From b344ef6f592ca6411cd1a471998cd5e10f4a7097 Mon Sep 17 00:00:00 2001
From: keboss-m <85340750+keboss-m@users.noreply.github.com>
Date: Mon, 1 Jun 2026 13:24:49 +0300
Subject: [PATCH] Fix _run_rapidocr: rapidocr API

---
 process_any_pdf.py | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

diff --git a/process_any_pdf.py b/process_any_pdf.py
index 01a7197..218d94a 100644
--- a/process_any_pdf.py
+++ b/process_any_pdf.py
@@ -157,13 +157,12 @@ def process_pdf(pdf_path: Path, out_dir: Path, use_qwen: bool = False, use_tilin
 def _run_rapidocr(img_path: Path):
     res = engine(img_path)
     ocr_lines = []
-    if res and res[0] is not None:
-        for item in res[0]:
-            box, txt, score = item
+    if res and res.txts is not None:
+        for txt, box, score in zip(res.txts, res.boxes, res.scores):
             ocr_lines.append({
                 "text": txt,
                 "confidence": float(score),
-                "bbox": box
+                "bbox": box.tolist() if hasattr(box, "tolist") else box
             })
     return ocr_lines