markitect-main/markitect/explode_variants/hierarchical_variant.py

"""
Hierarchical variant implementation for explode-implode operations.

This variant creates numbered directory structures with semantic hierarchy,
making it easier to understand document organization at a glance.
"""

import re
from pathlib import Path
from typing import Dict, List, Any, Optional, Tuple

from .base_variant import (
    BaseVariant, ExplodeOptions, ImplodeOptions,
    ExplodeResult, ImplodeResult
)
from .enums import ExplodeVariant
from .manifest_manager import ManifestManager, StructureEntry
from ..matter_frontmatter.parser import FrontmatterParser


class HierarchicalVariant(BaseVariant):
    """
    Hierarchical variant implementation.

    Creates numbered directory structures with nested organization.
    This provides clear document hierarchy and natural ordering.

    Structure example:
    book.mdd/
    ├── manifest.md
    ├── 01_introduction/
    │   ├── index.md
    │   ├── 01_overview.md
    │   └── 02_scope.md
    ├── 02_main_content/
    │   ├── index.md
    │   ├── 01_chapter_one.md
    │   └── 02_chapter_two.md
    └── 03_conclusion/
        └── index.md
    """

    def __init__(self):
        """Initialize the hierarchical variant."""
        super().__init__(ExplodeVariant.HIERARCHICAL)
        self.manifest_manager = ManifestManager()
        self.frontmatter_parser = FrontmatterParser()

    @property
    def name(self) -> str:
        """Human-readable name of the variant."""
        return "Hierarchical Structure"

    @property
    def description(self) -> str:
        """Description of the variant's behavior."""
        return ("Creates numbered directory structures with semantic hierarchy. "
                "Provides clear document organization and natural ordering.")

    def explode(
        self,
        input_file: Path,
        options: ExplodeOptions
    ) -> ExplodeResult:
        """
        Explode a markdown file using the hierarchical structure variant.

        Args:
            input_file: Path to the markdown file to explode
            options: Options controlling the explode operation

        Returns:
            Result of the explode operation
        """
        # Validate input
        validation_errors = self.validate_input_file(input_file)
        if validation_errors:
            return ExplodeResult(
                success=False,
                output_directory=options.output_dir or Path(),
                files_created=[],
                manifest_path=None,
                warnings=[],
                errors=validation_errors,
                variant_used=self.variant_type
            )

        # Determine output directory
        if options.output_dir:
            output_dir = options.output_dir
        else:
            suffix = ".mdd" if options.create_manifest else "_exploded"
            output_dir = input_file.parent / f"{input_file.stem}{suffix}"

        # Create output directory
        creation_errors = self.create_output_directory(output_dir, overwrite=True)
        if creation_errors:
            return ExplodeResult(
                success=False,
                output_directory=output_dir,
                files_created=[],
                manifest_path=None,
                warnings=[],
                errors=creation_errors,
                variant_used=self.variant_type
            )

        try:
            # Parse the markdown content
            content = input_file.read_text(encoding='utf-8')

            # Extract and save front matter if present and preservation is enabled
            files_created = []
            if options.preserve_front_matter:
                frontmatter, content_without_fm = self.frontmatter_parser.separate_frontmatter_and_content(content)
                if frontmatter:
                    # Save front matter to _frontmatter.yml
                    import yaml
                    fm_file = output_dir / "_frontmatter.yml"
                    fm_content = yaml.dump(frontmatter, default_flow_style=False)
                    fm_file.write_text(fm_content, encoding='utf-8')
                    files_created.append(fm_file)
                    # Use content without front matter for processing
                    content = content_without_fm

            # Analyze document structure
            sections = self._parse_hierarchical_structure(content)

            # Create hierarchical directory structure
            hierarchy_files = self._create_hierarchical_structure(
                output_dir, sections, options
            )

            # Create manifest if requested
            manifest_path = None
            if options.create_manifest:
                structure = self._build_structure_entries(sections)
                manifest_path = self.manifest_manager.create_manifest(
                    output_dir=output_dir,
                    original_file=input_file,
                    variant=self.variant_type,
                    structure=structure,
                    preservation_options={
                        "front_matter": options.preserve_front_matter,
                        "section_order": True,
                        "heading_levels": True,
                        "numbering_scheme": "hierarchical"
                    }
                )
                hierarchy_files.append(manifest_path)

            # Combine all created files
            all_files = files_created + hierarchy_files

            return ExplodeResult(
                success=True,
                output_directory=output_dir,
                files_created=all_files,
                manifest_path=manifest_path,
                warnings=[],
                errors=[],
                variant_used=self.variant_type
            )

        except Exception as e:
            return ExplodeResult(
                success=False,
                output_directory=output_dir,
                files_created=[],
                manifest_path=None,
                warnings=[],
                errors=[f"Error during hierarchical explosion: {e}"],
                variant_used=self.variant_type
            )

    def implode(
        self,
        input_directory: Path,
        options: ImplodeOptions
    ) -> ImplodeResult:
        """
        Implode a hierarchical directory structure back into a markdown file.

        Args:
            input_directory: Path to the directory to implode
            options: Options controlling the implode operation

        Returns:
            Result of the implode operation
        """
        # Validate input
        validation_errors = self.validate_input_directory(input_directory)
        if validation_errors:
            return ImplodeResult(
                success=False,
                output_file=options.output_file or Path(),
                files_processed=[],
                variant_detected=self.variant_type,
                warnings=[],
                errors=validation_errors
            )

        # Determine output file
        if options.output_file:
            output_file = options.output_file
        else:
            output_file = input_directory.parent / f"{input_directory.name}_imploded.md"

        try:
            # Read manifest if available
            manifest_data = self.manifest_manager.read_manifest(input_directory)

            # Reconstruct content from hierarchical structure
            content, files_processed = self._reconstruct_from_hierarchy(
                input_directory, manifest_data, options
            )

            # Add front matter if present and preservation is enabled
            if options.preserve_front_matter:
                fm_file = input_directory / '_frontmatter.yml'
                if fm_file.exists():
                    try:
                        import yaml
                        frontmatter_content = fm_file.read_text(encoding='utf-8').strip()
                        content = f"---\n{frontmatter_content}\n---\n\n{content}"
                    except Exception:
                        pass  # Ignore errors reading front matter

            # Write output file
            if not options.dry_run:
                output_file.write_text(content, encoding='utf-8')

            return ImplodeResult(
                success=True,
                output_file=output_file,
                files_processed=files_processed,
                variant_detected=self.variant_type,
                warnings=[],
                errors=[]
            )

        except Exception as e:
            return ImplodeResult(
                success=False,
                output_file=output_file,
                files_processed=[],
                variant_detected=self.variant_type,
                warnings=[],
                errors=[f"Error during hierarchical implosion: {e}"]
            )

    def can_handle_directory(self, directory: Path) -> bool:
        """
        Check if this variant can handle the given directory structure.

        Args:
            directory: Path to the directory to check

        Returns:
            True if this variant can handle the directory
        """
        if not directory.exists() or not directory.is_dir():
            return False

        # Check for manifest indicating hierarchical variant
        manifest_data = self.manifest_manager.read_manifest(directory)
        if manifest_data and manifest_data.explosion_type == "hierarchical":
            return True

        # Check for hierarchical structure patterns
        subdirs = [d for d in directory.iterdir() if d.is_dir()]

        # Look for numbered prefixes (strong hierarchical indicator)
        numbered_dirs = sum(1 for d in subdirs if re.match(r'^\d+_', d.name))

        # High ratio of numbered directories indicates hierarchical structure
        return (numbered_dirs / len(subdirs) if subdirs else 0) > 0.6

    def get_detection_patterns(self) -> Dict[str, Any]:
        """
        Get patterns used for auto-detecting this variant.

        Returns:
            Dictionary of detection patterns and weights
        """
        return {
            "manifest_type": "hierarchical",
            "numbered_directory_ratio": {"min": 0.6, "weight": 0.8},
            "index_file_count": {"min": 2, "weight": 0.5},
            "max_depth": {"min": 2, "weight": 0.4},
            "nested_numbered_dirs": {"weight": 0.7}
        }

    def _parse_hierarchical_structure(self, content: str) -> List[Dict[str, Any]]:
        """
        Parse markdown content into hierarchical sections.

        Args:
            content: Markdown content to parse

        Returns:
            List of section dictionaries with hierarchy information
        """
        sections = []
        lines = content.split('\n')
        current_section = None
        current_content = []
        section_counter = 1

        for i, line in enumerate(lines):
            # Check for headings
            heading_match = re.match(r'^(#{1,6})\s+(.+)', line)

            if heading_match:
                # Save previous section
                if current_section:
                    current_section['content'] = '\n'.join(current_content)
                    current_section['end_line'] = i
                    sections.append(current_section)

                # Start new section
                level = len(heading_match.group(1))
                title = heading_match.group(2).strip()

                current_section = {
                    'level': level,
                    'title': title,
                    'start_line': i + 1,
                    'order': section_counter,
                    'parent': self._find_parent_section(sections, level),
                    'numbering': self._generate_numbering(sections, level, section_counter)
                }
                current_content = [line]
                section_counter += 1
            else:
                if current_content:
                    current_content.append(line)

        # Handle last section
        if current_section:
            current_section['content'] = '\n'.join(current_content)
            current_section['end_line'] = len(lines)
            sections.append(current_section)

        return sections

    def _find_parent_section(self, sections: List[Dict[str, Any]], level: int) -> Optional[str]:
        """
        Find the parent section for the current heading level.

        Args:
            sections: Previously parsed sections
            level: Current heading level

        Returns:
            Parent section title or None
        """
        # Look for the most recent section with a lower level
        for section in reversed(sections):
            if section['level'] < level:
                return section['title']
        return None

    def _generate_numbering(self, sections: List[Dict[str, Any]], level: int, order: int) -> str:
        """
        Generate hierarchical numbering for a section.

        Args:
            sections: Previously parsed sections
            level: Current heading level
            order: Overall section order

        Returns:
            Hierarchical numbering string (e.g., "01", "02_01", etc.)
        """
        if level == 1:
            # Count h1 sections
            h1_count = sum(1 for s in sections if s['level'] == 1) + 1
            return f"{h1_count:02d}"

        # Find parent numbering and append subsection number
        parent_title = self._find_parent_section(sections, level)
        if parent_title:
            parent_section = next((s for s in sections if s['title'] == parent_title), None)
            if parent_section:
                # Count subsections at this level under the same parent
                subsection_count = sum(
                    1 for s in sections
                    if s['level'] == level and s.get('parent') == parent_title
                ) + 1
                return f"{parent_section['numbering']}_{subsection_count:02d}"

        # Fallback numbering
        return f"{order:02d}"

    def _create_hierarchical_structure(
        self,
        output_dir: Path,
        sections: List[Dict[str, Any]],
        options: ExplodeOptions
    ) -> List[Path]:
        """
        Create the hierarchical directory structure from parsed sections.

        Args:
            output_dir: Output directory for the structure
            sections: Parsed sections with hierarchy information
            options: Explode options

        Returns:
            List of created file paths
        """
        files_created = []

        for section in sections:
            # Generate directory name
            safe_title = self._sanitize_filename(section['title'])
            dir_name = f"{section['numbering']}_{safe_title}"

            # Create section directory
            section_dir = output_dir / dir_name
            section_dir.mkdir(exist_ok=True)

            # Create index.md for this section
            index_path = section_dir / "index.md"

            # Process content - extract subsections if any
            main_content, subsections = self._extract_subsections(
                section['content'], section['level']
            )

            # Write main content to index.md
            index_path.write_text(main_content, encoding='utf-8')
            files_created.append(index_path)

            # Create files for subsections
            for i, subsection in enumerate(subsections, 1):
                subsection_title = subsection.get('title', f'subsection_{i}')
                safe_sub_title = self._sanitize_filename(subsection_title)
                sub_file_name = f"{i:02d}_{safe_sub_title}.md"

                sub_file_path = section_dir / sub_file_name
                sub_file_path.write_text(subsection['content'], encoding='utf-8')
                files_created.append(sub_file_path)

        return files_created

    def _extract_subsections(self, content: str, parent_level: int) -> Tuple[str, List[Dict[str, Any]]]:
        """
        Extract subsections from section content.

        Args:
            content: Section content
            parent_level: Level of the parent section

        Returns:
            Tuple of (main_content, subsections_list)
        """
        lines = content.split('\n')
        main_content_lines = []
        subsections = []
        current_subsection = None
        current_subsection_lines = []

        for line in lines:
            heading_match = re.match(r'^(#{1,6})\s+(.+)', line)

            if heading_match:
                level = len(heading_match.group(1))
                title = heading_match.group(2).strip()

                if level > parent_level:
                    # This is a subsection
                    if current_subsection:
                        # Save previous subsection
                        current_subsection['content'] = '\n'.join(current_subsection_lines)
                        subsections.append(current_subsection)

                    # Start new subsection
                    current_subsection = {
                        'level': level,
                        'title': title
                    }
                    current_subsection_lines = [line]
                elif level <= parent_level:
                    # This is the main section heading or a peer section
                    if level == parent_level:
                        main_content_lines.append(line)
                    else:
                        # Higher-level heading that shouldn't be here in normal parsing
                        main_content_lines.append(line)
            else:
                # Regular content line
                if current_subsection:
                    current_subsection_lines.append(line)
                else:
                    main_content_lines.append(line)

        # Handle last subsection
        if current_subsection:
            current_subsection['content'] = '\n'.join(current_subsection_lines)
            subsections.append(current_subsection)

        main_content = '\n'.join(main_content_lines)
        return main_content, subsections

    def _sanitize_filename(self, title: str) -> str:
        """
        Sanitize a title for use as a filename/directory name.

        Args:
            title: Original title

        Returns:
            Sanitized filename
        """
        # Remove special characters
        safe_title = re.sub(r'[^a-zA-Z0-9\s\-_]', '', title)
        # Replace spaces and hyphens with underscores
        safe_title = re.sub(r'[\s\-]+', '_', safe_title)
        # Convert to lowercase
        safe_title = safe_title.lower()
        # Remove leading/trailing underscores
        safe_title = safe_title.strip('_')
        # Limit length
        if len(safe_title) > 50:
            safe_title = safe_title[:50].rstrip('_')

        return safe_title or 'untitled'

    def _build_structure_entries(self, sections: List[Dict[str, Any]]) -> List[StructureEntry]:
        """
        Build structure entries for manifest from parsed sections.

        Args:
            sections: Parsed sections

        Returns:
            List of structure entries
        """
        entries = []

        for section in sections:
            safe_title = self._sanitize_filename(section['title'])
            dir_name = f"{section['numbering']}_{safe_title}"
            path = f"{dir_name}/index.md"

            entry = StructureEntry(
                type=f"h{section['level']}",
                title=section['title'],
                path=path,
                order=section['order'],
                parent=section.get('parent'),
                level=section['level'],
                original_line=section.get('start_line')
            )
            entries.append(entry)

        return entries

    def _reconstruct_from_hierarchy(
        self,
        input_directory: Path,
        manifest_data: Any,
        options: ImplodeOptions
    ) -> Tuple[str, List[Path]]:
        """
        Reconstruct markdown content from hierarchical directory structure.

        Args:
            input_directory: Directory containing hierarchical structure
            manifest_data: Manifest data if available
            options: Implode options

        Returns:
            Tuple of (reconstructed_content, files_processed)
        """
        content_parts = []
        files_processed = []

        # Get all directories and sort them properly
        if manifest_data and hasattr(manifest_data, 'structure'):
            # Use manifest data to determine proper order
            subdirs = []
            dir_mapping = {}

            # Create mapping of directory names to Path objects
            all_dirs = [d for d in input_directory.iterdir()
                       if d.is_dir() and not d.name.startswith('.')]
            for d in all_dirs:
                dir_mapping[d.name] = d

            # Sort manifest entries by original order
            for entry in sorted(manifest_data.structure, key=lambda x: x.order):
                dir_name = Path(entry.path).parts[0] if entry.path else ""
                if dir_name in dir_mapping and dir_mapping[dir_name] not in subdirs:
                    subdirs.append(dir_mapping[dir_name])

            # Add any remaining directories not in manifest (fallback)
            for d in all_dirs:
                if d not in subdirs:
                    subdirs.append(d)
        else:
            # Fallback: sort by numbering prefix, then by name
            subdirs = sorted([
                d for d in input_directory.iterdir()
                if d.is_dir() and not d.name.startswith('.')
            ], key=lambda d: (
                int(d.name.split('_')[0]) if re.match(r'^\d+_', d.name) else 999,
                d.name
            ))

        for subdir in subdirs:
            self._process_directory_recursively(subdir, content_parts, files_processed)

        # Join with appropriate spacing
        spacing = '\n' * (options.section_spacing + 1)
        full_content = spacing.join(content_parts)

        return full_content, files_processed

    def _process_directory_recursively(self, directory: Path, content_parts: List[str], files_processed: List[Path]):
        """
        Recursively process a directory and its subdirectories for hierarchical content.

        Args:
            directory: Directory to process
            content_parts: List to append content to
            files_processed: List to append processed files to
        """
        # Read index.md if it exists
        index_file = directory / "index.md"
        if index_file.exists():
            index_content = index_file.read_text(encoding='utf-8')
            content_parts.append(index_content)
            files_processed.append(index_file)

        # Read other markdown files in this directory
        md_files = sorted([
            f for f in directory.glob("*.md")
            if f.name != "index.md"
        ], key=lambda f: f.name)

        for md_file in md_files:
            file_content = md_file.read_text(encoding='utf-8')
            content_parts.append(file_content)
            files_processed.append(md_file)

        # Recursively process subdirectories
        subdirs = sorted([
            d for d in directory.iterdir()
            if d.is_dir() and not d.name.startswith('.')
        ], key=lambda d: (
            int(d.name.split('_')[0]) if re.match(r'^\d+_', d.name) else 999,
            d.name
        ))

        for subdir in subdirs:
            self._process_directory_recursively(subdir, content_parts, files_processed)