markitect-main/markitect/explode_variants/flat_variant.py

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

This variant represents the current default behavior where h1 headings
become top-level directories with content organized beneath them.
"""

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

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 FlatVariant(BaseVariant):
    """
    Flat variant implementation.

    Creates directories based on h1 headings with nested content.
    This is the current default behavior for backward compatibility.

    Structure example:
    book.mdd/
    ├── manifest.md
    ├── book_title/
    │   ├── index.md
    │   ├── chapter_1.md
    │   └── chapter_2.md
    └── conclusion.md
    """

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

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

    @property
    def description(self) -> str:
        """Description of the variant's behavior."""
        return ("Creates directories based on h1 headings with content organized beneath them. "
                "This is the default structure for backward compatibility.")

    def explode(
        self,
        input_file: Path,
        options: ExplodeOptions
    ) -> ExplodeResult:
        """
        Explode a markdown file using the flat 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')

            # Implement flat explode logic directly
            files_created = self._explode_flat_structure(
                input_file, output_dir, content, options
            )

            # Create manifest if requested
            manifest_path = None
            if options.create_manifest:
                structure = self._analyze_structure(content, output_dir)
                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
                    }
                )
                files_created.append(manifest_path)

            return ExplodeResult(
                success=True,
                output_directory=output_dir,
                files_created=files_created,
                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 explosion: {e}"],
                variant_used=self.variant_type
            )

    def implode(
        self,
        input_directory: Path,
        options: ImplodeOptions
    ) -> ImplodeResult:
        """
        Implode a 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)

            # Implement flat implode logic directly
            content, files_processed = self._implode_flat_structure(
                input_directory, manifest_data, options
            )

            # 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 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 flat variant
        manifest_data = self.manifest_manager.read_manifest(directory)
        if manifest_data and manifest_data.explosion_type == "flat":
            return True

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

        # Look for typical flat patterns (no numbered prefixes, no semantic grouping)
        numbered_dirs = sum(1 for d in subdirs if re.match(r'^\d+_', d.name))
        semantic_dirs = sum(1 for d in subdirs
                          if any(name in d.name.lower()
                               for name in ['parts', 'chapters', 'sections', 'appendices']))

        # Flat structure has minimal numbered or semantic directories
        return (numbered_dirs / len(subdirs) if subdirs else 0) < 0.3 and \
               (semantic_dirs / len(subdirs) if subdirs else 0) < 0.3

    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": "flat",
            "numbered_directory_ratio": {"max": 0.3, "weight": 0.6},
            "semantic_directory_ratio": {"max": 0.3, "weight": 0.5},
            "index_file_count": {"min": 0, "weight": 0.3},
            "fallback_score": 0.6  # Default choice
        }

    def _explode_flat_structure(
        self,
        input_file: Path,
        output_dir: Path,
        content: str,
        options: ExplodeOptions
    ) -> List[Path]:
        """
        Implement flat structure explosion directly.

        Creates directories based on h1 headings with nested content.
        This is the traditional behavior for backward compatibility.
        """
        files_created = []

        # Extract and save front matter if present and preservation is enabled
        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

        # Parse sections based on headings
        sections = self._parse_flat_sections(content)

        for section in sections:
            if section['level'] == 1:
                # Create directory for h1 sections
                safe_title = self._sanitize_filename(section['title'])
                section_dir = output_dir / safe_title
                section_dir.mkdir(exist_ok=True)

                # Create index.md for the main content
                index_file = section_dir / "index.md"

                # Extract main content and subsections
                main_content, subsections = self._extract_content_and_subsections(
                    section['content'], section['level']
                )

                index_file.write_text(main_content, encoding='utf-8')
                files_created.append(index_file)

                # Create files for subsections
                for subsection in subsections:
                    sub_title = self._sanitize_filename(subsection['title'])
                    sub_file = section_dir / f"{sub_title}.md"
                    sub_file.write_text(subsection['content'], encoding='utf-8')
                    files_created.append(sub_file)

            else:
                # Handle standalone sections (not under h1)
                safe_title = self._sanitize_filename(section['title'])
                standalone_file = output_dir / f"{safe_title}.md"
                standalone_file.write_text(section['content'], encoding='utf-8')
                files_created.append(standalone_file)

        return files_created

    def _implode_flat_structure(
        self,
        input_directory: Path,
        manifest_data: Any,
        options: ImplodeOptions
    ) -> tuple[str, List[Path]]:
        """
        Implement flat structure implosion directly.

        Reconstructs markdown content from flat directory structure.
        """
        content_parts = []
        files_processed = []

        # If we have manifest data, use it for proper ordering
        if manifest_data and hasattr(manifest_data, 'structure'):
            # Use manifest to determine file order
            output_file = options.output_file
            for entry in sorted(manifest_data.structure, key=lambda x: x.order):
                file_path = input_directory / entry.path
                if (file_path.exists() and
                    file_path.name != "manifest.md" and
                    (output_file is None or file_path.resolve() != output_file.resolve())):
                    file_content = file_path.read_text(encoding='utf-8')
                    content_parts.append(file_content)
                    files_processed.append(file_path)
        else:
            # Fallback: collect all markdown files recursively (legacy behavior)
            # This ensures compatibility with tests that expect all nested files to be processed
            all_md_files = []

            # Collect all markdown files recursively, excluding output file if it exists
            output_file = options.output_file
            for md_file in input_directory.rglob("*.md"):
                if (md_file.name != "manifest.md" and
                    (output_file is None or md_file.resolve() != output_file.resolve())):
                    all_md_files.append(md_file)

            # Sort files by their path to ensure consistent ordering
            all_md_files.sort(key=lambda f: str(f.relative_to(input_directory)))

            # Process all found markdown files
            for md_file in all_md_files:
                content = md_file.read_text(encoding='utf-8')
                content_parts.append(content)
                files_processed.append(md_file)

        # Check for legacy front matter file (from old explode system)
        legacy_front_matter = None
        fm_file = input_directory / '_frontmatter.yml'
        if fm_file.exists() and options.preserve_front_matter:
            try:
                legacy_front_matter = fm_file.read_text(encoding='utf-8').strip()
            except Exception:
                pass  # Ignore errors reading front matter

        # Normalize content parts - remove excessive leading/trailing whitespace but preserve content
        normalized_parts = []
        for part in content_parts:
            if part:
                # Remove excessive leading/trailing newlines but preserve internal structure
                normalized = part.strip('\r\n')
                if normalized:
                    normalized_parts.append(normalized)

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

        # Add front matter to the beginning if found
        if legacy_front_matter and options.preserve_front_matter:
            full_content = f"---\n{legacy_front_matter}\n---\n\n{full_content}"

        return full_content, files_processed

    def _parse_flat_sections(self, content: str) -> List[Dict[str, Any]]:
        """Parse content into sections for flat structure."""
        sections = []
        lines = content.split('\n')
        current_section = None
        current_content = []
        section_order = 1

        for i, line in enumerate(lines):
            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)
                    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,
                    'order': section_order,
                    'start_line': i + 1
                }
                current_content = [line]
                section_order += 1
            else:
                if current_content:
                    current_content.append(line)

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

        return sections

    def _extract_content_and_subsections(self, content: str, parent_level: int) -> tuple[str, List[Dict[str, Any]]]:
        """Extract main content and subsections from a section."""
        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]
                else:
                    # This is the main section heading or higher level
                    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."""
        # Remove markdown heading markers
        title = re.sub(r'^#+\s*', '', title)
        # 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 _basic_explode_implementation(
        self,
        input_file: Path,
        output_dir: Path,
        content: str
    ) -> List[Path]:
        """Basic explode implementation for testing purposes."""
        files_created = []

        # Simple h1-based splitting
        sections = re.split(r'\n# ', content)

        for i, section in enumerate(sections):
            if not section.strip():
                continue

            if i == 0:
                # First section might not have leading #
                if not section.startswith('#'):
                    section = '# ' + section
            else:
                # Add back the # that was removed by split
                section = '# ' + section

            # Extract title
            lines = section.split('\n')
            title_line = lines[0]
            title = re.sub(r'^#\s*', '', title_line).strip()

            # Create directory and file
            safe_title = re.sub(r'[^\w\s-]', '', title).strip()
            safe_title = re.sub(r'[-\s]+', '_', safe_title).lower()

            section_dir = output_dir / safe_title
            section_dir.mkdir(exist_ok=True)

            file_path = section_dir / "index.md"
            file_path.write_text(section, encoding='utf-8')
            files_created.append(file_path)

        return files_created

    def _basic_implode_implementation(self, input_directory: Path) -> tuple[str, List[Path]]:
        """Basic implode implementation for testing purposes."""
        content_parts = []
        files_processed = []

        # Find all markdown files
        md_files = sorted(input_directory.glob("**/*.md"))

        for file_path in md_files:
            if file_path.name == "manifest.md":
                continue

            file_content = file_path.read_text(encoding='utf-8')
            content_parts.append(file_content)
            files_processed.append(file_path)

        # Join with appropriate spacing
        full_content = '\n\n\n\n'.join(content_parts)

        return full_content, files_processed

    def _analyze_structure(self, content: str, output_dir: Path) -> List[StructureEntry]:
        """Analyze the content structure for manifest generation."""
        structure = []
        lines = content.split('\n')

        order = 1
        for i, line in enumerate(lines):
            # Check for headings
            heading_match = re.match(r'^(#{1,6})\s+(.+)', line)
            if heading_match:
                level = len(heading_match.group(1))
                title = heading_match.group(2).strip()

                # Generate path based on title using same sanitization as file creation
                safe_title = self._sanitize_filename(title)

                if level == 1:
                    path = f"{safe_title}/index.md"
                else:
                    path = f"{safe_title}.md"

                structure.append(StructureEntry(
                    type=f"h{level}",
                    title=title,
                    path=path,
                    order=order,
                    level=level,
                    original_line=i + 1
                ))
                order += 1

        return structure