Fix schema validation and add custom validators

luigi/parameter.py

@@ -1092,6 +1092,26 @@ def run(self):
 
         $ luigi --module my_tasks MyTask --tags '{"role": "UNKNOWN_VALUE", "env": "staging"}'
 
+    Finally, the provided schema can be a custom validator:
+
+    .. code-block:: python
+
+        custom_validator = jsonschema.Draft4Validator(
+          schema={
+            "type": "object",
+            "patternProperties": {
+              ".*": {"type": "string", "enum": ["web", "staging"]},
+            }
+          }
+        )
+
+        class MyTask(luigi.Task):
+          tags = luigi.DictParameter(schema=custom_validator)
+
+          def run(self):
+            logging.info("Find server with role: %s", self.tags['role'])
+            server = aws.ec2.find_my_resource(self.tags)
+
     """
 
     def __init__(
@@ -1105,7 +1125,9 @@ def __init__(
                 "The 'jsonschema' package is not installed so the parameter can not be validated "
                 "even though a schema is given."
             )
-        self.schema = schema
+            self.schema = None
+        else:
+            self.schema = schema
         super().__init__(
             *args,
             **kwargs,
@@ -1117,7 +1139,12 @@ def normalize(self, value):
         """
         frozen_value = recursively_freeze(value)
         if self.schema is not None:
-            jsonschema.validate(instance=recursively_unfreeze(frozen_value), schema=self.schema)
+            unfrozen_value = recursively_unfreeze(frozen_value)
+            try:
+                self.schema.validate(unfrozen_value)  # Validators may update the instance inplace
+                frozen_value = super().normalize(unfrozen_value)
+            except AttributeError:
+                jsonschema.validate(instance=unfrozen_value, schema=self.schema)
         return frozen_value
 
     def parse(self, source):
@@ -1212,6 +1239,31 @@ def run(self):
         $ luigi --module my_tasks MyTask --numbers '[]'  # must have at least 1 element
         $ luigi --module my_tasks MyTask --numbers '[-999, 999]'  # elements must be in [0, 10]
 
+    Finally, the provided schema can be a custom validator:
+
+    .. code-block:: python
+
+        custom_validator = jsonschema.Draft4Validator(
+          schema={
+            "type": "array",
+            "items": {
+              "type": "number",
+              "minimum": 0,
+              "maximum": 10
+            },
+            "minItems": 1
+          }
+        )
+
+        class MyTask(luigi.Task):
+          grades = luigi.ListParameter(schema=custom_validator)
+
+          def run(self):
+                sum = 0
+                for element in self.grades:
+                    sum += element
+                avg = sum / len(self.grades)
+
     """
 
     def __init__(
@@ -1225,7 +1277,9 @@ def __init__(
                 "The 'jsonschema' package is not installed so the parameter can not be validated "
                 "even though a schema is given."
             )
-        self.schema = schema
+            self.schema = None
+        else:
+            self.schema = schema
         super().__init__(
             *args,
             **kwargs,
@@ -1240,7 +1294,12 @@ def normalize(self, x):
         """
         frozen_value = recursively_freeze(x)
         if self.schema is not None:
-            jsonschema.validate(instance=recursively_unfreeze(frozen_value), schema=self.schema)
+            unfrozen_value = recursively_unfreeze(frozen_value)
+            try:
+                self.schema.validate(unfrozen_value)  # Validators may update the instance inplace
+                frozen_value = super().normalize(unfrozen_value)
+            except AttributeError:
+                jsonschema.validate(instance=unfrozen_value, schema=self.schema)
         return frozen_value
 
     def parse(self, x):

test/dict_parameter_test.py

@@ -15,6 +15,7 @@
 # limitations under the License.
 #
 
+from jsonschema import Draft4Validator
 from jsonschema.exceptions import ValidationError
 from helpers import unittest, in_parse
 
@@ -113,6 +114,7 @@ def test_schema(self):
         with pytest.raises(ValidationError, match=r"'UNKNOWN_VALUE' is not one of \['web', 'staging'\]"):
             b.normalize({"role": "UNKNOWN_VALUE", "env": "staging"})
 
+        # Check that warnings are properly emitted
         with mock.patch('luigi.parameter._JSONSCHEMA_ENABLED', False):
             with pytest.warns(
                 UserWarning,
@@ -122,3 +124,17 @@ def test_schema(self):
                 )
             ):
                 luigi.ListParameter(schema={"type": "object"})
+
+        # Test with a custom validator
+        validator = Draft4Validator(
+            schema={
+              "type": "object",
+              "patternProperties": {
+                ".*": {"type": "string", "enum": ["web", "staging"]},
+              },
+            }
+        )
+        c = luigi.DictParameter(schema=validator)
+        c.normalize({"role": "web", "env": "staging"})
+        with pytest.raises(ValidationError, match=r"'UNKNOWN_VALUE' is not one of \['web', 'staging'\]"):
+            c.normalize({"role": "UNKNOWN_VALUE", "env": "staging"})

test/list_parameter_test.py

@@ -15,6 +15,7 @@
 # limitations under the License.
 #
 
+from jsonschema import Draft4Validator
 from jsonschema.exceptions import ValidationError
 from helpers import unittest, in_parse
 
@@ -76,10 +77,7 @@ def test_schema(self):
         )
 
         # Check that the default value is validated
-        with pytest.raises(
-            ValidationError,
-            match=r"'INVALID_ATTRIBUTE' is not of type 'number'",
-        ):
+        with pytest.raises(ValidationError, match=r"'INVALID_ATTRIBUTE' is not of type 'number'"):
             a.normalize(["INVALID_ATTRIBUTE"])
 
         # Check that empty list is not valid
@@ -100,6 +98,7 @@ def test_schema(self):
         with pytest.raises(ValidationError, match="-999 is less than the minimum of 0"):
             a.normalize(invalid_list_value)
 
+        # Check that warnings are properly emitted
         with mock.patch('luigi.parameter._JSONSCHEMA_ENABLED', False):
             with pytest.warns(
                 UserWarning,
@@ -109,3 +108,20 @@ def test_schema(self):
                 )
             ):
                 luigi.ListParameter(schema={"type": "array", "items": {"type": "number"}})
+
+        # Test with a custom validator
+        validator = Draft4Validator(
+            schema={
+                "type": "array",
+                "items": {
+                    "type": "number",
+                    "minimum": 0,
+                    "maximum": 10,
+                },
+                "minItems": 1,
+            }
+        )
+        c = luigi.DictParameter(schema=validator)
+        c.normalize(valid_list)
+        with pytest.raises(ValidationError, match=r"'INVALID_ATTRIBUTE' is not of type 'number'",):
+            c.normalize(["INVALID_ATTRIBUTE"])