ocr-service/test_metrics.py

@@ -444,6 +444,38 @@ async def test_ocr_models_ready_gauge_is_one_after_lifespan_startup(fresh_metric
             assert fresh_metrics.ocr_models_ready._value.get() == 1.0
 
 
+@pytest.mark.asyncio
+async def test_ocr_jobs_total_not_incremented_when_pdf_download_fails_in_stream(fresh_metrics):
+    """If `_download_and_convert_pdf` raises, ocr_jobs_total is NOT incremented.
+
+    Mirrors the /ocr endpoint's semantics: the counter only records jobs that
+    actually started OCR work, not failed downloads.
+    """
+    async def fail_download(url):
+        raise RuntimeError("synthetic download failure")
+
+    with patch("main.kraken_engine.load_models"), \
+         patch("main.load_spell_checker"), \
+         patch("main._download_and_convert_pdf", new=fail_download):
+        transport = ASGITransport(app=app, raise_app_exceptions=False)
+        async with AsyncClient(transport=transport, base_url="http://test") as client:
+            import main as main_module
+            main_module._models_ready = True
+            try:
+                response = await client.post("/ocr/stream", json={
+                    "pdfUrl": "http://minio/doc.pdf",
+                    "scriptType": "TYPEWRITER",
+                    "language": "de",
+                })
+            finally:
+                main_module._models_ready = False
+
+    assert response.status_code == 500
+    assert fresh_metrics.ocr_jobs_total.labels(
+        engine="surya", script_type="TYPEWRITER"
+    )._value.get() == 0.0
+
+
 def test_uvicorn_access_log_filter_skips_metrics_path():
     """The MetricsPathFilter drops uvicorn.access log records that target /metrics."""
     import logging as _logging