Update app.py

app.py

@@ -281,7 +281,8 @@ def run_ocr(uploaded_file):
                         hdr = extract_pdf_text_in_band(path, page_num, 0, PDF_HEADER_BAND_FRAC)
                 if not (hdr and hdr.strip()):
                     hdr = ocr_zone(img_path, 0, he)
-                if hdr and hdr.strip():
+                # Skip header if already present in page_md (avoids duplicating body content)
+                if hdr and hdr.strip() and hdr.strip()[:80] not in page_md:
                     parts.append(hdr.strip())
 
             if page_md: