Spaces:
Sleeping
Sleeping
| from fastapi import FastAPI, File, UploadFile | |
| from fastapi.responses import HTMLResponse | |
| import fitz # PyMuPDF | |
| import pytesseract | |
| from PIL import Image | |
| import io | |
| from pdfminer.high_level import extract_pages | |
| from pdfminer.layout import LTTextContainer | |
| app = FastAPI() | |
| async def home(): | |
| html_content = """ | |
| <!DOCTYPE html> | |
| <html> | |
| <head> | |
| <title>PDF Text Extraction API</title> | |
| <style> | |
| body { font-family: Arial, sans-serif; padding: 2rem; background: #f0f0f0; } | |
| .container { max-width: 600px; margin: auto; background: white; padding: 2rem; border-radius: 8px; box-shadow: 0 0 10px rgba(0,0,0,0.1);} | |
| h1 { color: #333; } | |
| p { color: #555; } | |
| a { color: #1a73e8; text-decoration: none; } | |
| a:hover { text-decoration: underline; } | |
| </style> | |
| </head> | |
| <body> | |
| <div class="container"> | |
| <h1>Welcome to PDF Text Extraction API</h1> | |
| <p>This API allows you to upload PDFs and extract text — including optional OCR for images.</p> | |
| <h2>Available endpoints:</h2> | |
| <ul> | |
| <li><b>POST /extract-text</b> - Extract plain text from PDF pages.</li> | |
| <li><b>POST /extract-text-ocr</b> - Extract text including OCR from image-based PDFs.</li> | |
| <li><b>POST /extract-text-structured</b> - Extract structured text using pdfminer.</li> | |
| </ul> | |
| <p>Use a tool like <a href="https://www.postman.com/" target="_blank">Postman</a> or write your own client to send PDF files to the endpoints.</p> | |
| </div> | |
| </body> | |
| </html> | |
| """ | |
| return HTMLResponse(content=html_content, status_code=200) | |
| async def extract_text(file: UploadFile = File(...)): | |
| try: | |
| contents = await file.read() | |
| doc = fitz.open(stream=contents, filetype="pdf") | |
| extracted_text = "" | |
| for i, page in enumerate(doc): | |
| extracted_text += f"\n\n--- Page {i + 1} ---\n\n" + page.get_text() | |
| return {"filename": file.filename, "text": extracted_text} | |
| except Exception as e: | |
| return {"error": str(e)} | |
| async def extract_text_ocr(file: UploadFile = File(...)): | |
| try: | |
| contents = await file.read() | |
| doc = fitz.open(stream=contents, filetype="pdf") | |
| full_text = "" | |
| for i in range(len(doc)): | |
| page = doc.load_page(i) | |
| # Normal text | |
| text = page.get_text() | |
| # Render page to an image | |
| pix = page.get_pixmap() | |
| img = Image.open(io.BytesIO(pix.tobytes())) | |
| # OCR text | |
| ocr_text = pytesseract.image_to_string(img) | |
| full_text += f"\n\n--- Page {i + 1} ---\n\n" | |
| full_text += text + "\n" | |
| full_text += "[OCR Text]\n" + ocr_text | |
| return {"filename": file.filename, "text": full_text} | |
| except Exception as e: | |
| return {"error": str(e)} | |
| async def extract_text_structured(file: UploadFile = File(...)): | |
| try: | |
| contents = await file.read() | |
| # Save to temp file to use with extract_pages | |
| import tempfile | |
| with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as tmp_file: | |
| tmp_file.write(contents) | |
| temp_pdf_path = tmp_file.name | |
| structured_text = "" | |
| for i, page_layout in enumerate(extract_pages(temp_pdf_path)): | |
| structured_text += f"\n\n--- Page {i + 1} ---\n\n" | |
| for element in page_layout: | |
| if isinstance(element, LTTextContainer): | |
| structured_text += element.get_text() | |
| return {"filename": file.filename, "text": structured_text} | |
| except Exception as e: | |
| return {"error": str(e)} | |