aigenrec/luminabackend

1import os2from typing import Optional3import pymupdf4llm4from docx import Document5from bs4 import BeautifulSoup6import markdown7from utils.logger import logger8 9class FileParser:10    @staticmethod11    def extract_text(file_path: str) -> Optional[str]:12        """Extract text from various file formats"""13        try:14            _, ext = os.path.splitext(file_path)15            ext = ext.lower()16            17            if ext == '.pdf':18                return FileParser._extract_pdf(file_path)19            elif ext in ['.docx', '.doc']:20                return FileParser._extract_docx(file_path)21            elif ext == '.txt':22                return FileParser._extract_txt(file_path)23            elif ext == '.html':24                return FileParser._extract_html(file_path)25            elif ext == '.md':26                return FileParser._extract_markdown(file_path)27            else:28                logger.warning(f"Unsupported file type: {ext}")29                return None30                31        except Exception as e:32            logger.error(f"Error extracting text from {file_path}: {str(e)}")33            return None34    35    @staticmethod36    def _extract_pdf(file_path: str) -> str:37        """Extract text from PDF as Markdown using PyMuPDF4LLM"""38        try:39            # Convert PDF to Markdown40            # This handles tables, images (metadata), and headers much better than raw text extraction41            text = pymupdf4llm.to_markdown(file_path)42            return text.strip()43        except Exception as e:44            logger.error(f"PyMuPDF4LLM extraction failed: {str(e)}")45            # Fallback (optional, but let's fail hard or return empty for now to debug)46            raise e47    48    @staticmethod49    def _extract_docx(file_path: str) -> str:50        """Extract text from DOCX"""51        doc = Document(file_path)52        text = "\n".join([paragraph.text for paragraph in doc.paragraphs])53        return text.strip()54    55    @staticmethod56    def _extract_txt(file_path: str) -> str:57        """Extract text from TXT"""58        with open(file_path, 'r', encoding='utf-8', errors='ignore') as f:59            return f.read().strip()60    61    @staticmethod62    def _extract_html(file_path: str) -> str:63        """Extract text from HTML"""64        with open(file_path, 'r', encoding='utf-8', errors='ignore') as f:65            soup = BeautifulSoup(f.read(), 'html.parser')66            return soup.get_text().strip()67    68    @staticmethod69    def _extract_markdown(file_path: str) -> str:70        """Extract text from Markdown"""71        with open(file_path, 'r', encoding='utf-8', errors='ignore') as f:72            md_text = f.read()73            html = markdown.markdown(md_text)74            soup = BeautifulSoup(html, 'html.parser')75            return soup.get_text().strip()76