file_paths: List of markdown file paths to combine
        section_spacing: Number of blank lines between sections

    Returns:
        Combined markdown content as a string
    """
    combined_parts = []

    for file_path in file_paths:
        if file_path.exists() and file_path.is_file():
            content = file_path.read_text().strip()
            if content:
                combined_parts.append(content)

    spacing = "\n" * (section_spacing + 1)  # +1 for the natural line break
    return spacing.join(combined_parts)


def preserve_markdown_formatting(file_paths: list[Path]) -> str:
    """Preserve markdown formatting while combining files.

    Args:
        file_paths: List of markdown file paths

    Returns:
        Combined content with all formatting preserved
    """
    # This function focuses on preserving formatting during combination
    # For now, it's equivalent to combine_markdown_files but could be extended
    # with specific formatting preservation logic
    return combine_markdown_files(file_paths, section_spacing=2)


def handle_index_files(directory: Path) -> str:
    """Handle index.md files as parent section content.

    Args:
        directory: Directory to scan for index files

    Returns:
        Combined content from all index files and other markdown files
    """
    all_content = []

    # Collect all markdown files including index files
    markdown_files = []

    # First, collect index files and regular files separately
    for path in directory.rglob("*.md"):
        if path.is_file():
            markdown_files.append(path)

    # Sort files hierarchically: depth-first traversal with index.md files first in each directory
    def hierarchical_sort_key(path: Path):
        # Calculate relative path from the root directory
        try:
            rel_path = path.relative_to(directory)
        except ValueError:
            rel_path = path

        # Build path components for hierarchical ordering
        path_parts = list(rel_path.parts)

        # Index files come first within their directory
        is_index = path.name == "index.md"

        # For depth-first traversal with index.md first:
        # 1. Sort by directory path components
        # 2. Within each directory, index.md comes first (priority 0), others come after (priority 1)
        # 3. For non-index files, sort alphabetically by filename

        if is_index:
            # Index files: replace filename with empty string and priority 0
            sort_parts = path_parts[:-1] + ['', 0]
        else:
            # Regular files: keep full path with priority 1
            sort_parts = path_parts[:-1] + [path_parts[-1], 1]

        return sort_parts

    markdown_files.sort(key=hierarchical_sort_key)

    # Combine all content
    for file_path in markdown_files:
        content = file_path.read_text().strip()
        if content:
            all_content.append(content)

    # Combine with proper spacing
    return "\n\n\n".join(all_content)


def process_front_matter(content_or_path) -> tuple[dict, str]:
    """Process YAML front matter from markdown content or file.

    Args:
        content_or_path: Markdown content string or Path to markdown file

    Returns:
        Tuple of (front_matter_dict, content_without_front_matter)
    """
    import re
    import yaml
    from pathlib import Path

    # Handle both string content and file paths
    if isinstance(content_or_path, (str, Path)):
        if isinstance(content_or_path, Path):
            if content_or_path.exists():
                content = content_or_path.read_text()
            else:
                return {}, ""
        else:
            content = content_or_path
    else:
        content = str(content_or_path)

    # Match YAML front matter
    fm_match = re.match(r'^---\n(.*?)\n---\n(.*)$', content, re.DOTALL)

    if fm_match:
        front_matter_yaml = fm_match.group(1)
        content_without_fm = fm_match.group(2).strip()

        try:
            front_matter = yaml.safe_load(front_matter_yaml)
            return front_matter or {}, content_without_fm
        except yaml.YAMLError:
            # If YAML parsing fails, return content as-is
            return {}, content
    else:
        return {}, content


def aggregate_content(directory: Path, output_file: Path = None,
                     preserve_structure: bool = True, preserve_front_matter: bool = False) -> str:
    """Aggregate content from a directory structure into a single markdown document.

    Args:
        directory: Source directory containing markdown files
        output_file: Optional output file path
        preserve_structure: Whether to preserve hierarchical structure
        preserve_front_matter: Whether to preserve and consolidate front matter

    Returns:
        Aggregated markdown content
    """
    # Collect all markdown files
    markdown_files = []
    for path in directory.rglob("*.md"):
        if path.is_file() and path.name.lower() not in ["readme.md"]:
            # Exclude output file if specified
            if output_file and path == output_file:
                continue
            markdown_files.append(path)

    # Sort files for consistent ordering
    markdown_files.sort()

    if preserve_front_matter:
        # Handle front matter consolidation
        consolidator = FrontMatterConsolidator(conflict_strategy="merge")
        consolidated_fm, combined_content = consolidator.consolidate(markdown_files)

        if consolidated_fm:
            import yaml
            # Add front matter to the beginning
            front_matter_yaml = yaml.dump(consolidated_fm, default_flow_style=False).strip()
            return f"---\n{front_matter_yaml}\n---\n\n{combined_content}"
        else:
            return combined_content
    elif preserve_structure:
        # Handle index files and hierarchy - use the comprehensive approach
        return handle_index_files(directory)
    else:
        return combine_markdown_files(markdown_files)


class ContentAggregator:
    """Aggregator for combining markdown content from multiple sources."""

    def __init__(self, section_spacing: int = 2, preserve_formatting: bool = True,
                 handle_front_matter: bool = True, include_toc: bool = False,
                 recursive: bool = True, sort_files: bool = True):
        self.section_spacing = section_spacing
        self.preserve_formatting = preserve_formatting
        self.handle_front_matter = handle_front_matter
        self.include_toc = include_toc
        self.recursive = recursive
        self.sort_files = sort_files
        self.aggregated_content = []

    def add_file(self, file_path: Path):
        """Add a file to the aggregation."""
        if file_path.exists() and file_path.is_file():
            content = file_path.read_text().strip()
            if content:
                self.aggregated_content.append(content)

    def add_content(self, content: str):
        """Add raw content to the aggregation."""
        if content.strip():
            self.aggregated_content.append(content.strip())

    def get_combined_content(self) -> str:
        """Get the combined content."""
        spacing = "\n" * (self.section_spacing + 1)
        return spacing.join(self.aggregated_content)

    def aggregate(self, directory: Path) -> str:
        """Aggregate content from a directory.

        Args:
            directory: Directory to aggregate content from

        Returns:
            Aggregated content string
        """
        # Use the existing aggregate_content function but with our settings
        return aggregate_content(
            directory,
            preserve_structure=True,
            preserve_front_matter=self.handle_front_matter
        )

    def reset(self):
        """Reset the aggregator."""
        self.aggregated_content.clear()


class FrontMatterConsolidator:
    """Consolidator for handling front matter from multiple files."""

    def __init__(self, conflict_strategy: str = "merge"):
        self.front_matters = []
        self.consolidated = {}
        self.conflict_strategy = conflict_strategy

    def add_front_matter(self, front_matter: dict):
        """Add front matter from a file."""
        if front_matter:
            self.front_matters.append(front_matter)

    def consolidate(self, files: list[Path] = None) -> tuple[dict, str]:
        """Consolidate front matter from files and return combined content.

        Args:
            files: List of file paths to process (optional if front matter already added)

        Returns:
            Tuple of (consolidated_front_matter, combined_content)
        """
        if files:
            # Process files and extract front matter
            all_content = []
            for file_path in files:
                front_matter, content = process_front_matter(file_path)
                if front_matter:
                    self.add_front_matter(front_matter)
                if content.strip():
                    all_content.append(content.strip())

            combined_content = "\n\n\n".join(all_content)
        else:
            combined_content = ""

        # Consolidate front matter
        consolidated = {}
        for fm in self.front_matters:
            for key, value in fm.items():
                if key in consolidated:
                    # Handle conflicts - for now, use list aggregation
                    if not isinstance(consolidated[key], list):
                        consolidated[key] = [consolidated[key]]
                    if isinstance(value, list):
                        consolidated[key].extend(value)
                    else:
                        consolidated[key].append(value)
                else:
                    consolidated[key] = value

        self.consolidated = consolidated
        return consolidated, combined_content

    def to_yaml(self) -> str:
        """Convert consolidated front matter to YAML string."""
        import yaml
        if self.consolidated:
            return yaml.dump(self.consolidated, default_flow_style=False)
        return ""


@register_plugin("markdown_commands")
class MarkdownCommandsPlugin(CommandPlugin):
    """Plugin providing core markdown file operations."""

    @property
    def metadata(self) -> PluginMetadata:
        return PluginMetadata(
            name="markdown_commands",
            version="1.0.0",
            description="Core markdown file operations with md- prefixes",
            author="MarkiTect Core Team",
            plugin_type=PluginType.COMMAND,
            markitect_version=">=0.1.0"
        )

    def get_commands(self) -> Dict[str, Any]:
        """Return the markdown commands with md- prefixes."""
        return {
            'md-ingest': md_ingest_command,
            'md-get': md_get_command,
            'md-list': md_list_command,
            'md-render': md_render_command,
            'themes': themes_list_command,
            'md-index': md_index_command,
            'md-explode': md_explode_command,
            'md-implode': md_implode_command,
            'md-package': md_package_command,
            'md-transclude': md_transclude_command
        }


# Define commands as standalone functions

@click.command()
@click.argument('file_path', type=click.Path(exists=True))
@click.pass_context
def md_ingest_command(ctx, file_path):
    """
    Process and store a markdown file.

    Ingests a markdown file into the MarkiTect system, parsing its content,
    extracting front matter, generating AST cache, and storing metadata
    in the database.

    FILE_PATH: Path to the markdown file to process

    Examples:
        markitect md-ingest README.md
        markitect md-ingest docs/guide.md
    """
    config = ctx.obj or {}
    try:
        if config.get('verbose', False):
            click.echo(f"Processing file: {file_path}")

        # Initialize document manager with database manager
        from markitect.clean_document_manager import CleanDocumentManager
        doc_manager = CleanDocumentManager(config.get('db_manager'))

        # Process the file
        result = doc_manager.ingest_file(Path(file_path))

        if config.get('verbose', False):
            click.echo(f"Processing results:")
            click.echo(f"  File: {result['metadata']['filename']}")
            click.echo(f"  AST nodes: {len(result['ast'])} nodes")
            click.echo(f"  Cache file: {result['ast_cache_path']}")
            click.echo(f"  Parse time: {result['parse_time']:.2f}s")
            click.echo(f"  Cache time: {result['cache_time']:.2f}s")

        click.echo(f"✓ Successfully ingested: {Path(file_path).name}")

    except Exception as e:
        click.echo(f"Error processing file: {e}", err=True)
        raise click.Abort()


@click.command()
@click.argument('file_path', type=str)
@click.option('--output', '-o', default='-',
              help='Output file (default: stdout)')
@click.pass_context
def md_get_command(ctx, file_path, output):
    """
    Retrieve content from a markdown file with metadata.

    Fetches a markdown file from the MarkiTect system, returning its content
    along with metadata, front matter, and optional AST information.

    FILE_PATH: Path to the markdown file to retrieve

    Examples:
        markitect md-get README.md
        markitect md-get docs/guide.md --output processed.md
    """
    config = ctx.obj or {}
    try:
        # Initialize document manager
        from markitect.clean_document_manager import CleanDocumentManager
        doc_manager = CleanDocumentManager(config.get('db_manager'))

        # Get file information
        result = doc_manager.get_file(file_path)

        # Output to file or stdout
        if output == '-':
            click.echo(result['content'])
        else:
            output_path = Path(output)
            output_path.write_text(result['content'], encoding='utf-8')
            click.echo(f"✓ Content written to: {output_path}")

        if config.get('verbose', False):
            metadata = result['metadata']
            click.echo(f"File: {metadata['filename']}", err=True)
            click.echo(f"Size: {metadata.get('size', 'unknown')} bytes", err=True)
            click.echo(f"Modified: {metadata.get('modified', 'unknown')}", err=True)

    except FileNotFoundError as e:
        click.echo(f"Error: File not found in database - {e}", err=True)
        raise click.Abort()
    except Exception as e:
        click.echo(f"Error retrieving file: {e}", err=True)
        raise click.Abort()


@click.command()
@click.option('--output-format', '-f', default='table',
              type=click.Choice(['table', 'json', 'yaml', 'simple']),
              help='Output format (default: table)')
@click.option('--names-only', is_flag=True,
              help='Show only filenames, no metadata')
@click.pass_context
def md_list_command(ctx, output_format, names_only):
    """
    List all markdown files in the MarkiTect system.

    Shows a list of all ingested markdown files with their metadata,
    including file sizes, modification dates, and processing status.

    Examples:
        markitect md-list
        markitect md-list --output-format json
        markitect md-list --names-only
    """
    config = ctx.obj or {}
    try:
        # Initialize document manager
        from markitect.clean_document_manager import CleanDocumentManager
        doc_manager = CleanDocumentManager(config.get('db_manager'))

        # Get file listing
        files = doc_manager.list_files()

        if not files:
            click.echo("No markdown files found in the system.")
            return

        if names_only:
            for file_info in files:
                click.echo(file_info['filename'])
        elif output_format == 'json':
            click.echo(json.dumps(files, indent=2))
        elif output_format == 'yaml':
            import yaml
            click.echo(yaml.dump(files, default_flow_style=False))
        else:  # table or simple
            click.echo(f"{'Filename':<40} {'Size':<10} {'Modified':<20}")
            click.echo("-" * 72)
            for file_info in files:
                size = file_info.get('size', 'unknown')
                modified = file_info.get('modified', 'unknown')
                click.echo(f"{file_info['filename']:<40} {size:<10} {modified:<20}")

    except Exception as e:
        click.echo(f"Error listing files: {e}", err=True)
        raise click.Abort()


@click.command()
@click.argument('input_file', type=click.Path(exists=True))
@click.option('--output', '-o', type=click.Path(),
              help='Output HTML file (default: <input>.html)')
@click.option('--theme', type=ThemeType(),
              help='Theme(s) to apply. Single: dark or layered: dark,academic or light,github,corporate. Available: basic, github, dark, academic, light, corporate, startup')
@click.option('--css', type=click.Path(),
              help='Custom CSS file to include')
@click.option('--edit', is_flag=True,
              help='Open in interactive edit mode with stable section editing')
@click.option('--insert', is_flag=True,
              help='Open in interactive insert mode with heading protection (levels 1-3 read-only)')
@click.option('--engine', type=str, default=None,
              help='Rendering engine to use (default: testdrive-jsui for edit/insert, standard for view)')
@click.option('--editor-theme', default='github',
              type=click.Choice(['github', 'monokai', 'tomorrow', 'dark']),
              help='Editor theme for live edit mode (default: github)')
@click.option('--keyboard-shortcuts', is_flag=True, default=True,
              help='Enable keyboard shortcuts in live edit mode')
@click.option('--use-publication-dir', is_flag=True,
              help='Use publication directory for output')
@click.option('--dont-use-publication-dir', is_flag=True,
              help='Don\'t use publication directory for output')
@click.option('--nodogtag', is_flag=True,
              help='Don\'t add HTML generation dogtag at end of document')
@click.option('--ship-assets', is_flag=True, default=None,
              help='Copy referenced assets to output directory')
@click.option('--no-ship-assets', is_flag=True,
              help='Don\'t copy referenced assets to output directory')
@click.option('--verbose', '-v', is_flag=True,
              help='Show detailed output including asset operations')
@click.option('--silent', '-s', is_flag=True,
              help='Suppress non-essential output')
@click.option('--image-max-width', type=str, default=None,
              help='Maximum width for images (default: 12cm, supports px, em, %, cm, in, etc.)')
@click.option('--image-max-height', type=str, default=None,
              help='Maximum height for images (default: 20cm, supports px, em, %, cm, in, etc.)')
@click.pass_context
def md_render_command(ctx, input_file, output, theme, css, edit, insert, engine, editor_theme,
                     keyboard_shortcuts, use_publication_dir, dont_use_publication_dir, nodogtag,
                     ship_assets, no_ship_assets, verbose, silent, image_max_width, image_max_height):
    """
    Render a markdown file to HTML with basic templates and live preview capabilities.

    Converts a markdown file to HTML using customizable layered themes and styles.
    Supports live editing mode with real-time preview and syntax highlighting.

    Theme Layering:
    - Single themes: basic, github, dark, academic, light, corporate, startup
    - Layered themes: dark,academic combines dark UI with academic typography
    - Later themes override settings from earlier themes

    INPUT_FILE: Path to the markdown file to render

    Examples:
        markitect md-render README.md
        markitect md-render docs/guide.md --output guide.html --theme github
        markitect md-render draft.md --edit --editor-theme monokai
        markitect md-render draft.md --insert --editor-theme monokai
        markitect md-render doc.md --theme dark --css custom.css
        markitect md-render doc.md --theme dark,academic
        markitect md-render doc.md --theme light,github,corporate
    """
    config = ctx.obj or {}

    try:
        input_path = Path(input_file)

        # Validate mode flags
        if edit and insert:
            raise click.BadParameter("Cannot use both --edit and --insert flags simultaneously. Choose one mode.")

        # Check environment variables for edit/insert modes (if not set via CLI flags)
        import os
        if not edit and not insert:
            if os.environ.get('MARKITECT_EDIT_MODE', '').lower() in ('true', '1', 'yes'):
                edit = True
            elif os.environ.get('MARKITECT_INSERT_MODE', '').lower() in ('true', '1', 'yes'):
                insert = True

        # Validate asset shipping flags
        if ship_assets and no_ship_assets:
            raise click.BadParameter("Cannot use both --ship-assets and --no-ship-assets flags simultaneously.")

        # Validate verbosity flags
        if verbose and silent:
            raise click.BadParameter("Cannot use both --verbose and --silent flags simultaneously.")

        # Handle image size configuration with environment variable support
        import os

        # Get image max width (CLI > ENV > default)
        final_image_max_width = image_max_width
        if final_image_max_width is None:
            final_image_max_width = os.environ.get('MARKITECT_IMAGE_MAX_WIDTH', '12cm')

        # Get image max height (CLI > ENV > default)
        final_image_max_height = image_max_height
        if final_image_max_height is None:
            final_image_max_height = os.environ.get('MARKITECT_IMAGE_MAX_HEIGHT', '20cm')

        # Determine output path with environment variable support
        if output:
            output_path = Path(output)
            # If output is a directory, use canonical filename within that directory
            if output_path.is_dir() or (not output_path.suffix and not output_path.exists()):
                # Ensure the directory exists
                output_path.mkdir(parents=True, exist_ok=True)
                # Use canonical filename (input name + .html) in the specified directory
                canonical_filename = input_path.with_suffix('.html').name
                output_path = output_path / canonical_filename
                output_is_directory = True
            else:
                output_is_directory = False
        else:
            # Check for environment variable
            import os
            env_output_dir = os.environ.get('MARKITECT_OUTPUT_DIR')
            if env_output_dir:
                output_path = Path(env_output_dir)
                output_path.mkdir(parents=True, exist_ok=True)
                canonical_filename = input_path.with_suffix('.html').name
                output_path = output_path / canonical_filename
                output_is_directory = True
            else:
                output_path = input_path.with_suffix('.html')
                output_is_directory = False

        # Use publication directory if specified
        if use_publication_dir and not dont_use_publication_dir:
            pub_dir = get_publication_directory()
            ensure_publication_directory(pub_dir)
            output_path = pub_dir / get_output_filename(input_path)
            output_is_directory = True  # Publication dir is always a directory output

        # Determine if we should ship assets
        should_ship_assets = False
        if no_ship_assets:
            should_ship_assets = False
        elif ship_assets:
            should_ship_assets = True
        elif output_is_directory:
            # Default: ship assets when output is a directory
            should_ship_assets = True


        # Ship markdown-referenced assets first if needed
        if should_ship_assets:
            if output_is_directory:
                # For directory output, ship to the same directory as the HTML file
                _ship_assets(input_path, output_path.parent, verbose, silent)

        # Determine rendering engine to use
        if engine is None:
            # Default engine selection
            if edit or insert:
                engine = 'testdrive-jsui'  # Default to testdrive-jsui for interactive modes
            else:
                engine = 'standard'  # Use standard CleanDocumentManager for non-interactive

        # Use plugin system for rendering engines, fallback to standard
        if engine != 'standard':
            try:
                from markitect.plugins import PluginManager, RenderingEngineManager, RenderingConfig
                plugin_manager = PluginManager()
                rendering_manager = RenderingEngineManager(plugin_manager)

                rendering_engine = rendering_manager.get_engine(engine)
                if rendering_engine is None:
                    if not silent:
                        click.echo(f"⚠️ Rendering engine '{engine}' not found, falling back to standard", err=True)
                    engine = 'standard'
                elif not silent:
                    modes = rendering_engine.get_supported_modes()
                    current_mode = 'edit' if edit else ('insert' if insert else 'view')
                    if not rendering_engine.validate_mode(current_mode):
                        click.echo(f"⚠️ Engine '{engine}' doesn't support mode '{current_mode}', falling back to standard", err=True)
                        engine = 'standard'
                    else:
                        click.echo(f"🎯 Using rendering engine: {engine} (supports: {', '.join(modes)})")

            except ImportError as e:
                if not silent:
                    click.echo(f"⚠️ Plugin system not available ({e}), using standard rendering", err=True)
                engine = 'standard'

        # Initialize document manager or rendering engine
        if engine == 'standard':
            from markitect.clean_document_manager import CleanDocumentManager
            doc_manager = CleanDocumentManager(config.get('db_manager'))

        # Render the file
        if edit:
            if engine != 'standard':
                # Plugin-based rendering for edit mode
                try:
                    # Read markdown content
                    content = input_path.read_text(encoding='utf-8')

                    # Configure rendering
                    render_config = RenderingConfig(
                        asset_base_url="_markitect",
                        development_mode=False,  # Production deployment for CLI usage
                        output_directory=output_path.parent
                    )

                    # Render using plugin
                    html_content = rendering_engine.render_document(content, 'edit', render_config)

                    # Deploy plugin assets
                    if not silent:
                        click.echo(f"📦 Deploying assets for engine '{engine}'...")

                    deployed_assets = rendering_manager.deploy_engine_assets(engine, render_config)

                    if verbose and deployed_assets:
                        total_assets = sum(len(files) for files in deployed_assets.values() if isinstance(files, list))
                        click.echo(f"   📄 Deployed {total_assets} asset files")
                        for asset_type, files in deployed_assets.items():
                            if isinstance(files, list) and files:
                                click.echo(f"   {asset_type}: {len(files)} files")

                    # Write output
                    output_path.write_text(html_content, encoding='utf-8')
                    result = True

                except Exception as e:
                    if not silent:
                        click.echo(f"❌ Plugin rendering failed: {e}", err=True)
                        click.echo("   Falling back to standard rendering...", err=True)
                    engine = 'standard'

            if engine == 'standard':
                # Standard edit mode - generate HTML with editing capabilities
                result = doc_manager.render_file(input_file, str(output_path),
                                               template=theme, css=css,
                                               edit_mode=True,
                                               editor_theme=editor_theme,
                                               keyboard_shortcuts=keyboard_shortcuts,
                                               nodogtag=nodogtag,
                                               image_max_width=final_image_max_width,
                                               image_max_height=final_image_max_height)

            if not silent:
                click.echo(f"✅ Rendered with INTERACTIVE editing mode to: {output_path}")
                click.echo(f"   Edit mode is fully functional with interactive section editing.")

            if verbose:
                click.echo(f"Editor theme: {editor_theme}")
                click.echo(f"Keyboard shortcuts: {'enabled' if keyboard_shortcuts else 'disabled'}")
                click.echo(f"Theme: {theme or 'default'}")
                click.echo(f"CSS: {css or 'default'}")
        elif insert:
            if engine != 'standard':
                # Plugin-based rendering for insert mode
                try:
                    # Read markdown content
                    content = input_path.read_text(encoding='utf-8')

                    # Configure rendering
                    render_config = RenderingConfig(
                        asset_base_url="_markitect",
                        development_mode=False,  # Production deployment for CLI usage
                        output_directory=output_path.parent
                    )

                    # Render using plugin (note: insert mode may not be supported by all plugins)
                    if rendering_engine.validate_mode('insert'):
                        html_content = rendering_engine.render_document(content, 'insert', render_config)
                    else:
                        # Fallback to edit mode if insert not supported
                        html_content = rendering_engine.render_document(content, 'edit', render_config)
                        if not silent:
                            click.echo(f"ℹ️ Engine '{engine}' doesn't support insert mode, using edit mode instead")

                    # Deploy plugin assets
                    if not silent:
                        click.echo(f"📦 Deploying assets for engine '{engine}'...")

                    deployed_assets = rendering_manager.deploy_engine_assets(engine, render_config)

                    if verbose and deployed_assets:
                        total_assets = sum(len(files) for files in deployed_assets.values() if isinstance(files, list))
                        click.echo(f"   📄 Deployed {total_assets} asset files")
                        for asset_type, files in deployed_assets.items():
                            if isinstance(files, list) and files:
                                click.echo(f"   {asset_type}: {len(files)} files")

                    # Write output
                    output_path.write_text(html_content, encoding='utf-8')
                    result = True

                except Exception as e:
                    if not silent:
                        click.echo(f"❌ Plugin rendering failed: {e}", err=True)
                        click.echo("   Falling back to standard rendering...", err=True)
                    engine = 'standard'

            if engine == 'standard':
                # Standard insert mode - generate HTML with insert capabilities and heading protection
                result = doc_manager.render_file(input_file, str(output_path),
                                               template=theme, css=css,
                                               insert_mode=True,
                                               editor_theme=editor_theme,
                                               keyboard_shortcuts=keyboard_shortcuts,
                                               nodogtag=nodogtag,
                                               image_max_width=final_image_max_width,
                                               image_max_height=final_image_max_height)

            if not silent:
                click.echo(f"✅ Rendered with INTERACTIVE insert mode to: {output_path}")
                click.echo(f"   Insert mode is fully functional with protected heading editing.")

            if verbose:
                click.echo(f"Editor theme: {editor_theme}")
                click.echo(f"Keyboard shortcuts: {'enabled' if keyboard_shortcuts else 'disabled'}")
                click.echo(f"Heading protection: levels 1-3 read-only")
                click.echo(f"Theme: {theme or 'default'}")
                click.echo(f"CSS: {css or 'default'}")
        else:
            # Static render
            result = doc_manager.render_file(input_file, str(output_path),
                                           template=theme, css=css,
                                           edit_mode=False,
                                           insert_mode=False,
                                           nodogtag=nodogtag,
                                           image_max_width=final_image_max_width,
                                           image_max_height=final_image_max_height)
            if not silent:
                click.echo(f"✓ Rendered to: {output_path}")

            if verbose:
                click.echo(f"Theme: {theme or 'default'}")
                click.echo(f"CSS: {css or 'default'}")

        # Ship HTML-referenced assets (JavaScript, CSS) after HTML generation
        if should_ship_assets and output_is_directory and output_path.exists():
            _ship_html_assets(output_path, output_path.parent, verbose, silent)

    except Exception as e:
        click.echo(f"Error rendering file: {e}", err=True)
        raise click.Abort()


@click.command()
@click.option('--format', type=click.Choice(['table', 'list', 'json']), default='table',
              help='Output format: table (default), list, or json')
@click.option('--scope', type=click.Choice(['mode', 'ui', 'document', 'branding', 'all']), default='all',
              help='Filter themes by scope: mode (light/dark), ui (editor interface), document (typography), branding (colors), or all (default)')
def themes_list_command(format, scope):
    """
    List all available themes and their properties.

    Shows the available themes that can be used with md-render and other commands.
    Themes can be used individually or combined in layers.

    Examples:
        markitect themes list
        markitect themes list --format json
        markitect themes list --scope ui
        markitect themes list --scope document --format list
    """
    from tabulate import tabulate
    import json

    # Get theme data
    layered_themes = []
    legacy_mappings = []

    # Process layered themes
    for theme_name, theme_data in LAYERED_THEMES.items():
        theme_scope = theme_data['scope']
        if scope == 'all' or scope == theme_scope:
            properties = theme_data['properties']
            # Get key properties for display based on scope
            key_props = []
            if theme_scope == 'mode':
                if 'body_background' in properties:
                    key_props.append(f"bg:{properties['body_background']}")
                if 'link_color' in properties:
                    key_props.append(f"links:{properties['link_color']}")
            elif theme_scope == 'ui':
                if 'editor_panel_bg' in properties:
                    key_props.append(f"panel:{properties['editor_panel_bg']}")
                if 'editor_text_color' in properties:
                    key_props.append(f"text:{properties['editor_text_color']}")
                if 'editor_focus_color' in properties:
                    key_props.append(f"focus:{properties['editor_focus_color']}")
            elif theme_scope == 'document':
                if 'font_family' in properties:
                    family = properties['font_family'].split(',')[0].strip().strip('"\'')
                    key_props.append(f"font:{family}")
                if 'link_color' in properties:
                    key_props.append(f"links:{properties['link_color']}")
            elif theme_scope == 'branding':
                if 'accent_color' in properties:
                    key_props.append(f"accent:{properties['accent_color']}")

            layered_themes.append({
                'name': theme_name,
                'scope': theme_scope,
                'properties': ', '.join(key_props) if key_props else 'default styling'
            })

    # Process legacy mappings
    for legacy_name, expanded_themes in LEGACY_THEME_MAPPING.items():
        legacy_mappings.append({
            'name': legacy_name,
            'expands_to': ' + '.join(expanded_themes)
        })

    if format == 'json':
        # JSON output
        output_data = {
            'layered_themes': layered_themes,
            'legacy_mappings': legacy_mappings,
            'usage': {
                'single': 'markitect md-render file.md --theme dark',
                'layered': 'markitect md-render file.md --theme dark,academic',
                'legacy': 'markitect md-render file.md --theme github'
            }
        }
        click.echo(json.dumps(output_data, indent=2))

    elif format == 'list':
        # Simple list output
        click.echo("Available themes:")
        for theme in layered_themes:
            click.echo(f"  {theme['name']} ({theme['scope']})")
        if legacy_mappings:
            click.echo("\nLegacy mappings:")
            for mapping in legacy_mappings:
                click.echo(f"  {mapping['name']} -> {mapping['expands_to']}")

    else:  # table format (default)
        # Table output
        if layered_themes:
            click.echo("Layered themes (can be combined):")
            headers = ['Theme', 'Scope', 'Key Properties']
            table_data = [[t['name'], t['scope'], t['properties']] for t in layered_themes]
            click.echo(tabulate(table_data, headers=headers, tablefmt='grid'))

        if legacy_mappings:
            click.echo("\nLegacy theme mappings:")
            headers = ['Legacy Name', 'Expands To']
            table_data = [[m['name'], m['expands_to']] for m in legacy_mappings]
            click.echo(tabulate(table_data, headers=headers, tablefmt='grid'))

        click.echo("\nUsage examples:")
        click.echo("  Single theme:    markitect md-render file.md --theme dark")
        click.echo("  Layered themes:  markitect md-render file.md --theme dark,academic")
        click.echo("  Legacy mapping:  markitect md-render file.md --theme github")


@click.command()
@click.argument('directory', type=click.Path(exists=True, file_okay=False, dir_okay=True))
@click.option('--output', '-o', type=click.Path(),
              help='Output index file (default: <directory>/index.html)')
@click.option('--theme', type=ThemeType(),
              help='Theme(s) to apply to index. Single: dark or layered: dark,github. Available: basic, github, dark, academic, light, corporate, startup')
@click.option('--recursive', '-r', is_flag=True,
              help='Include subdirectories recursively')
@click.pass_context
def md_index_command(ctx, directory, output, theme, recursive):
    """
    Generate an index page for HTML files in a directory.

    Creates an HTML index page listing all HTML files in the specified
    directory, with links and extracted titles.

    DIRECTORY: Path to the directory to index

    Examples:
        markitect md-index docs/
        markitect md-index . --recursive --output site-index.html
    """
    config = ctx.obj or {}

    try:
        dir_path = Path(directory)

        # Determine output path
        if output:
            output_path = Path(output)
        else:
            output_path = dir_path / 'index.html'

        # Find HTML files
        html_files = find_html_files(dir_path, recursive=recursive)

        if not html_files:
            click.echo(f"No HTML files found in: {dir_path}")

        # Create file info list, excluding the index file itself
        file_info_list = []
        for html_file in html_files:
            if html_file.name != output_path.name:
                title = extract_html_title(html_file)
                # Calculate relative path from output directory
                try:
                    relative_path = html_file.relative_to(dir_path)
                except ValueError:
                    # If html_file is not under dir_path, use absolute path
                    relative_path = html_file

                file_info_list.append({
                    'path': html_file,
                    'title': title,
                    'relative_path': str(relative_path)
                })

        # Generate index page title
        index_title = f"Index - {dir_path.name}"

        # Generate HTML content
        html_content = generate_index_html(file_info_list, index_title, theme)

        # Write index file
        output_path.parent.mkdir(parents=True, exist_ok=True)
        output_path.write_text(html_content, encoding='utf-8')

        click.echo(f"✓ Generated index: {output_path}")
        click.echo(f"📄 Indexed {len(file_info_list)} files")

        if config.get('verbose', False):
            click.echo("Files indexed:")
            for file_info in file_info_list:
                click.echo(f"  {file_info['title']} ({file_info['relative_path']})")

    except Exception as e:
        click.echo(f"Error generating index: {e}", err=True)
        raise click.Abort()


# ==============================================================================
# Enhanced Explode/Implode Commands with Variant System
# ==============================================================================

@click.command()
@click.argument('input_file', type=click.Path(exists=True))
@click.option('--output-dir', '-o', type=click.Path(),
              help='Output directory for exploded files (default: <filename>.mdd)')
@click.option('--variant', type=click.Choice(['flat', 'hierarchical', 'semantic']),
              default='flat', help='Directory organization variant (default: flat)')
@click.option('--max-depth', type=int, default=10,
              help='Maximum directory nesting depth (default: 10)')
@click.option('--create-manifest/--no-manifest', default=True,
              help='Create manifest.md for reversibility (default: true)')
@click.option('--dry-run', is_flag=True,
              help='Show what would be done without creating files')
@click.option('--verbose', '-v', is_flag=True,
              help='Show detailed output during processing')
@click.pass_context
def md_explode_command(ctx, input_file, output_dir, variant, max_depth, create_manifest, dry_run, verbose):
    """
    Explode a markdown file into a directory structure.

    Takes a markdown file with hierarchical headings (# ## ### etc.) and creates
    a directory structure where each heading becomes a directory or file, with
    content distributed appropriately. Supports multiple organization variants
    for different use cases.

    INPUT_FILE: Path to the markdown file to explode

    Variants:
        flat: Creates directories based on h1 headings (traditional)
        hierarchical: Numbered structure reflecting heading hierarchy
        semantic: Content-based grouping (parts, chapters, appendices)

    Examples:
        # Explode book.md into book.mdd/ directory (flat structure)
        markitect md-explode book.md

        # Use hierarchical structure with numbered directories
        markitect md-explode book.md --variant hierarchical

        # Explode into custom output directory
        markitect md-explode book.md --output-dir /path/to/chapters

        # Preview what would be created
        markitect md-explode book.md --dry-run --verbose --variant semantic

        # Explode without creating manifest (legacy mode)
        markitect md-explode book.md --no-manifest
    """
    config = ctx.obj or {}

    try:
        input_path = Path(input_file)

        # Import variant system
        from markitect.explode_variants import ExplodeVariant, ExplodeOptions, get_variant_factory

        # Convert string variant to enum
        try:
            variant_enum = ExplodeVariant(variant)
        except ValueError:
            click.echo(f"❌ Error: Unknown variant '{variant}'. Available: flat, hierarchical, semantic", err=True)
            raise click.Abort()

        # Determine output directory
        if output_dir:
            output_path = Path(output_dir)
        else:
            suffix = ".mdd" if create_manifest else "_exploded"
            output_path = input_path.parent / f"{input_path.stem}{suffix}"

        is_verbose = verbose or config.get('verbose', False)

        # Create explode options
        options = ExplodeOptions(
            variant=variant_enum,
            output_dir=output_path,
            max_depth=max_depth,
            create_manifest=create_manifest,
            dry_run=dry_run,
            verbose=is_verbose
        )

        if dry_run:
            click.echo(f"📋 Would explode using {variant.title()} Structure")
            click.echo(f"📁 Input file: {input_path}")
            click.echo(f"📁 Output directory: {output_path}")
            click.echo(f"📄 Create manifest: {create_manifest}")
            return

        # Use the variant system to explode the file
        factory = get_variant_factory()
        variant_instance = factory.create_variant(variant_enum)

        result = variant_instance.explode(input_path, options)

        if not result.success:
            click.echo(f"❌ Error exploding markdown file:", err=True)
            for error in result.errors:
                click.echo(f"   {error}", err=True)
            if result.warnings:
                click.echo("⚠️  Warnings:")
                for warning in result.warnings:
                    click.echo(f"   {warning}")
            raise click.Abort()

        click.echo(f"✅ Successfully exploded markdown file using {variant_instance.name}!")
        click.echo(f"📁 Created structure in: {result.output_directory}")

        if result.manifest_path:
            click.echo(f"📄 Created manifest: {result.manifest_path.name}")

        if is_verbose:
            click.echo(f"📄 Input file: {input_path}")
            click.echo(f"🔧 Variant used: {result.variant_used.value}")

            if result.files_created:
                click.echo(f"📄 Created {len(result.files_created)} files:")
                for file_path in sorted(result.files_created):
                    try:
                        relative_path = file_path.relative_to(result.output_directory)
                        click.echo(f"   {relative_path}")
                    except ValueError:
                        click.echo(f"   {file_path}")

    except Exception as e:
        click.echo(f"❌ Error exploding markdown file: {e}", err=True)
        raise click.Abort()


@click.command()
@click.argument('input_dir', type=click.Path(exists=True, file_okay=False, dir_okay=True))
@click.option('--output', '-o', type=click.Path(),
              help='Output markdown file (default: <dirname>_imploded.md)')
@click.option('--force-variant', type=click.Choice(['flat', 'hierarchical', 'semantic']),
              help='Force specific variant instead of auto-detection')
@click.option('--dry-run', is_flag=True,
              help='Preview what would be created without writing files')
@click.option('--verbose', '-v', is_flag=True,
              help='Show detailed processing information')
@click.option('--overwrite', is_flag=True,
              help='Overwrite existing output file')
@click.option('--section-spacing', type=int, default=2,
              help='Number of blank lines between sections (default: 2)')
@click.option('--preserve-front-matter/--no-front-matter', default=True,
              help='Preserve YAML front matter from files (default: preserve)')
@click.pass_context
def md_implode_command(ctx, input_dir, output, force_variant, dry_run, verbose, overwrite,
                      section_spacing, preserve_front_matter):
    """
    Implode a directory structure back into a single markdown file.

    Takes a directory structure (like one created by md-explode) and combines
    all markdown files back into a single document, reconstructing the original
    hierarchical heading structure. Automatically detects the variant used
    during explosion for optimal reconstruction.

    INPUT_DIR: Path to the directory to implode

    Auto-Detection:
        The command automatically detects the variant type by analyzing:
        - manifest.md file (highest priority)
        - Directory naming patterns
        - Content organization structure

    Examples:
        # Implode exploded directory back to markdown (auto-detect variant)
        markitect md-implode book.mdd/

        # Force specific variant instead of auto-detection
        markitect md-implode chapters/ --force-variant hierarchical

        # Specify custom output file
        markitect md-implode chapters/ --output reconstructed.md

        # Preview what would be created with detection info
        markitect md-implode content/ --dry-run --verbose
    """
    config = ctx.obj or {}

    try:
        input_path = Path(input_dir)

        # Determine output file
        if output:
            output_path = Path(output)
        else:
            output_path = input_path.parent / f"{input_path.name}_imploded.md"

        # Check if output file exists and overwrite not specified
        if output_path.exists() and not overwrite:
            click.echo(f"❌ Error: Output file {output_path} already exists. Use --overwrite to overwrite.", err=True)
            raise click.Abort()

        # Create implode options
        options = ImplodeOptions(
            output_file=output_path,
            preserve_front_matter=preserve_front_matter,
            section_spacing=section_spacing,
            overwrite=overwrite
        )

        if dry_run:
            # Collect files that would be processed
            markdown_files = []
            for path in input_path.rglob("*.md"):
                if path.is_file() and path.name.lower() != "readme.md":
                    markdown_files.append(path)
            markdown_files.sort()

            click.echo(f"📋 Would implode directory structure")
            click.echo(f"📁 Source directory: {input_path}")
            click.echo(f"📄 Would create file: {output_path}")
            click.echo(f"📄 Would process {len(markdown_files)} files")

            if verbose:
                click.echo(f"\nℹ️  Files to process:")
                for file_path in markdown_files:
                    try:
                        relative_path = file_path.relative_to(input_path)
                        click.echo(f"   {relative_path}")
                    except ValueError:
                        click.echo(f"   {file_path}")
        else:
            # Actually perform the implode operation
            result = cli_implode_directory(input_dir=input_path, options=options)

            if result.success:
                click.echo(f"✅ Successfully imploded directory")
                click.echo(f"📁 Source directory: {input_path}")
                click.echo(f"📄 Created file: {result.output_file}")

                if verbose:
                    # Count processed files for feedback
                    markdown_files = []
                    for path in input_path.rglob("*.md"):
                        if path.is_file() and path.name.lower() != "readme.md":
                            markdown_files.append(path)
                    click.echo(f"📄 Processed {len(markdown_files)} files")
            else:
                click.echo(f"❌ Failed to implode directory:", err=True)
                for error in result.errors:
                    click.echo(f"   {error}", err=True)
                raise click.Abort()

    except Exception as e:
        click.echo(f"❌ Error during implode: {e}", err=True)
        if ctx.obj and ctx.obj.get('debug'):
            import traceback
            traceback.print_exc()
        raise click.Abort()


# ==============================================================================
# Advanced Packaging Commands
# ==============================================================================

@click.command()
@click.argument('action', type=click.Choice(['create', 'extract', 'info']))
@click.argument('input_path', type=click.Path(exists=True))
@click.option('--output', '-o', type=click.Path(),
              help='Output path for package or extraction')
@click.option('--format', '-f', type=click.Choice(['mdz', 'mdt']), default='mdz',
              help='Package format (mdz for Markdown Zip, mdt for Markdown Transcluded)')
@click.option('--compression', '-c', type=click.IntRange(0, 9), default=6,
              help='Compression level for MDZ packages (0-9)')
@click.option('--include-assets', is_flag=True, default=True,
              help='Include assets when creating packages')
@click.option('--variables', type=click.Path(exists=True),
              help='JSON file with variables for MDT processing')
@click.option('--dry-run', is_flag=True,
              help='Show what would be done without making changes')
@click.option('--verbose', '-v', is_flag=True,
              help='Enable verbose output')
@click.pass_context
def md_package_command(ctx, action, input_path, output, format, compression,
                      include_assets, variables, dry_run, verbose):
    """
    Advanced package management for markdown documents.

    Actions:
    - create: Create MDZ/MDT package from source
    - extract: Extract package contents
    - info: Show package information

    Examples:

      markitect md-package create document.md --format mdz --output document.mdz
      markitect md-package extract document.mdz --output extracted/
      markitect md-package info document.mdz
    """
    try:
        input_path = Path(input_path)

        if action == 'create':
            # Import packaging modules
            from markitect.packaging.mdz_variant import MdzVariant
            from markitect.packaging.transclusion import TransclusionEngine

            if not output:
                if format == 'mdz':
                    output = input_path.with_suffix('.mdz')
                else:
                    output = input_path.with_suffix('.mdt')
            else:
                output = Path(output)

            if verbose:
                click.echo(f"📦 Creating {format.upper()} package")
                click.echo(f"📄 Source: {input_path}")
                click.echo(f"📦 Output: {output}")

            if dry_run:
                click.echo("🔍 Dry run - no files would be created")
                return

            if format == 'mdz':
                mdz = MdzVariant()
                result = mdz.create_package(
                    source_path=input_path,
                    options={
                        'output_path': output,
                        'compression_level': compression
                    }
                )

                click.echo(f"✅ MDZ package created successfully")
                click.echo(f"📦 Package: {result.get('package_path', output)}")
                click.echo(f"📊 Assets embedded: {result.get('assets_embedded', 0)}")
                click.echo(f"💾 Package size: {result.get('package_size', 0):,} bytes")

            else:  # mdt format
                if not input_path.is_file():
                    click.echo("❌ MDT format requires a single markdown file", err=True)
                    raise click.Abort()

                # For MDT, we just copy the file with transclusion processing
                content = input_path.read_text(encoding='utf-8')

                # Process with transclusion engine if variables provided
                if variables:
                    variables_path = Path(variables)
                    if variables_path.exists():
                        import json
                        var_data = json.loads(variables_path.read_text())

                        engine = TransclusionEngine(
                            base_path=input_path.parent,
                            variables=var_data
                        )
                        content = engine.process_content(content)

                output.write_text(content, encoding='utf-8')
                click.echo(f"✅ MDT template created successfully")
                click.echo(f"📄 Template: {output}")

        elif action == 'extract':
            from markitect.packaging.mdz_variant import MdzVariant

            if not output:
                output = input_path.parent / f"{input_path.stem}_extracted"
            else:
                output = Path(output)

            if verbose:
                click.echo(f"📂 Extracting package")
                click.echo(f"📦 Source: {input_path}")
                click.echo(f"📁 Output: {output}")

            if dry_run:
                click.echo("🔍 Dry run - no files would be extracted")
                return

            mdz = MdzVariant()
            result = mdz.extract_package(
                package_path=input_path,
                options={'output_dir': output}
            )

            click.echo(f"✅ Package extracted successfully")
            click.echo(f"📁 Output directory: {result['output_directory']}")
            click.echo(f"📄 Files extracted: {result['files_extracted']}")

        elif action == 'info':
            from markitect.packaging.mdz_variant import MdzVariant

            if verbose:
                click.echo(f"ℹ️  Package information for: {input_path}")

            mdz = MdzVariant()
            metadata = mdz.get_package_metadata(input_path)

            click.echo(f"📋 Package Format: {metadata.format}")
            click.echo(f"🏷️  Format Version: {metadata.version}")
            click.echo(f"⏰ Created: {metadata.created}")
            click.echo(f"🛠️  MarkiTect Version: {metadata.markitect_version}")
            click.echo(f"📊 Assets: {len(metadata.assets) if metadata.assets else 0}")

            if verbose and metadata.assets:
                click.echo("\n📁 Assets:")
                for asset in metadata.assets:
                    click.echo(f"   - {asset.path} ({asset.size:,} bytes)")

    except Exception as e:
        click.echo(f"❌ Error during package operation: {e}", err=True)
        if ctx.obj and ctx.obj.get('debug'):
            import traceback
            traceback.print_exc()
        raise click.Abort()


@click.command()
@click.argument('action', type=click.Choice(['process', 'validate']))
@click.argument('input_file', type=click.Path(exists=True))
@click.option('--output', '-o', type=click.Path(),
              help='Output file for processed content')
@click.option('--variables', type=click.Path(exists=True),
              help='JSON file containing template variables')
@click.option('--base-path', type=click.Path(exists=True),
              help='Base path for resolving includes (defaults to input file directory)')
@click.option('--max-depth', type=int, default=10,
              help='Maximum inclusion depth to prevent infinite recursion')
@click.option('--dry-run', is_flag=True,
              help='Show what would be processed without creating output')
@click.option('--verbose', '-v', is_flag=True,
              help='Enable verbose output with processing details')
@click.pass_context
def md_transclude_command(ctx, action, input_file, output, variables, base_path,
                         max_depth, dry_run, verbose):
    """
    Process markdown files with transclusion directives.

    Actions:
    - process: Process transclusion directives and generate output
    - validate: Check template for errors without processing

    Transclusion directives supported:
    - {{include "file.md"}} - Include another markdown file
    - {{variable_name}} - Substitute variables
    - {{if condition}} content {{endif}} - Conditional content

    Examples:

      markitect md-transclude process template.mdt --variables vars.json
      markitect md-transclude validate template.mdt
      markitect md-transclude process template.mdt --output result.md
    """
    try:
        from markitect.packaging.transclusion import TransclusionEngine
        from markitect.packaging.errors import TransclusionError, CircularReferenceError

        input_file = Path(input_file)

        # Load variables if provided
        var_data = {}
        if variables:
            variables_path = Path(variables)
            if verbose:
                click.echo(f"📋 Loading variables from: {variables_path}")
            import json
            var_data = json.loads(variables_path.read_text())

        # Set base path
        if base_path:
            base_path = Path(base_path)
        else:
            base_path = input_file.parent

        if verbose:
            click.echo(f"📄 Processing template: {input_file}")
            click.echo(f"📁 Base path: {base_path}")
            click.echo(f"📋 Variables: {len(var_data)} loaded")
            click.echo(f"🔢 Max depth: {max_depth}")

        # Create transclusion engine
        engine = TransclusionEngine(
            base_path=base_path,
            variables=var_data,
            max_depth=max_depth
        )

        if action == 'validate':
            # Validate template without full processing
            try:
                content = input_file.read_text(encoding='utf-8')

                # Parse directives to check syntax
                from markitect.packaging.transclusion.directives import DirectiveParser
                directives = DirectiveParser.parse_directives(content)

                click.echo(f"✅ Template validation successful")
                click.echo(f"📊 Found {len(directives)} transclusion directives")

                if verbose:
                    for directive in directives:
                        click.echo(f"   - {directive.type}: {directive.args}")

                # Check for potential circular references
                file_includes = DirectiveParser.extract_file_includes(content)
                if file_includes:
                    click.echo(f"📁 File includes: {len(file_includes)}")
                    if verbose:
                        for include in file_includes:
                            include_path = base_path / include
                            status = "✅" if include_path.exists() else "❌"
                            click.echo(f"   {status} {include}")

            except Exception as e:
                click.echo(f"❌ Template validation failed: {e}", err=True)
                raise click.Abort()

        elif action == 'process':
            if not output:
                output = input_file.with_suffix('.processed.md')
            else:
                output = Path(output)

            if verbose:
                click.echo(f"🔄 Processing transclusion directives")
                click.echo(f"📤 Output: {output}")

            if dry_run:
                click.echo("🔍 Dry run - no output file would be created")
                try:
                    result = engine.process_file(input_file)
                    click.echo(f"✅ Template processed successfully ({len(result)} characters)")
                except CircularReferenceError as e:
                    click.echo(f"❌ Circular reference detected: {e}", err=True)
                    raise click.Abort()
                except TransclusionError as e:
                    click.echo(f"❌ Transclusion error: {e}", err=True)
                    raise click.Abort()
                return

            # Process the template
            try:
                result = engine.process_file(input_file)

                # Write output
                output.write_text(result, encoding='utf-8')

                click.echo(f"✅ Transclusion processing completed")
                click.echo(f"📄 Input: {input_file}")
                click.echo(f"📄 Output: {output}")
                click.echo(f"📊 Output size: {len(result):,} characters")

                if verbose:
                    # Count lines for additional stats
                    lines = result.count('\n') + 1
                    click.echo(f"📊 Output lines: {lines:,}")

            except CircularReferenceError as e:
                click.echo(f"❌ Circular reference detected: {e}", err=True)
                click.echo("💡 Check your include directives for loops", err=True)
                raise click.Abort()
            except TransclusionError as e:
                click.echo(f"❌ Transclusion error: {e}", err=True)
                raise click.Abort()

    except Exception as e:
        click.echo(f"❌ Error during transclusion: {e}", err=True)
        if ctx.obj and ctx.obj.get('debug'):
            import traceback
            traceback.print_exc()
        raise click.Abort()


# ==============================================================================
# Utility Functions
# ==============================================================================

def normalize_filename(title):
    """
    Normalize a title string for use as a filename.

    Args:
        title: The title string to normalize

    Returns:
        A safe filename string
    """
    # Remove markdown formatting
    title = re.sub(r'[*_`~]', '', title)

    # Handle special characters
    title = unicodedata.normalize('NFKD', title)
    title = title.encode('ascii', 'ignore').decode('ascii')

    # Replace spaces and special chars with underscores
    title = re.sub(r'[^\w\s-]', '', title).strip()
    title = re.sub(r'[-\s]+', '_', title)

    # Convert to lowercase and limit length
    title = title.lower()[:50]

    return title or 'untitled'


def generate_safe_path(base_path, filename):
    """
    Generate a safe file path, avoiding conflicts.

    Args:
        base_path: Base directory path
        filename: Desired filename

    Returns:
        Path object for a safe, non-conflicting file
    """
    output_path = Path(base_path) / filename
    counter = 1

    while output_path.exists():
        name_part = output_path.stem
        ext_part = output_path.suffix
        output_path = output_path.parent / f"{name_part}_{counter}{ext_part}"
        counter += 1

    return output_path


# Directory Structure Analysis Functions

class DirectoryNode:
    """Represents a node in a directory structure analysis."""

    def __init__(self, path: Path, name: str, depth: int, is_directory: bool):
        self.path = path
        self.name = name
        self.depth = depth
        self.is_directory = is_directory
        self.children = []
        self.markdown_files = []
        self.parent = None

    def add_child(self, child: 'DirectoryNode'):
        """Add a child node to this directory node."""
        self.children.append(child)
        child.parent = self

    def add_markdown_file(self, file_path: Path):
        """Add a markdown file to this directory node."""
        self.markdown_files.append(file_path)

    def __repr__(self):
        return f"DirectoryNode(path={self.path}, name='{self.name}', depth={self.depth}, is_directory={self.is_directory})"


class DirectoryAnalysis:
    """Result of directory structure analysis."""

    def __init__(self):
        self.index_file = None
        self.content_files = []
        self.subdirectories = []

    def add_content_file(self, file_path: Path):
        """Add a content file to the analysis."""
        self.content_files.append(file_path)

    def add_subdirectory(self, dir_path: Path):
        """Add a subdirectory to the analysis."""
        self.subdirectories.append(dir_path)


class DirectoryStructure:
    """Complete directory structure analysis result."""

    def __init__(self):
        self.root_nodes = []
        self.all_nodes = []

    def add_root_node(self, node: DirectoryNode):
        """Add a root-level node."""
        self.root_nodes.append(node)
        self.all_nodes.append(node)

    def add_node(self, node: DirectoryNode):
        """Add any node to the complete list."""
        self.all_nodes.append(node)


def scan_markdown_files(directory: Path, recursive: bool = False) -> list[Path]:
    """Scan directory for markdown files.

    Args:
        directory: Directory to scan
        recursive: Whether to scan recursively

    Returns:
        List of markdown file paths
    """
    directory = Path(directory)
    markdown_files = []

    if recursive:
        # Use rglob for recursive search
        for file_path in directory.rglob("*.md"):
            if file_path.is_file():
                markdown_files.append(file_path)
    else:
        # Use glob for non-recursive search
        for file_path in directory.glob("*.md"):
            if file_path.is_file():
                markdown_files.append(file_path)

    # Sort for consistent ordering
    markdown_files.sort()
    return markdown_files


def detect_hierarchy_from_structure(directory: Path) -> list[DirectoryNode]:
    """Detect hierarchy levels based on directory depth.

    Args:
        directory: Root directory to analyze

    Returns:
        List of DirectoryNode objects representing the hierarchy
    """
    directory = Path(directory)
    nodes = []

    # Walk through all directories and files
    for root_path in directory.rglob("*"):
        if root_path.is_file() and root_path.suffix == ".md":
            # Calculate depth relative to base directory
            try:
                relative_path = root_path.relative_to(directory)
                depth = len(relative_path.parts) - 1  # File depth (subtract file itself)

                # Create node for the file
                node = DirectoryNode(
                    path=root_path,
                    name=root_path.name,
                    depth=depth,
                    is_directory=False
                )
                nodes.append(node)
            except ValueError:
                # Skip files outside the directory
                continue

    # Also add directory nodes
    for root_path in directory.rglob("*"):
        if root_path.is_dir():
            try:
                relative_path = root_path.relative_to(directory)
                depth = len(relative_path.parts)

                # Create node for the directory
                node = DirectoryNode(
                    path=root_path,
                    name=root_path.name,
                    depth=depth,
                    is_directory=True
                )
                nodes.append(node)
            except ValueError:
                continue

    # Sort by depth and name for consistent ordering
    nodes.sort(key=lambda n: (n.depth, n.name))
    return nodes


def identify_index_files(directory: Path) -> DirectoryAnalysis:
    """Identify index.md files vs regular content files.

    Args:
        directory: Directory to analyze

    Returns:
        DirectoryAnalysis object with index and content files categorized
    """
    directory = Path(directory)
    analysis = DirectoryAnalysis()

    # Scan for markdown files in the directory (non-recursive)
    for file_path in directory.glob("*.md"):
        if file_path.is_file():
            if file_path.name == "index.md":
                analysis.index_file = file_path
            else:
                analysis.add_content_file(file_path)

    # Also identify subdirectories
    for dir_path in directory.iterdir():
        if dir_path.is_dir():
            analysis.add_subdirectory(dir_path)

    return analysis


def analyze_directory_structure(directory: Path) -> DirectoryStructure:
    """Analyze complete directory structure for hierarchical organization.

    Args:
        directory: Root directory to analyze

    Returns:
        DirectoryStructure object with complete hierarchy analysis
    """
    directory = Path(directory)
    structure = DirectoryStructure()
    node_map = {}  # Path -> DirectoryNode mapping

    # First pass: create all nodes
    all_paths = [directory]  # Add the root directory itself

    # Add all subdirectories and files (rglob doesn't include the root)
    for path in directory.rglob("*"):
        all_paths.append(path)

    # Create nodes for all paths
    for path in all_paths:
        try:
            if path == directory:
                relative_path = Path(".")
                depth = 0
            else:
                relative_path = path.relative_to(directory)
                # Both files and directories: depth = number of path components
                depth = len(relative_path.parts)

            node = DirectoryNode(
                path=path,
                name=path.name if path != directory else directory.name,
                depth=depth,
                is_directory=path.is_dir()
            )

            node_map[path] = node
            structure.add_node(node)

            # Add to root nodes if at depth 1 (direct children of root)
            if depth == 1:
                structure.add_root_node(node)

        except ValueError:
            # Skip paths outside the directory
            continue

    # Special handling for flat directories (only files, no subdirectories)
    has_subdirectories = any(node.is_directory for node in structure.all_nodes if node.depth > 0)
    if not has_subdirectories:
        # This is a flat directory - adjust file depths to 0 and add them to root_nodes
        structure.root_nodes.clear()
        for node in structure.all_nodes:
            if node.depth == 1 and not node.is_directory:
                node.depth = 0
                structure.add_root_node(node)

    # Second pass: establish parent-child relationships
    for path, node in node_map.items():
        if path != directory:
            parent_path = path.parent
            if parent_path in node_map:
                parent_node = node_map[parent_path]
                parent_node.add_child(node)

        # Add markdown files to directory nodes
        if node.is_directory:
            for md_file in node.path.glob("*.md"):
                node.add_markdown_file(md_file)

    return structure


def implode_directory(input_dir: Path, output_file: Path) -> Path:
    """Implode a directory structure back into a markdown file.

    Simple wrapper around cli_implode_directory for use in tests and scripts.

    Args:
        input_dir: Directory containing markdown files to implode
        output_file: Output markdown file path

    Returns:
        Path to the created output file

    Raises:
        Exception: If the implode operation fails
    """
    from pathlib import Path

    input_dir = Path(input_dir)
    output_file = Path(output_file)

    # Use the existing cli_implode_directory function with round-trip compatibility options
    options = ImplodeOptions(
        input_dir=input_dir,
        output_file=output_file,
        overwrite=True,
        preserve_heading_levels=True,  # Preserve original heading levels for round-trip consistency
        include_readme_files=True      # Include README.md files created by explode process
    )
    result = cli_implode_directory(options=options)

    if not result.success:
        error_msg = result.error_message or "Implode operation failed"
        raise Exception(error_msg)

    return result.output_file


# =============================================================================
# Filename Decoding Functions for Issue #139
# =============================================================================
# These functions convert filesystem-safe names back to readable headings

def restore_special_characters(encoded_text: str) -> str:
    """Restore special characters that were encoded for filesystem safety."""
    # First convert underscores to spaces
    result = encoded_text.replace('_', ' ')

    # Handle specific patterns for special characters (before title casing)
    special_patterns = {
        'whats': "what's",
        'file path issues': 'file/path issues',
        'questions and answers': 'questions & answers',
        'cafe resume': 'café & résumé',
        'colon separated': 'colon: separated',
        'parentheses content': 'parentheses (content)',
        'brackets and more': 'brackets [and more]'
    }

    # Handle version patterns like v2 1 -> v2.1
    result = re.sub(r'\bv(\d+)\s+(\d+)', r'v\1.\2', result)

    for pattern, replacement in special_patterns.items():
        result = result.replace(pattern, replacement)

    # Apply title case to the result
    return apply_title_case(result)


def reconstruct_number_format(encoded_text: str) -> str:
    """Reconstruct proper number formats from encoded versions."""
    # Convert patterns like "section_1_1_1" to "Section 1.1.1"

    # Pattern for numbered sections with underscores (including letter sections like "appendix_a_1")
    pattern = r'(section|version|appendix|figure|table)_([a-zA-Z0-9]+)(_[a-zA-Z0-9]+)*'

    def replace_numbers(match):
        prefix = match.group(1).title()
        parts = match.group(0).split('_')[1:]  # Get all parts after the prefix

        # Convert underscores to dots in numeric parts, keep letters as uppercase
        formatted_parts = []
        for part in parts:
            if part.isdigit():
                formatted_parts.append(part)
            elif len(part) == 1 and part.isalpha():
                formatted_parts.append(part.upper())
            else:
                formatted_parts.append(part)

        number_str = '.'.join(formatted_parts)
        return f"{prefix} {number_str}"

    result = re.sub(pattern, replace_numbers, encoded_text, flags=re.IGNORECASE)
    return result


def apply_title_case(text: str) -> str:
    """Apply appropriate title case to reconstructed headings."""
    # Simple title case with some exceptions
    exceptions = {'and', 'or', 'the', 'a', 'an', 'with', 'of', 'in', 'on', 'at', 'to', 'for'}

    # Split on spaces and handle special characters within words
    words = text.split()
    result = []

    for i, word in enumerate(words):
        # Handle common acronyms first (overrides other rules)
        if word.lower() in ['api', 'sql', 'http', 'json', 'xml', 'css']:
            result.append(word.upper())
        # Handle words with brackets or parentheses - always capitalize content inside
        elif '[' in word or ']' in word or '(' in word or ')' in word:
            result.append(_capitalize_word(word))
        # Always capitalize first and last word
        elif i == 0 or i == len(words) - 1:
            result.append(_capitalize_word(word))
        # Don't capitalize exceptions unless they're the first word (but be more lenient with single letters)
        elif word.lower() in exceptions and len(word) > 1:
            result.append(word.lower())
        # Single letter words like "a" should generally be capitalized unless they're truly exceptions
        elif len(word) == 1 and word.lower() in ['a', 'i']:
            result.append(word.upper())
        else:
            result.append(_capitalize_word(word))

    return ' '.join(result)


def _capitalize_word(word: str) -> str:
    """Capitalize a word, handling special characters within the word."""
    if not word:
        return word

    # Handle words with special characters like "file/path"
    if '/' in word:
        parts = word.split('/')
        return '/'.join(part.capitalize() for part in parts)
    elif ':' in word:
        parts = word.split(':')
        return ':'.join(part.capitalize() for part in parts)
    elif '(' in word and ')' in word:
        # Handle parentheses - capitalize content inside
        before_paren = word[:word.index('(')]
        inside_parens = word[word.index('(')+1:word.index(')')]
        after_paren = word[word.index(')')+1:]
        return before_paren.capitalize() + '(' + inside_parens.capitalize() + ')' + after_paren.capitalize()
    elif '[' in word and ']' in word:
        # Handle brackets - capitalize content inside
        before_bracket = word[:word.index('[')]
        inside_brackets = word[word.index('[')+1:word.index(']')]
        after_bracket = word[word.index(']')+1:]
        return before_bracket.capitalize() + '[' + inside_brackets.capitalize() + ']' + after_bracket.capitalize()
    elif word.startswith('[') or word.endswith(']'):
        # Handle partial bracket words like "[and" or "more]"
        result = ""
        if word.startswith('['):
            result += '['
            word = word[1:]
        if word.endswith(']'):
            end_bracket = ']'
            word = word[:-1]
        else:
            end_bracket = ''
        result += word.capitalize() + end_bracket
        return result
    elif word.startswith('(') or word.endswith(')'):
        # Handle partial parenthesis words like "(content" or "content)"
        result = ""
        if word.startswith('('):
            result += '('
            word = word[1:]
        if word.endswith(')'):
            end_paren = ')'
            word = word[:-1]
        else:
            end_paren = ''
        result += word.capitalize() + end_paren
        return result
    else:
        return word.capitalize()


def decode_filename_to_heading(filename: str) -> str:
    """Decode filesystem-safe filename to readable heading."""
    if isinstance(filename, Path):
        filename = filename.name

    # Remove .md extension
    name = filename
    if name.endswith('.md'):
        name = name[:-3]

    # Handle special cases
    if name.lower() == 'index':
        return ""
    if name.lower() == 'readme':
        return "Readme"

    # Handle special API/version patterns like "api_v2_1_reference" (put early to avoid conflicts)
    api_version_pattern = r'(\w+)_v(\d+)_(\d+)_(.+)'
    api_version_match = re.match(api_version_pattern, name, re.IGNORECASE)
    if api_version_match:
        prefix, major, minor, title = api_version_match.groups()
        formatted_prefix = prefix.upper() if prefix.lower() in ['api', 'sql', 'http', 'json', 'xml', 'css'] else prefix.title()
        formatted_title = apply_title_case(restore_special_characters(title))
        return f"{formatted_prefix} v{major}.{minor}: {formatted_title}"

    # Handle numbered prefixes
    numbered_pattern = r'^(\d+)_(.+)$'
    numbered_match = re.match(numbered_pattern, name)
    if numbered_match:
        number, rest = numbered_match.groups()
        return f"{number}: {apply_title_case(restore_special_characters(rest))}"

    # Handle private sections (starting with _)
    if name.startswith('_'):
        name = name[1:]
        return apply_title_case(restore_special_characters(name))

    # Handle common patterns like "chapter_1_getting_started" or "section_a_getting_started"
    # First try pattern with multiple numeric parts like "1_2_3"
    multi_id_pattern = r'(chapter|section|part|appendix)_(\d+(?:_\d+)+)_(.+)'
    multi_id_match = re.match(multi_id_pattern, name, re.IGNORECASE)
    if multi_id_match:
        prefix, numbers, title = multi_id_match.groups()
        # Convert underscores in numbers to dots
        formatted_numbers = numbers.replace('_', '.')
        formatted_title = apply_title_case(restore_special_characters(title))
        return f"{prefix.title()} {formatted_numbers}: {formatted_title}"

    # Then try pattern with single letter/number identifier (but not if it looks like a multi-number pattern)
    single_id_pattern = r'(chapter|section|part|appendix)_([a-zA-Z]|\d+)_(.+)'
    single_id_match = re.match(single_id_pattern, name, re.IGNORECASE)
    if single_id_match:
        prefix, identifier, title = single_id_match.groups()
        # Capitalize single letters, keep numbers as-is
        if identifier.isalpha():
            formatted_id = identifier.upper()
        else:
            formatted_id = identifier
        formatted_title = apply_title_case(restore_special_characters(title))
        return f"{prefix.title()} {formatted_id}: {formatted_title}"

    # Handle simple prefix+title patterns like "appendix_troubleshooting"
    simple_prefix_pattern = r'(chapter|section|part|appendix)_(.+)'
    simple_prefix_match = re.match(simple_prefix_pattern, name, re.IGNORECASE)
    if simple_prefix_match:
        prefix, title = simple_prefix_match.groups()
        formatted_title = apply_title_case(restore_special_characters(title))
        return f"{prefix.title()}: {formatted_title}"

    # Handle simple numbered patterns like "section_2_3_4_advanced"
    simple_numbered = r'(\w+)_(\d+(?:_\d+)*)_(.+)'
    simple_match = re.match(simple_numbered, name, re.IGNORECASE)
    if simple_match:
        prefix, numbers, title = simple_match.groups()
        formatted_numbers = numbers.replace('_', '.')
        formatted_title = apply_title_case(restore_special_characters(title))
        return f"{prefix.title()} {formatted_numbers}: {formatted_title}"

    # Default case - just apply title case and restore special characters
    return apply_title_case(restore_special_characters(name))


def decode_directory_name_to_heading(dirname: str) -> str:
    """Decode directory name to heading."""
    # Use the same logic as filename decoding but without .md extension handling
    return decode_filename_to_heading(dirname)


class FilenameDecoder:
    """Comprehensive filename decoder for batch processing and configuration."""

    def __init__(self, preserve_acronyms=True, title_case_enabled=True,
                 number_format_reconstruction=True, context_aware=False,
                 flexible_parsing=False):
        """Initialize the decoder with configuration options."""
        self.preserve_acronyms = preserve_acronyms
        self.title_case_enabled = title_case_enabled
        self.number_format_reconstruction = number_format_reconstruction
        self.context_aware = context_aware
        self.flexible_parsing = flexible_parsing

    def decode(self, filename_or_path, parent_context=None):
        """Decode a single filename or path."""
        if isinstance(filename_or_path, Path):
            filename = filename_or_path.name
        else:
            filename = str(filename_or_path)

        return decode_filename_to_heading(filename)

    def decode_batch(self, filenames):
        """Process multiple filenames in batch."""
        return [self.decode(filename) for filename in filenames]


def _ship_assets(input_path: Path, output_dir: Path, verbose: bool = False, silent: bool = False):
    """
    Ship (copy) assets referenced in markdown file to output directory.

    Args:
        input_path: Path to the markdown file
        output_dir: Directory where assets should be copied
        verbose: Whether to print detailed output
        silent: Whether to suppress non-essential output
    """
    import shutil
    import hashlib
    from markitect.assets.discovery import discover_assets_from_markdown

    def get_file_hash(file_path):
        """Get SHA-256 hash of file content for content comparison."""
        hash_sha256 = hashlib.sha256()
        with open(file_path, "rb") as f:
            for chunk in iter(lambda: f.read(4096), b""):
                hash_sha256.update(chunk)
        return hash_sha256.hexdigest()

    try:
        # Read the markdown content
        markdown_content = input_path.read_text(encoding='utf-8')

        # Discover assets
        base_path = input_path.parent
        assets = discover_assets_from_markdown(markdown_content, base_path)

        shipped_count = 0
        skipped_count = 0
        missing_count = 0

        for asset_ref in assets:
            # Skip URLs and broken assets
            if asset_ref.asset_path.startswith(('http:', 'https:', 'mailto:', 'data:')):
                continue

            if asset_ref.is_broken or not asset_ref.resolved_path:
                missing_count += 1
                if verbose:
                    click.echo(f"  ⚠ Missing asset: {asset_ref.asset_path}", err=True)
                continue

            # Determine output path (preserve relative directory structure)
            clean_path = asset_ref.asset_path.lstrip('./')
            dest_path = output_dir / clean_path

            # Create destination directory
            dest_path.parent.mkdir(parents=True, exist_ok=True)

            # Check if we need to copy (smart comparison for cross-filesystem compatibility)
            should_copy = True
            if dest_path.exists():
                source_stat = asset_ref.resolved_path.stat()
                dest_stat = dest_path.stat()

                # Detect if we're in a cross-filesystem scenario where timestamps might be unreliable
                # Heuristics: different filesystems, or timestamps that don't make sense
                is_cross_fs = (
                    # Different device IDs suggests different filesystems
                    source_stat.st_dev != dest_stat.st_dev or
                    # Destination path starts with /mnt/ (common WSL Windows mount)
                    str(dest_path).startswith('/mnt/') or
                    # Very large timestamp differences (>1 hour) for same content suggest sync issues
                    abs(source_stat.st_mtime - dest_stat.st_mtime) > 3600
                )

                if is_cross_fs:
                    # Use content-based comparison for cross-filesystem scenarios
                    if source_stat.st_size == dest_stat.st_size:
                        try:
                            source_hash = get_file_hash(asset_ref.resolved_path)
                            dest_hash = get_file_hash(dest_path)

                            if source_hash == dest_hash:
                                should_copy = False
                                skipped_count += 1
                                if verbose:
                                    click.echo(f"  → Content verified (cross-fs): {asset_ref.asset_path}")
                            # If hashes differ, should_copy remains True
                        except (OSError, IOError):
                            if verbose:
                                click.echo(f"  ⚠ Could not verify content, will copy: {asset_ref.asset_path}")
                            pass
                    # If sizes differ, should_copy remains True
                else:
                    # Use fast timestamp comparison for same-filesystem scenarios
                    if source_stat.st_mtime <= dest_stat.st_mtime and source_stat.st_size == dest_stat.st_size:
                        should_copy = False
                        skipped_count += 1
                        if verbose:
                            click.echo(f"  → Timestamp verified: {asset_ref.asset_path}")
                    # If timestamp suggests newer source or different size, should_copy remains True

            if should_copy:
                shutil.copy2(asset_ref.resolved_path, dest_path)
                shipped_count += 1
                if verbose:
                    click.echo(f"  ✓ Copied: {asset_ref.asset_path}")
            elif verbose:
                click.echo(f"  → Skipped (up-to-date): {asset_ref.asset_path}")

        # Summary - provide feedback based on verbosity settings
        total_assets = shipped_count + skipped_count + missing_count

        if total_assets > 0 and not silent:
            if shipped_count > 0:
                click.echo(f"✓ Shipped {shipped_count} assets")
            elif skipped_count > 0:
                click.echo(f"✓ All {skipped_count} assets up-to-date")

            # Additional details for verbose or when there are mixed results
            if verbose or (shipped_count > 0 and skipped_count > 0):
                if skipped_count > 0 and shipped_count > 0:
                    click.echo(f"  → {skipped_count} already up-to-date")

            # Always show missing assets as it's important information
            if missing_count > 0:
                click.echo(f"  ⚠ {missing_count} assets not found", err=True)

    except Exception as e:
        if verbose:
            click.echo(f"Error shipping assets: {e}", err=True)


def _ship_html_assets(html_path: Path, output_dir: Path, verbose: bool = False, silent: bool = False):
    """
    Ship (copy) assets referenced in HTML file to output directory.

    This function scans the generated HTML file for JavaScript and CSS references,
    then copies those assets to the output directory for deployment.

    Args:
        html_path: Path to the generated HTML file
        output_dir: Directory where assets should be copied
        verbose: Whether to print detailed output
        silent: Whether to suppress non-essential output
    """
    import shutil
    import hashlib
    from markitect.assets.discovery import discover_assets_from_html

    def get_file_hash(file_path):
        """Get SHA-256 hash of file content for content comparison."""
        hash_sha256 = hashlib.sha256()
        with open(file_path, "rb") as f:
            for chunk in iter(lambda: f.read(4096), b""):
                hash_sha256.update(chunk)
        return hash_sha256.hexdigest()

    try:
        # Read the HTML content
        html_content = html_path.read_text(encoding='utf-8')

        # Discover HTML assets (JavaScript, CSS)
        # Use the project root as base path for resolving markitect/static/js paths
        project_root = Path(__file__).parent.parent.parent.parent  # Go up to project root (markitect/plugins/builtin/markdown_commands.py -> project_root)
        assets = discover_assets_from_html(html_content, project_root)

        if not assets:
            if verbose:
                click.echo("  No HTML assets (JS/CSS) found to ship")
            return

        shipped_count = 0
        skipped_count = 0
        missing_count = 0

        if not silent:
            click.echo(f"📦 Shipping {len(assets)} HTML assets...")

        for asset_ref in assets:
            # Skip URLs and broken assets
            if asset_ref.asset_path.startswith(('http:', 'https:', 'mailto:', 'data:')):
                continue

            if asset_ref.is_broken or not asset_ref.resolved_path:
                missing_count += 1
                if verbose:
                    click.echo(f"  ⚠ Missing HTML asset: {asset_ref.asset_path}", err=True)
                continue

            # Determine output path (preserve relative directory structure)
            clean_path = asset_ref.asset_path.lstrip('./')
            dest_path = output_dir / clean_path

            # Create destination directory
            dest_path.parent.mkdir(parents=True, exist_ok=True)

            # Check if we need to copy (smart comparison for cross-filesystem compatibility)
            should_copy = True
            if dest_path.exists():
                source_stat = asset_ref.resolved_path.stat()
                dest_stat = dest_path.stat()

                # Detect if we're in a cross-filesystem scenario where timestamps might be unreliable
                # Heuristics: different filesystems, or timestamps that don't make sense
                is_cross_fs = (
                    # Different device IDs suggests different filesystems
                    source_stat.st_dev != dest_stat.st_dev or
                    # Destination path starts with /mnt/ (common WSL Windows mount)
                    str(dest_path).startswith('/mnt/') or
                    # Very large timestamp differences (>1 hour) for same content suggest sync issues
                    abs(source_stat.st_mtime - dest_stat.st_mtime) > 3600
                )

                if is_cross_fs:
                    # Use content-based comparison for cross-filesystem scenarios
                    if source_stat.st_size == dest_stat.st_size:
                        try:
                            source_hash = get_file_hash(asset_ref.resolved_path)
                            dest_hash = get_file_hash(dest_path)

                            if source_hash == dest_hash:
                                should_copy = False
                                skipped_count += 1
                                if verbose:
                                    click.echo(f"  → Content verified (cross-fs): {asset_ref.asset_path}")
                            # If hashes differ, should_copy remains True
                        except (OSError, IOError):
                            if verbose:
                                click.echo(f"  ⚠ Could not verify content, will copy: {asset_ref.asset_path}")
                            pass
                    # If sizes differ, should_copy remains True
                else:
                    # Use fast timestamp comparison for same-filesystem scenarios
                    if source_stat.st_mtime <= dest_stat.st_mtime and source_stat.st_size == dest_stat.st_size:
                        should_copy = False
                        skipped_count += 1
                        if verbose:
                            click.echo(f"  → Timestamp verified: {asset_ref.asset_path}")
                    # If timestamp suggests newer source or different size, should_copy remains True

            if should_copy:
                shutil.copy2(asset_ref.resolved_path, dest_path)
                shipped_count += 1
                if verbose:
                    click.echo(f"  ✓ Shipped HTML asset: {asset_ref.asset_path}")

        # Report results
        if not silent:
            click.echo(f"✓ Shipped {shipped_count} HTML assets, skipped {skipped_count} up-to-date")
            if missing_count > 0:
                click.echo(f"  ⚠ {missing_count} HTML assets not found", err=True)

    except Exception as e:
        if verbose:
            click.echo(f"Error shipping HTML assets: {e}", err=True)