pdfrenderer.cpp: Ignore non-text blocks

Fix tesseract-ocr#3957.
amitdo · Nov 7, 2022 · 1b38293 · 1b38293
1 parent 490611e
commit 1b38293
Showing 1 changed file with 4 additions and 0 deletions.
diff --git a/src/api/pdfrenderer.cpp b/src/api/pdfrenderer.cpp
@@ -354,6 +354,10 @@ 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 (!res_it->PTIsTextType(block_type)) {
+        continue; // ignore non-text blocks
+      }
       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