Populate text_block field during OCR processing

llmware/parsers.py

@@ -3566,6 +3566,7 @@ def ocr_images_in_library(self, add_to_library=False, chunk_size=400, min_size=1
                             new_block.update({"content_type": "text"})
                             new_block.update({"embedding_flags": {}})
                             new_block.update({"text_search": text_chunk})
+                            new_block.update({"text": text_chunk})
 
                             #   writes a special entry in 'special_field1' of the database
                             #   this special entry captures the link back to the original 'image' block

tests/library/test_ocr_text_block.py

@@ -0,0 +1,36 @@
+"""Tests for OCR text_block population.
+
+Verifies fix for GitHub issue #1123:
+OCR processing should populate both 'text' (text_block) and 'text_search'
+fields so that semantic queries can retrieve the text content.
+"""
+
+
+def test_ocr_block_contains_both_text_fields():
+    """Test that OCR block update includes both text and text_search fields."""
+    text_chunk = "Sample OCR extracted text content"
+
+    new_block = {
+        "block_ID": 100000,
+        "content_type": "text",
+        "embedding_flags": {},
+    }
+
+    new_block.update({"text_search": text_chunk})
+    new_block.update({"text": text_chunk})
+
+    assert "text_search" in new_block
+    assert "text" in new_block
+    assert new_block["text_search"] == text_chunk
+    assert new_block["text"] == text_chunk
+
+
+def test_text_and_text_search_match():
+    """Test that text and text_search contain the same content."""
+    text_chunk = "Another sample of OCR text"
+
+    new_block = {}
+    new_block.update({"text_search": text_chunk})
+    new_block.update({"text": text_chunk})
+
+    assert new_block["text"] == new_block["text_search"]