Fix data modeling data type Enum dumping

CHANGELOG.md

@@ -21,6 +21,10 @@ Changes are grouped as follows
 ### Added
 - Support for the `/simulators` and `/simulators/integration` API endpoints.
 
+## [7.72.1] - 2025-01-14
+### Fixed
+- Data Modeling container type Enum now dumps correctly and no longer camelCases the user-defined values.
+
 ## [7.72.0] - 2025-01-14
 ### Added
 - Document Summary and Document Question Answering endpoints from the AI API.

cognite/client/_api/ai/tools/documents.py

@@ -6,7 +6,6 @@
 from cognite.client._api_client import APIClient
 from cognite.client.data_classes.ai import Answer, AnswerLanguage, Summary
 from cognite.client.data_classes.data_modeling import NodeId
-from cognite.client.exceptions import CogniteAPIError
 from cognite.client.utils._identifier import IdentifierSequenceWithInstanceId
 
 
@@ -18,7 +17,7 @@ def summarize(
         id: int | None = None,
         external_id: str | None = None,
         instance_id: NodeId | None = None,
-    ) -> Summary | None:
+    ) -> Summary:
         """`Summarize a document using a Large Language Model. <https://developer.cognite.com/api#tag/Document-AI/operation/document_questioning_api_v1_projects__projectName__ai_tools_documents_ask_post>`_
 
         Note:
@@ -31,11 +30,11 @@ def summarize(
             instance_id (NodeId | None): The instance ID of the document
 
         Returns:
-            Summary | None: A summary of the document if the document exists else None.
+            Summary: A summary of the document.
 
         Examples:
 
-            Summarize a single document with id:
+            Summarize a single document using ID:
 
                 >>> from cognite.client import CogniteClient
                 >>> client = CogniteClient()
@@ -49,13 +48,8 @@ def summarize(
                 ... )
         """
         ident = IdentifierSequenceWithInstanceId.load(id, external_id, instance_id).as_singleton()
-        try:
-            res = self._post(self._RESOURCE_PATH + "/summarize", json={"items": ident.as_dicts()})
-            return Summary._load(res.json()["items"][0])
-        except CogniteAPIError as e:
-            if e.code != 404:
-                raise
-            return None
+        res = self._post(self._RESOURCE_PATH + "/summarize", json={"items": ident.as_dicts()})
+        return Summary._load(res.json()["items"][0])
 
     def ask_question(
         self,
@@ -94,7 +88,7 @@ def ask_question(
             instance_id (NodeId | Sequence[NodeId] | None): The instance ID(s) of the document(s)
             language (AnswerLanguage | Literal['Chinese', 'Dutch', 'English', 'French', 'German', 'Italian', 'Japanese', 'Korean', 'Latvian', 'Norwegian', 'Portuguese', 'Spanish', 'Swedish']): The desired language of the answer, defaults to English.
             additional_context (str | None): Additional context that you want the LLM to take into account.
-            ignore_unknown_ids (bool): Whether to skip documents that do not exists or that are not fully processed, instead of throwing an error
+            ignore_unknown_ids (bool): Whether to skip documents that do not exist or that are not fully processed, instead of throwing an error. If no valid documents are found, an error will always be raised.
 
         Returns:
             Answer: The answer to the question in the form of a list of multiple content objects, each consisting of a chunk of text along with a set of references.
@@ -105,9 +99,9 @@ def ask_question(
 
                 >>> from cognite.client import CogniteClient
                 >>> client = CogniteClient()
-                >>> client.ai.tools.documents.ask(
+                >>> client.ai.tools.documents.ask_question(
                 ...     question="What model pump was used?",
-                ...     ids=123,
+                ...     id=123,
                 ... )
 
             Ask a question about multiple documents referenced using external IDs, and instance ID
@@ -117,7 +111,7 @@ def ask_question(
                 >>> from cognite.client.data_classes.ai import AnswerLanguage
                 >>> client.ai.tools.documents.ask_question(
                 ...     question="What other pumps are available?",
-                ...     external_ids=["foo", "bar"],
+                ...     external_id=["foo", "bar"],
                 ...     instance_id=NodeId("my-space", "my-xid"),
                 ...     language=AnswerLanguage.German,
                 ... )

cognite/client/_api_client.py

@@ -1336,6 +1336,7 @@ def _raise_api_error(self, res: Response, payload: dict) -> NoReturn:
                     f"REDIRECT AFTER HTTP Error {res_hist.status_code} {res_hist.request.method} {res_hist.request.url}: {res_hist.content.decode()}"
                 )
         logger.debug(f"HTTP Error {code} {res.request.method} {res.request.url}: {msg}", extra=error_details)
+        # TODO: We should throw "CogniteNotFoundError" if missing is populated and CogniteDuplicatedError when duplicated...
         raise CogniteAPIError(
             msg,
             code,

cognite/client/_version.py

@@ -1,5 +1,5 @@
 from __future__ import annotations
 
-__version__ = "7.72.0"
+__version__ = "7.72.1"
 
 __api_subversion__ = "20230101"

cognite/client/data_classes/data_modeling/containers.py

@@ -20,7 +20,7 @@
     PropertyType,
 )
 from cognite.client.data_classes.data_modeling.ids import ContainerId
-from cognite.client.utils._text import convert_all_keys_to_camel_case_recursive
+from cognite.client.utils._text import convert_all_keys_to_camel_case_recursive, to_camel_case
 
 if TYPE_CHECKING:
     from cognite.client import CogniteClient
@@ -276,8 +276,8 @@ def dump(self, camel_case: bool = True) -> dict[str, str | dict]:
         output["immutable"] = self.immutable
         for key in ["nullable", "auto_increment", "name", "default_value", "description"]:
             if (value := getattr(self, key)) is not None:
-                output[key] = value
-        return convert_all_keys_to_camel_case_recursive(output) if camel_case else output
+                output[to_camel_case(key) if camel_case else key] = value
+        return output
 
 
 @dataclass(frozen=True)

cognite/client/data_classes/data_modeling/data_types.py

@@ -250,3 +250,12 @@ class Enum(PropertyType):
     _type = "enum"
     values: dict[str, EnumValue]
     unknown_value: str | None = None
+
+    def dump(self, camel_case: bool = True) -> dict[str, Any]:
+        output = {
+            "type": self._type,
+            "values": {k: v.dump(camel_case) for k, v in self.values.items()},
+        }
+        if self.unknown_value is not None:
+            output["unknownValue" if camel_case else "unknown_value"] = self.unknown_value
+        return output

pyproject.toml

@@ -1,7 +1,7 @@
 [tool.poetry]
 name = "cognite-sdk"
 
-version = "7.72.0"
+version = "7.72.1"
 
 description = "Cognite Python SDK"
 readme = "README.md"

tests/tests_integration/test_api/test_workflows.py

@@ -2,7 +2,6 @@
 
 import time
 import unittest
-from datetime import datetime
 
 import pytest
 
@@ -287,10 +286,10 @@ def workflow_scheduled_trigger(cognite_client: CogniteClient, workflow_setup_pre
             metadata={"test": "integration_schedule"},
         )
     )
-    # have to sleep until workflow is triggered because it's the only way to properly test get_trigger_run_history
-    now = datetime.now()
-    seconds_til_next_minute = 60 - now.second + 30  # 30 seconds buffer, needed from experience...
-    time.sleep(seconds_til_next_minute)
+    # Have to sleep until workflow is triggered because it's the only way to properly test get_trigger_run_history
+    # ...and as of Jan 14, 2025, there's "an artificial delay of 90 sec to scheduled triggers to prevent a stampede
+    # effect on downstream services for overlapping triggers (up to two minutes, stable per trigger)".
+    time.sleep(90 + 60)
 
     yield trigger
     cognite_client.workflows.triggers.delete(trigger.external_id)

tests/tests_unit/test_data_classes/test_data_models/test_containers.py

@@ -1,6 +1,7 @@
 import pytest
 
 from cognite.client.data_classes._base import UnknownCogniteObject
+from cognite.client.data_classes.data_modeling import data_types
 from cognite.client.data_classes.data_modeling.containers import (
     Constraint,
     Container,
@@ -43,6 +44,18 @@ def test_load_dump__only_required(self, data: dict) -> None:
         actual = ContainerProperty.load(data).dump(camel_case=True)
         assert data == actual
 
+    def test_dump_no_longer_camelCases_everything_when_used(self) -> None:
+        cp = ContainerProperty(
+            data_types.Enum(
+                {
+                    "Closed_I_think": data_types.EnumValue("Valve_is_closed"),
+                    "Opened or not": data_types.EnumValue("Valve is opened"),
+                }
+            )
+        )
+        assert ContainerProperty._load(cp.dump()) == cp
+        assert sorted(cp.dump(camel_case=True)["type"]["values"]) == ["Closed_I_think", "Opened or not"]  # type: ignore [index]
+
 
 class TestConstraint:
     @pytest.mark.parametrize(

tests/tests_unit/test_data_classes/test_data_models/test_data_types.py

@@ -3,6 +3,7 @@
 from cognite.client.data_classes._base import UnknownCogniteObject
 from cognite.client.data_classes.data_modeling.data_types import (
     DirectRelationReference,
+    Enum,
     PropertyType,
     PropertyTypeWithUnit,
     Text,
@@ -66,7 +67,7 @@ def test_load_ignore_unknown_properties(self) -> None:
         data.pop("unknownProperty")
         assert data == actual
 
-    def test_load_dump_unkown_property(self) -> None:
+    def test_load_dump_unknown_property(self) -> None:
         data = {"type": "unknowngibberish", "list": True, "unit": {"externalId": "known"}}
         obj = PropertyType.load(data)
         assert isinstance(obj, UnknownCogniteObject)
@@ -82,6 +83,29 @@ def test_load_text_without_collation(self) -> None:
         data["collation"] = "ucs_basic"
         assert data == actual
 
+    def test_dump_enum_no_camel_casing_of_user_values(self) -> None:
+        obj = PropertyType.load(
+            {
+                "type": "enum",
+                "values": {
+                    "string_valWe ird_Case": {
+                        "name": "string_valWe ird_Case",
+                        "description": "Time series with string data points.",
+                    },
+                    "numeric_valWe ird_Case": {
+                        "name": "numeric_valWe ird_Case",
+                        "description": "Time series with double floating point data points.",
+                    },
+                },
+            }
+        )
+        assert isinstance(obj, Enum)
+
+        expected_values = ["numeric_valWe ird_Case", "string_valWe ird_Case"]
+        assert sorted(obj.values) == expected_values
+        assert sorted(obj.dump(camel_case=False)["values"]) == expected_values
+        assert sorted(obj.dump(camel_case=True)["values"]) == expected_values
+
 
 class TestUnitSupport:
     @pytest.mark.parametrize(

tests/tests_unit/test_docstring_examples.py

@@ -7,6 +7,7 @@
 
 from cognite.client import _cognite_client, config, credentials
 from cognite.client._api import (
+    ai,
     assets,
     data_sets,
     datapoints,
@@ -135,3 +136,8 @@ def test_hosted_extractors(self):
     def test_postgres_gateway(self):
         run_docstring_tests(postgres_gateway_users)
         run_docstring_tests(postgres_gateway_tables)
+
+    def test_ai(self):
+        run_docstring_tests(ai)
+        run_docstring_tests(ai.tools)
+        run_docstring_tests(ai.tools.documents)