AI Book Agent MCP Server

"""EPUB parser for extracting and structuring book content.""" import ebooklib from ebooklib import epub from bs4 import BeautifulSoup import re import json from pathlib import Path from typing import List, Dict, Any, Optional import logging from dataclasses import dataclass logger = logging.getLogger(__name__) @dataclass class BookChapter: """Represents a chapter from a book.""" chapter_id: int title: str content: str word_count: int book_id: str book_title: str author: str @dataclass class BookMetadata: """Book metadata.""" id: str title: str author: str publication_date: str language: str description: str class EPUBParser: """Parser for EPUB files.""" def __init__(self): self.logger = logging.getLogger(__name__) def parse_epub(self, epub_path: str) -> tuple[BookMetadata, List[BookChapter]]: """ Parse an EPUB file and extract metadata and chapters. Args: epub_path: Path to the EPUB file Returns: Tuple of (metadata, chapters) """ epub_path = Path(epub_path) if not epub_path.exists(): raise FileNotFoundError(f"EPUB file not found: {epub_path}") self.logger.info(f"Parsing EPUB: {epub_path}") try: book = epub.read_epub(str(epub_path)) metadata = self._extract_metadata(book, epub_path) chapters = self._extract_chapters(book, metadata) self.logger.info(f"Successfully parsed {len(chapters)} chapters from {metadata.title}") return metadata, chapters except Exception as e: self.logger.error(f"Error parsing EPUB {epub_path}: {e}") raise def _extract_metadata(self, book: epub.EpubBook, epub_path: Path) -> BookMetadata: """Extract metadata from EPUB book.""" # Generate book ID from filename book_id = self._generate_book_id(epub_path.stem) # Extract metadata with fallbacks title = self._get_metadata_value(book, 'DC', 'title') or epub_path.stem author = self._get_metadata_value(book, 'DC', 'creator') or "Unknown Author" publication_date = self._get_metadata_value(book, 'DC', 'date') or "Unknown" language = self._get_metadata_value(book, 'DC', 'language') or "en" description = self._get_metadata_value(book, 'DC', 'description') or "" return BookMetadata( id=book_id, title=title, author=author, publication_date=publication_date, language=language, description=description ) def _extract_chapters(self, book: epub.EpubBook, metadata: BookMetadata) -> List[BookChapter]: """Extract chapters from EPUB book.""" chapters = [] chapter_id = 0 # Get all document items for item in book.get_items(): if item.get_type() == ebooklib.ITEM_DOCUMENT: content = self._extract_text_from_html(item.get_content()) if content and len(content.strip()) > 100: # Skip very short content # Try to extract title from the HTML title = self._extract_title_from_html(item.get_content()) or f"Chapter {chapter_id + 1}" chapter = BookChapter( chapter_id=chapter_id, title=title, content=content, word_count=len(content.split()), book_id=metadata.id, book_title=metadata.title, author=metadata.author ) chapters.append(chapter) chapter_id += 1 return chapters def _get_metadata_value(self, book: epub.EpubBook, namespace: str, name: str) -> Optional[str]: """Safely extract metadata value.""" try: metadata = book.get_metadata(namespace, name) if metadata: return metadata[0][0] # Get first value except Exception: pass return None def _extract_text_from_html(self, html_content: bytes) -> str: """Extract clean text from HTML content.""" try: soup = BeautifulSoup(html_content, 'html.parser') # Remove script and style elements for script in soup(["script", "style"]): script.decompose() # Extract text text = soup.get_text() # Clean up the text text = re.sub(r'\s+', ' ', text) # Replace multiple whitespace with single space text = re.sub(r'\n\s*\n', '\n\n', text) # Clean up line breaks text = text.strip() return text except Exception as e: self.logger.warning(f"Error extracting text from HTML: {e}") return "" def _extract_title_from_html(self, html_content: bytes) -> Optional[str]: """Extract title from HTML content.""" try: soup = BeautifulSoup(html_content, 'html.parser') # Look for title in various header tags for tag in ['h1', 'h2', 'h3', 'title']: title_elem = soup.find(tag) if title_elem: title = title_elem.get_text().strip() if title and len(title) < 200: # Reasonable title length return title # Look for class names that might indicate titles for class_name in ['title', 'chapter-title', 'heading']: title_elem = soup.find(class_=class_name) if title_elem: title = title_elem.get_text().strip() if title and len(title) < 200: return title except Exception: pass return None def _generate_book_id(self, filename: str) -> str: """Generate a clean book ID from filename.""" # Remove common file extensions and clean up book_id = filename.lower() book_id = re.sub(r'\.(epub|pdf)$', '', book_id) book_id = re.sub(r'[^\w\-_]', '-', book_id) book_id = re.sub(r'-+', '-', book_id) book_id = book_id.strip('-') return book_id or "unknown-book" def save_processed_book(self, metadata: BookMetadata, chapters: List[BookChapter], output_dir: str) -> str: """ Save processed book data to JSON file. Args: metadata: Book metadata chapters: List of chapters output_dir: Directory to save the processed data Returns: Path to the saved file """ output_dir = Path(output_dir) output_dir.mkdir(parents=True, exist_ok=True) output_file = output_dir / f"{metadata.id}.json" data = { "metadata": { "id": metadata.id, "title": metadata.title, "author": metadata.author, "publication_date": metadata.publication_date, "language": metadata.language, "description": metadata.description, "total_chapters": len(chapters), "total_words": sum(ch.word_count for ch in chapters) }, "chapters": [ { "chapter_id": ch.chapter_id, "title": ch.title, "content": ch.content, "word_count": ch.word_count } for ch in chapters ] } with open(output_file, 'w', encoding='utf-8') as f: json.dump(data, f, indent=2, ensure_ascii=False) self.logger.info(f"Saved processed book to: {output_file}") return str(output_file)