Update app.py

app.py

@@ -8,10 +8,9 @@ import logging
 import asyncio
 import uuid
 import httpx
-import pymupdf  # Replaced 'fitz' to fix the deprecation warning
+import pymupdf
 from PIL import Image
 from pypdf import PdfReader, PdfWriter
-# FIX: Rename the Gradio handle_file to avoid collision with the Telegram handler
 from gradio_client import Client, handle_file as gradio_handle_file
 from telegram import Update
 from telegram.ext import (
@@ -54,7 +53,6 @@ except Exception as e:
 
 # --- 4. PDF Handling & Image Conversion ---
 def split_pdf_by_size(input_path: str, max_size_mb: float = MAX_PDF_SIZE_MB) -> list[str]:
-    """Splits a PDF into chunks smaller than max_size_mb."""
     chunks = []
     temp_dir = tempfile.mkdtemp(prefix="ocr_chunks_")
     reader = PdfReader(input_path)
@@ -93,12 +91,10 @@ def split_pdf_by_size(input_path: str, max_size_mb: float = MAX_PDF_SIZE_MB) ->
     return chunks
 
 def convert_pdf_to_images(pdf_path: str, output_dir: str) -> list[str]:
-    """Converts each page of a PDF to a high-res PNG image using PyMuPDF."""
     doc = pymupdf.open(pdf_path)
     image_paths = []
     for page_num in range(len(doc)):
         page = doc.load_page(page_num)
-        # Render at 2x zoom for better OCR accuracy
         pix = page.get_pixmap(matrix=pymupdf.Matrix(2, 2))
         img_path = os.path.join(output_dir, f"page_{page_num + 1}.png")
         pix.save(img_path)
@@ -106,7 +102,7 @@ def convert_pdf_to_images(pdf_path: str, output_dir: str) -> list[str]:
     doc.close()
     return image_paths
 
-# --- 5. OCR Call with Robust Image Normalization ---
+# --- 5. OCR Call ---
 def call_space_ocr(image_path: str) -> str:
     """
     Converts the image to a standard RGB JPEG, wraps it in gradio_handle_file, 
@@ -120,14 +116,13 @@ def call_space_ocr(image_path: str) -> str:
             rgb_img = img.convert("RGB")
             rgb_img.save(clean_image_path, "JPEG")
             
-        # 2. Call the API with gradio_handle_file and verbose=True
+        # 2. Call the API (verbose flag removed for compatibility)
         result = gradio_client.predict(
-            image=gradio_handle_file(clean_image_path), # FIX: Use the renamed import
+            image=gradio_handle_file(clean_image_path),
             model_size="Gundam (Recommended)",
             task_type="📄 Convert to Markdown",
             ref_text=" ",
             api_name="/process_ocr_task",
-            verbose=True, # Prints exact server errors if it fails
         )
         
         if isinstance(result, tuple) and len(result) > 0:
@@ -138,7 +133,6 @@ def call_space_ocr(image_path: str) -> str:
         logger.exception(f"OCR call failed for {image_path}")
         raise exc
     finally:
-        # 3. Cleanup
         if os.path.exists(clean_image_path):
             try:
                 os.remove(clean_image_path)
@@ -147,7 +141,6 @@ def call_space_ocr(image_path: str) -> str:
 
 # --- 6. Record Parsing ---
 def parse_records_from_text(raw_text: str) -> list[dict]:
-    # Look for <|ref|>Name<|/ref|><|det|>[[x1,y1,x2,y2]]<|/det|>
     pattern = re.compile(
         r"<\|ref\|>(.*?)<\|/ref\|>\s*<\|det\|>\[\[(\d+),\s*(\d+),\s*(\d+),\s*(\d+)\]\]<\|/det\|>",
         re.DOTALL