test(ocr): add business-logic tests for polygon extraction, Kraken routing, and confidence markers

Cover Surya polygon/word-level extraction, health endpoint states,
Kraken script-type routing, 503 when models not ready, 400 when
Kraken unavailable for Kurrent, and confidence marker application
during streaming. Production code coverage: 88%.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

ocr-service/test_engines.py

@@ -45,6 +45,52 @@ def test_surya_extract_page_blocks_returns_blocks_for_single_image():
     assert blocks[0]["height"] == 20 / 200
 
 
+def test_surya_extract_page_blocks_extracts_polygon_when_present():
+    image = Image.new("RGB", (100, 200))
+
+    mock_line = _make_surya_line("Text", [10, 20, 90, 40])
+    mock_line.polygon = [(10, 20), (90, 20), (90, 40), (10, 40)]
+
+    mock_pred = MagicMock()
+    mock_pred.text_lines = [mock_line]
+
+    with patch.object(surya, "_recognition_predictor") as mock_rec, \
+         patch.object(surya, "_loaded", True):
+        mock_rec.return_value = [mock_pred]
+
+        blocks = surya.extract_page_blocks(image, page_idx=1, language="de")
+
+    assert blocks[0]["polygon"] is not None
+    assert len(blocks[0]["polygon"]) == 4
+    assert blocks[0]["polygon"][0] == [10 / 100, 20 / 200]
+
+
+def test_surya_extract_page_blocks_extracts_word_level_confidence():
+    image = Image.new("RGB", (100, 200))
+
+    word1 = MagicMock()
+    word1.text = "Hallo"
+    word1.confidence = 0.95
+    word2 = MagicMock()
+    word2.text = "Welt"
+    word2.confidence = 0.3
+
+    mock_line = _make_surya_line("Hallo Welt", [10, 20, 90, 40], words=[word1, word2])
+
+    mock_pred = MagicMock()
+    mock_pred.text_lines = [mock_line]
+
+    with patch.object(surya, "_recognition_predictor") as mock_rec, \
+         patch.object(surya, "_loaded", True):
+        mock_rec.return_value = [mock_pred]
+
+        blocks = surya.extract_page_blocks(image, page_idx=1, language="de")
+
+    assert len(blocks[0]["words"]) == 2
+    assert blocks[0]["words"][0]["text"] == "Hallo"
+    assert blocks[0]["words"][0]["confidence"] == 0.95
+
+
 def test_surya_extract_blocks_delegates_to_extract_page_blocks():
     """After refactoring, extract_blocks should produce the same output."""
     image1 = Image.new("RGB", (100, 200))