replace assertions with errors & add invalid detector name tests

Author: saichandrapandraju

detectors/built_in/app.py

@@ -32,7 +32,8 @@ def detect_content(request: ContentAnalysisHttpRequest):
     for content in request.contents:
         message_detections = []
         for detector_kind, detector_registry in app.get_all_detectors().items():
-            assert isinstance(detector_registry, BaseDetectorRegistry), f"Detector {detector_kind} is not a valid BaseDetectorRegistry"
+            if not isinstance(detector_registry, BaseDetectorRegistry):
+                raise TypeError(f"Detector {detector_kind} is not a valid BaseDetectorRegistry")
             if detector_kind in request.detector_params:
                 try:
                     message_detections += detector_registry.handle_request(content, request.detector_params)
@@ -48,7 +49,8 @@ def detect_content(request: ContentAnalysisHttpRequest):
 def get_registry():
     result = {}
     for detector_type, detector_registry in app.get_all_detectors().items():
-        assert isinstance(detector_registry, BaseDetectorRegistry), f"Detector {detector_type} is not a valid BaseDetectorRegistry"
+        if not isinstance(detector_registry, BaseDetectorRegistry):
+            raise TypeError(f"Detector {detector_type} is not a valid BaseDetectorRegistry")
         result[detector_type] = {}
         for detector_name, detector_fn in detector_registry.get_registry().items():
             result[detector_type][detector_name] = detector_fn.__doc__

detectors/llm_judge/app.py

@@ -49,7 +49,7 @@ async def detector_unary_handler(
     """Analyze content using LLM-as-Judge evaluation."""
     detector: LLMJudgeDetector = app.get_detector()
     if not detector:
-        raise HTTPException(status_code=404, detail="Detector not found")
+        raise HTTPException(status_code=503, detail="Detector not found")
     return ContentsAnalysisResponse(root=await detector.run(request))
 
 
@@ -65,7 +65,7 @@ async def list_metrics():
     """List all available evaluation metrics."""
     detector: LLMJudgeDetector = app.get_detector()
     if not detector:
-        raise HTTPException(status_code=404, detail="Detector not found")
+        raise HTTPException(status_code=503, detail="Detector not found")
 
     metrics = detector.list_available_metrics()
     return MetricsListResponse(metrics=metrics, total=len(metrics))

tests/detectors/builtIn/test_filetype.py

@@ -284,4 +284,43 @@ def test_multiple_filetype_valid_and_invalid(self, client: TestClient):
         assert not any(d["detection"] == "invalid_yaml" for d in detections[5])
 
         # 7: invalid yaml
-        assert any(d["detection"] == "invalid_yaml" for d in detections[6])
+        assert any(d["detection"] == "invalid_yaml" for d in detections[6])
+    
+    
+    # === ERROR HANDLING & INVALID DETECTOR TYPES =================================================
+    def test_unregistered_detector_kind_ignored(self, client: TestClient):
+        """Test that requesting an unregistered detector kind is silently ignored"""
+        payload = {
+            "contents": ['{"a": 1}'],
+            "detector_params": {"nonexistent_detector": ["some_value"]}
+        }
+        resp = client.post("/api/v1/text/contents", json=payload)
+        assert resp.status_code == 200
+        # Should return empty list since nonexistent_detector is not registered
+        assert resp.json()[0] == []
+
+    def test_mixed_valid_invalid_detector_kinds(self, client: TestClient):
+        """Test mixing valid and invalid detector kinds"""
+        payload = {
+            "contents": ['{a: 1, b: 2}'],
+            "detector_params": {
+                "file_type": ["json"],  # valid detector kind
+                "nonexistent_detector": ["some_value"]  # invalid detector kind
+            }
+        }
+        resp = client.post("/api/v1/text/contents", json=payload)
+        assert resp.status_code == 200
+        detections = resp.json()[0]
+        # Should only process the valid detector kind
+        assert detections[0]["detection"] == "invalid_json"
+    
+    def test_empty_detector_params(self, client: TestClient):
+        """Test with empty detector_params"""
+        payload = {
+            "contents": ['{"a": 1}'],
+            "detector_params": {}
+        }
+        resp = client.post("/api/v1/text/contents", json=payload)
+        assert resp.status_code == 200
+        # Should return empty list since no detectors are specified
+        assert resp.json()[0] == []

tests/detectors/builtIn/test_regex.py

@@ -198,4 +198,41 @@ def test_single_detector(self, client):
         assert any("[email protected]" in d["text"] for d in results[0])
 
 
+    # === ERROR HANDLING & INVALID DETECTOR TYPES =================================================    
+    def test_unregistered_detector_kind_ignored(self, client):
+        """Test that requesting an unregistered detector kind is silently ignored"""
+        payload = {
+            "contents": ["[email protected]"],
+            "detector_params": {"nonexistent_detector": ["some_value"]}
+        }
+        resp = client.post("/api/v1/text/contents", json=payload)
+        assert resp.status_code == 200
+        # Should return empty list since nonexistent_detector is not registered
+        assert resp.json()[0] == []
+
+    def test_mixed_valid_invalid_detector_kinds(self, client):
+        """Test mixing valid and invalid detector kinds"""
+        payload = {
+            "contents": ["Contact me at [email protected]"],
+            "detector_params": {
+                "regex": ["email"],  # valid detector kind
+                "nonexistent_detector": ["some_value"]  # invalid detector kind
+            }
+        }
+        resp = client.post("/api/v1/text/contents", json=payload)
+        assert resp.status_code == 200
+        detections = resp.json()[0]
+        # Should only process the valid detector kind
+        assert detections[0]["text"] == "[email protected]"
+
+    def test_empty_detector_params(self, client):
+        """Test with empty detector_params"""
+        payload = {
+            "contents": ["[email protected]"],
+            "detector_params": {}
+        }
+        resp = client.post("/api/v1/text/contents", json=payload)
+        assert resp.status_code == 200
+        # Should return empty list since no detectors are specified
+        assert resp.json()[0] == []