From 696b71da5abe63915f823ac1a58740c51b89c79e Mon Sep 17 00:00:00 2001
From: Marcel <marcel@familienarchiv>
Date: Thu, 21 May 2026 16:03:37 +0200
Subject: [PATCH] feat(ocr): increment ocr_jobs_total with engine and
 script_type labels

Pick engine="kraken" for HANDWRITING_KURRENT, engine="surya" otherwise,
then increment after the blocks have been extracted.

Refs #652 (AC2)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---
 ocr-service/main.py         |  3 +++
 ocr-service/test_metrics.py | 38 +++++++++++++++++++++++++++++++++++++
 2 files changed, 41 insertions(+)

diff --git a/ocr-service/main.py b/ocr-service/main.py
index 0d068d15..83a092ee 100644
--- a/ocr-service/main.py
+++ b/ocr-service/main.py
@@ -106,6 +106,7 @@ async def run_ocr(request: OcrRequest):
         del img
 
     script_type = request.scriptType.upper()
+    engine_name = "kraken" if script_type == "HANDWRITING_KURRENT" else "surya"
 
     if script_type == "HANDWRITING_KURRENT":
         if not kraken_engine.is_available():
@@ -119,6 +120,8 @@ async def run_ocr(request: OcrRequest):
         # TYPEWRITER, HANDWRITING_LATIN, UNKNOWN — all use Surya
         blocks = await asyncio.to_thread(surya_engine.extract_blocks, images, request.language)
 
+    metrics.ocr_jobs_total.labels(engine=engine_name, script_type=script_type).inc()
+
     threshold = get_threshold(script_type)
     for block in blocks:
         if block.get("words"):
diff --git a/ocr-service/test_metrics.py b/ocr-service/test_metrics.py
index 0906beec..2f7b0db8 100644
--- a/ocr-service/test_metrics.py
+++ b/ocr-service/test_metrics.py
@@ -91,3 +91,41 @@ def test_build_metrics_registers_all_custom_metrics_on_given_registry():
     # A second registry yields a separate container — no shared state.
     other_metrics = build_metrics(CollectorRegistry())
     assert metrics is not other_metrics
+
+
+async def _drive_ocr(client: AsyncClient, *, script_type: str) -> None:
+    """Helper — fires /ocr with a single mocked page and asserts a 200."""
+    response = await client.post("/ocr", json={
+        "pdfUrl": "http://minio/doc.pdf",
+        "scriptType": script_type,
+        "language": "de",
+    })
+    assert response.status_code == 200, response.text
+
+
+@pytest.mark.asyncio
+async def test_ocr_jobs_total_incremented_with_kraken_engine_label_for_kurrent(fresh_metrics):
+    """A /ocr call with HANDWRITING_KURRENT increments engine=kraken."""
+    mock_images = [Image.new("RGB", (100, 100))]
+    mock_blocks = [{"pageNumber": 1, "x": 0.0, "y": 0.0, "width": 1.0, "height": 1.0,
+                    "polygon": None, "text": "hi", "words": []}]
+
+    with patch("main.kraken_engine.load_models"), \
+         patch("main.load_spell_checker"), \
+         patch("main.correct_text", side_effect=lambda t: t), \
+         patch("main._download_and_convert_pdf", new_callable=AsyncMock, return_value=mock_images), \
+         patch("main.preprocess_page", side_effect=lambda img: img), \
+         patch("main.kraken_engine.is_available", return_value=True), \
+         patch("main.kraken_engine.extract_blocks", return_value=mock_blocks):
+        async with AsyncClient(transport=ASGITransport(app=app), base_url="http://test") as client:
+            import main as main_module
+            main_module._models_ready = True
+            try:
+                await _drive_ocr(client, script_type="HANDWRITING_KURRENT")
+            finally:
+                main_module._models_ready = False
+
+    value = fresh_metrics.ocr_jobs_total.labels(
+        engine="kraken", script_type="HANDWRITING_KURRENT"
+    )._value.get()
+    assert value == 1.0