From c1964560b667a54ba4eb3bd9fbc7ed26c4cc86a2 Mon Sep 17 00:00:00 2001
From: Amit Dovev <amitdev2222@gmail.com>
Date: Tue, 8 Nov 2022 08:02:09 +0200
Subject: [PATCH] pdfrenderer.cpp: Ignore non-text blocks

Fix #3957.
---
 src/api/pdfrenderer.cpp | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/src/api/pdfrenderer.cpp b/src/api/pdfrenderer.cpp
index 774558aed7..81cf2e24d8 100644
--- a/src/api/pdfrenderer.cpp
+++ b/src/api/pdfrenderer.cpp
@@ -25,6 +25,7 @@
 
 #include <allheaders.h>
 #include <tesseract/baseapi.h>
+#include <tesseract/publictypes.h> // for PTIsTextType()
 #include <tesseract/renderer.h>
 #include <cmath>
 #include <cstring>
@@ -354,6 +355,12 @@ char *TessPDFRenderer::GetPDFTextObjects(TessBaseAPI *api, double width, double
   const std::unique_ptr</*non-const*/ ResultIterator> res_it(api->GetIterator());
   while (!res_it->Empty(RIL_BLOCK)) {
     if (res_it->IsAtBeginningOf(RIL_BLOCK)) {
+      auto block_type = res_it->BlockType();
+      if (!PTIsTextType(block_type)) {
+        // ignore non-text blocks
+        res_it->Next(RIL_BLOCK);
+        continue;
+      }
       pdf_str << "BT\n3 Tr"; // Begin text object, use invisible ink
       old_fontsize = 0;      // Every block will declare its fontsize
       new_block = true;      // Every block will declare its affine matrix