fix(resources): preserve multipart file uploads in generated client (HYBIM-882)

openapi.yaml

@@ -14069,7 +14069,7 @@ components:
       properties:
         file:
           type: string
-          contentMediaType: application/octet-stream
+          format: binary
           title: File
         validation_result:
           type: string
@@ -14102,7 +14102,7 @@ components:
         file:
           anyOf:
             - type: string
-              contentMediaType: application/octet-stream
+              format: binary
             - type: 'null'
           title: File
         copy_from_dataset_id:
@@ -14173,14 +14173,14 @@ components:
         query_files:
           items:
             type: string
-            contentMediaType: application/octet-stream
+            format: binary
           type: array
           title: Query Files
           default: []
         response_files:
           items:
             type: string
-            contentMediaType: application/octet-stream
+            format: binary
           type: array
           title: Response Files
           default: []
@@ -14192,7 +14192,7 @@ components:
       properties:
         file:
           type: string
-          contentMediaType: application/octet-stream
+          format: binary
           title: File
         dataset_id:
           type: string
@@ -14242,7 +14242,7 @@ components:
       properties:
         file:
           type: string
-          contentMediaType: application/octet-stream
+          format: binary
           title: File
         log_stream_id:
           anyOf:
@@ -14301,7 +14301,7 @@ components:
       properties:
         file:
           type: string
-          contentMediaType: application/octet-stream
+          format: binary
           title: File
         test_input:
           anyOf:

scripts/import-openapi-yaml.sh

@@ -32,6 +32,7 @@ curl -s "${HOST_URL}/openapi.json" | poetry run python -c 'import sys, json, yam
 #   - ListDatasetParams.properties.sort.default = "None"
 #   - ListPromptTemplateParams.properties.sort.default = "None"
 #   - ProjectCollectionParams.properties.sort.default = "None"
+#   - ListAnnotationQueueParams.properties.sort.default = "None"
 #   - galileo_core__schemas__shared__scorers__scorer_name__ScorerName.title = "CoreScorerName"
 #   - galileo_core__schemas__shared__scorers__scorer_name__ScorerName.enum |= unique
 #   - /llm_integrations/projects/{project_id}/runs/{run_id}.get.responses[200].schema.title = "GetRunIntegrationsResponse" (Windows filename length fix)
@@ -43,16 +44,40 @@ curl -s "${HOST_URL}/openapi.json" | poetry run python -c 'import sys, json, yam
 #     - NotNode_Annotated_Union_LogRecordsIDFilter__LogRecordsDateFilter__LogRecordsNumberFilter__LogRecordsBooleanFilter__LogRecordsCollectionFilter__LogRecordsTextFilter__LogRecordsFullyAnnotatedFilter___FieldInfo_annotation_NoneType__required_True__discriminator__type____.title = "NotNodeLogRecordsFilter"
 #     - OrNode_Annotated_Union_LogRecordsIDFilter__LogRecordsDateFilter__LogRecordsNumberFilter__LogRecordsBooleanFilter__LogRecordsCollectionFilter__LogRecordsTextFilter__LogRecordsFullyAnnotatedFilter___FieldInfo_annotation_NoneType__required_True__discriminator__type____.title = "OrNodeLogRecordsFilter"
 #   - StepType.enum += ["control"] |= unique: server omits this value from the spec but returns it for some log-stream columns (sc-62628); unique deduplicates if the server spec is later fixed
+#   - multipart/form-data binary fields: contentMediaType = "application/octet-stream" -> format = "binary"
 #
 # If you run into related issues with the auto-generate-api-client.sh script, add the openapi.yaml patches here.
 
 # Apply all patches using yq in a single command
-poetry run python -m yq --in-place -Y '.components.schemas.api__schemas__project_v2__GetProjectsPaginatedResponse.title = "GetProjectsPaginatedResponseV2" | .components.schemas.galileo_core__schemas__shared__message__Message.title = "MessagesListItem" | .components.schemas.galileo_core__schemas__shared__message_role__MessageRole.title = "MessagesListItemRole" | .components.schemas.ListDatasetParams.properties.sort.default = "None" | .components.schemas.ListPromptTemplateParams.properties.sort.default = "None" | .components.schemas.ProjectCollectionParams.properties.sort.default = "None" | .components.schemas.galileo_core__schemas__shared__scorers__scorer_name__ScorerName.title = "CoreScorerName" | .components.schemas.galileo_core__schemas__shared__scorers__scorer_name__ScorerName.enum |= unique | .paths["/llm_integrations/projects/{project_id}/runs/{run_id}"].get.responses["200"].content["application/json"].schema.title = "GetRunIntegrationsResponse" | .components.schemas.Document.properties.content = .components.schemas.Document.properties.page_content | del(.components.schemas.Document.properties.page_content) | .components.schemas.Document.properties.content.title = "Content" | .components.schemas.Document.required = ["content"] | .components.schemas["AndNode_Annotated_Union_LogRecordsIDFilter__LogRecordsDateFilter__LogRecordsNumberFilter__LogRecordsBooleanFilter__LogRecordsCollectionFilter__LogRecordsTextFilter__LogRecordsFullyAnnotatedFilter___FieldInfo_annotation_NoneType__required_True__discriminator__type____"].title = "AndNodeLogRecordsFilter" | .components.schemas["FilterExpression_Annotated_Union_LogRecordsIDFilter__LogRecordsDateFilter__LogRecordsNumberFilter__LogRecordsBooleanFilter__LogRecordsCollectionFilter__LogRecordsTextFilter__LogRecordsFullyAnnotatedFilter___FieldInfo_annotation_NoneType__required_True__discriminator__type____"].title = "FilterExpressionLogRecordsFilter" | .components.schemas["FilterLeaf_Annotated_Union_LogRecordsIDFilter__LogRecordsDateFilter__LogRecordsNumberFilter__LogRecordsBooleanFilter__LogRecordsCollectionFilter__LogRecordsTextFilter__LogRecordsFullyAnnotatedFilter___FieldInfo_annotation_NoneType__required_True__discriminator__type____"].title = "FilterLeafLogRecordsFilter" | .components.schemas["NotNode_Annotated_Union_LogRecordsIDFilter__LogRecordsDateFilter__LogRecordsNumberFilter__LogRecordsBooleanFilter__LogRecordsCollectionFilter__LogRecordsTextFilter__LogRecordsFullyAnnotatedFilter___FieldInfo_annotation_NoneType__required_True__discriminator__type____"].title = "NotNodeLogRecordsFilter" | .components.schemas["OrNode_Annotated_Union_LogRecordsIDFilter__LogRecordsDateFilter__LogRecordsNumberFilter__LogRecordsBooleanFilter__LogRecordsCollectionFilter__LogRecordsTextFilter__LogRecordsFullyAnnotatedFilter___FieldInfo_annotation_NoneType__required_True__discriminator__type____"].title = "OrNodeLogRecordsFilter" | .components.schemas.StepType.enum += ["control"] | .components.schemas.StepType.enum |= unique'  "$HOME_DIR/openapi.yaml"
+poetry run python -m yq --in-place -Y '.components.schemas.api__schemas__project_v2__GetProjectsPaginatedResponse.title = "GetProjectsPaginatedResponseV2" | .components.schemas.galileo_core__schemas__shared__message__Message.title = "MessagesListItem" | .components.schemas.galileo_core__schemas__shared__message_role__MessageRole.title = "MessagesListItemRole" | .components.schemas.ListDatasetParams.properties.sort.default = "None" | .components.schemas.ListPromptTemplateParams.properties.sort.default = "None" | .components.schemas.ProjectCollectionParams.properties.sort.default = "None" | .components.schemas.ListAnnotationQueueParams.properties.sort.default = "None" | .components.schemas.galileo_core__schemas__shared__scorers__scorer_name__ScorerName.title = "CoreScorerName" | .components.schemas.galileo_core__schemas__shared__scorers__scorer_name__ScorerName.enum |= unique | .paths["/llm_integrations/projects/{project_id}/runs/{run_id}"].get.responses["200"].content["application/json"].schema.title = "GetRunIntegrationsResponse" | .components.schemas.Document.properties.content = .components.schemas.Document.properties.page_content | del(.components.schemas.Document.properties.page_content) | .components.schemas.Document.properties.content.title = "Content" | .components.schemas.Document.required = ["content"] | .components.schemas["AndNode_Annotated_Union_LogRecordsIDFilter__LogRecordsDateFilter__LogRecordsNumberFilter__LogRecordsBooleanFilter__LogRecordsCollectionFilter__LogRecordsTextFilter__LogRecordsFullyAnnotatedFilter___FieldInfo_annotation_NoneType__required_True__discriminator__type____"].title = "AndNodeLogRecordsFilter" | .components.schemas["FilterExpression_Annotated_Union_LogRecordsIDFilter__LogRecordsDateFilter__LogRecordsNumberFilter__LogRecordsBooleanFilter__LogRecordsCollectionFilter__LogRecordsTextFilter__LogRecordsFullyAnnotatedFilter___FieldInfo_annotation_NoneType__required_True__discriminator__type____"].title = "FilterExpressionLogRecordsFilter" | .components.schemas["FilterLeaf_Annotated_Union_LogRecordsIDFilter__LogRecordsDateFilter__LogRecordsNumberFilter__LogRecordsBooleanFilter__LogRecordsCollectionFilter__LogRecordsTextFilter__LogRecordsFullyAnnotatedFilter___FieldInfo_annotation_NoneType__required_True__discriminator__type____"].title = "FilterLeafLogRecordsFilter" | .components.schemas["NotNode_Annotated_Union_LogRecordsIDFilter__LogRecordsDateFilter__LogRecordsNumberFilter__LogRecordsBooleanFilter__LogRecordsCollectionFilter__LogRecordsTextFilter__LogRecordsFullyAnnotatedFilter___FieldInfo_annotation_NoneType__required_True__discriminator__type____"].title = "NotNodeLogRecordsFilter" | .components.schemas["OrNode_Annotated_Union_LogRecordsIDFilter__LogRecordsDateFilter__LogRecordsNumberFilter__LogRecordsBooleanFilter__LogRecordsCollectionFilter__LogRecordsTextFilter__LogRecordsFullyAnnotatedFilter___FieldInfo_annotation_NoneType__required_True__discriminator__type____"].title = "OrNodeLogRecordsFilter" | .components.schemas.StepType.enum += ["control"] | .components.schemas.StepType.enum |= unique | .components.schemas["Body_create_code_scorer_version_scorers__scorer_id__version_code_post"].properties.file.format = "binary" | del(.components.schemas["Body_create_code_scorer_version_scorers__scorer_id__version_code_post"].properties.file.contentMediaType) | .components.schemas["Body_create_dataset_datasets_post"].properties.file.anyOf[0].format = "binary" | del(.components.schemas["Body_create_dataset_datasets_post"].properties.file.anyOf[0].contentMediaType) | .components.schemas["Body_manual_llm_validate_multipart_scorers_llm_validate_multipart_post"].properties.query_files.items.format = "binary" | del(.components.schemas["Body_manual_llm_validate_multipart_scorers_llm_validate_multipart_post"].properties.query_files.items.contentMediaType) | .components.schemas["Body_manual_llm_validate_multipart_scorers_llm_validate_multipart_post"].properties.response_files.items.format = "binary" | del(.components.schemas["Body_manual_llm_validate_multipart_scorers_llm_validate_multipart_post"].properties.response_files.items.contentMediaType) | .components.schemas["Body_validate_code_scorer_dataset_scorers_code_validate_dataset_post"].properties.file.format = "binary" | del(.components.schemas["Body_validate_code_scorer_dataset_scorers_code_validate_dataset_post"].properties.file.contentMediaType) | .components.schemas["Body_validate_code_scorer_log_record_scorers_code_validate_log_record_post"].properties.file.format = "binary" | del(.components.schemas["Body_validate_code_scorer_log_record_scorers_code_validate_log_record_post"].properties.file.contentMediaType) | .components.schemas["Body_validate_code_scorer_scorers_code_validate_post"].properties.file.format = "binary" | del(.components.schemas["Body_validate_code_scorer_scorers_code_validate_post"].properties.file.contentMediaType)'  "$HOME_DIR/openapi.yaml"
 
-# Check if the command was successful
-if [ $? -eq 0 ]; then
-  echo "OpenAPI YAML saved to $HOME_DIR/openapi.yaml"
-else
-  echo "Failed to fetch and convert OpenAPI JSON"
+if [ $? -ne 0 ]; then
+  echo "Failed to patch OpenAPI YAML"
   exit 1
 fi
+
+poetry run python -m yq -e '
+  [
+    .components.schemas.Body_create_code_scorer_version_scorers__scorer_id__version_code_post.properties.file.format == "binary",
+    (.components.schemas.Body_create_code_scorer_version_scorers__scorer_id__version_code_post.properties.file | has("contentMediaType") | not),
+    .components.schemas.Body_create_dataset_datasets_post.properties.file.anyOf[0].format == "binary",
+    (.components.schemas.Body_create_dataset_datasets_post.properties.file.anyOf[0] | has("contentMediaType") | not),
+    .components.schemas.Body_manual_llm_validate_multipart_scorers_llm_validate_multipart_post.properties.query_files.items.format == "binary",
+    (.components.schemas.Body_manual_llm_validate_multipart_scorers_llm_validate_multipart_post.properties.query_files.items | has("contentMediaType") | not),
+    .components.schemas.Body_manual_llm_validate_multipart_scorers_llm_validate_multipart_post.properties.response_files.items.format == "binary",
+    (.components.schemas.Body_manual_llm_validate_multipart_scorers_llm_validate_multipart_post.properties.response_files.items | has("contentMediaType") | not),
+    .components.schemas.Body_validate_code_scorer_dataset_scorers_code_validate_dataset_post.properties.file.format == "binary",
+    (.components.schemas.Body_validate_code_scorer_dataset_scorers_code_validate_dataset_post.properties.file | has("contentMediaType") | not),
+    .components.schemas.Body_validate_code_scorer_log_record_scorers_code_validate_log_record_post.properties.file.format == "binary",
+    (.components.schemas.Body_validate_code_scorer_log_record_scorers_code_validate_log_record_post.properties.file | has("contentMediaType") | not),
+    .components.schemas.Body_validate_code_scorer_scorers_code_validate_post.properties.file.format == "binary",
+    (.components.schemas.Body_validate_code_scorer_scorers_code_validate_post.properties.file | has("contentMediaType") | not)
+  ] | all
+' "$HOME_DIR/openapi.yaml" > /dev/null
+
+if [ $? -ne 0 ]; then
+  echo "Failed to patch multipart binary fields in OpenAPI YAML"
+  exit 1
+fi
+
+echo "OpenAPI YAML saved to $HOME_DIR/openapi.yaml"

src/splunk_ao/resources/models/body_create_code_scorer_version_scorers_scorer_id_version_code_post.py

@@ -1,30 +1,33 @@
 from __future__ import annotations
 
 from collections.abc import Mapping
+from io import BytesIO
 from typing import Any, TypeVar
 
 from attrs import define as _attrs_define
 from attrs import field as _attrs_field
 
 from .. import types
+from ..types import File
 
 T = TypeVar("T", bound="BodyCreateCodeScorerVersionScorersScorerIdVersionCodePost")
 
 
 @_attrs_define
 class BodyCreateCodeScorerVersionScorersScorerIdVersionCodePost:
     """
-    Attributes:
+    Attributes
+    ----------
         file (str):
-        validation_result (str): Pre-validated result as JSON string from the validate endpoint
+        validation_result (str): Pre-validated result as JSON string from the validate endpoint.
     """
 
-    file: str
+    file: File
     validation_result: str
     additional_properties: dict[str, Any] = _attrs_field(init=False, factory=dict)
 
     def to_dict(self) -> dict[str, Any]:
-        file = self.file
+        file = self.file.to_tuple()
 
         validation_result = self.validation_result
 
@@ -37,7 +40,7 @@ def to_dict(self) -> dict[str, Any]:
     def to_multipart(self) -> types.RequestFiles:
         files: types.RequestFiles = []
 
-        files.append(("file", (None, str(self.file).encode(), "text/plain")))
+        files.append(("file", self.file.to_tuple()))
 
         files.append(("validation_result", (None, str(self.validation_result).encode(), "text/plain")))
 
@@ -49,7 +52,7 @@ def to_multipart(self) -> types.RequestFiles:
     @classmethod
     def from_dict(cls: type[T], src_dict: Mapping[str, Any]) -> T:
         d = dict(src_dict)
-        file = d.pop("file")
+        file = File(payload=BytesIO(d.pop("file")))
 
         validation_result = d.pop("validation_result")
 

src/splunk_ao/resources/models/body_create_dataset_datasets_post.py

@@ -1,13 +1,14 @@
 from __future__ import annotations
 
 from collections.abc import Mapping
+from io import BytesIO
 from typing import Any, TypeVar, cast
 
 from attrs import define as _attrs_define
 from attrs import field as _attrs_field
 
 from .. import types
-from ..types import UNSET, Unset
+from ..types import UNSET, File, FileTypes, Unset
 
 T = TypeVar("T", bound="BodyCreateDatasetDatasetsPost")
 
@@ -20,7 +21,7 @@ class BodyCreateDatasetDatasetsPost:
         hidden (bool | Unset):  Default: False.
         name (None | str | Unset):
         append_suffix_if_duplicate (bool | Unset):  Default: False.
-        file (None | str | Unset):
+        file (File | None | Unset):
         copy_from_dataset_id (None | str | Unset):
         copy_from_dataset_version_index (int | None | Unset):
         project_id (None | str | Unset):
@@ -31,7 +32,7 @@ class BodyCreateDatasetDatasetsPost:
     hidden: bool | Unset = False
     name: None | str | Unset = UNSET
     append_suffix_if_duplicate: bool | Unset = False
-    file: None | str | Unset = UNSET
+    file: File | None | Unset = UNSET
     copy_from_dataset_id: None | str | Unset = UNSET
     copy_from_dataset_version_index: int | None | Unset = UNSET
     project_id: None | str | Unset = UNSET
@@ -51,9 +52,11 @@ def to_dict(self) -> dict[str, Any]:
 
         append_suffix_if_duplicate = self.append_suffix_if_duplicate
 
-        file: None | str | Unset
+        file: FileTypes | None | Unset
         if isinstance(self.file, Unset):
             file = UNSET
+        elif isinstance(self.file, File):
+            file = self.file.to_tuple()
         else:
             file = self.file
 
@@ -126,8 +129,8 @@ def to_multipart(self) -> types.RequestFiles:
             )
 
         if not isinstance(self.file, Unset):
-            if isinstance(self.file, str):
-                files.append(("file", (None, str(self.file).encode(), "text/plain")))
+            if isinstance(self.file, File):
+                files.append(("file", self.file.to_tuple()))
             else:
                 files.append(("file", (None, str(self.file).encode(), "text/plain")))
 
@@ -188,12 +191,18 @@ def _parse_name(data: object) -> None | str | Unset:
 
         append_suffix_if_duplicate = d.pop("append_suffix_if_duplicate", UNSET)
 
-        def _parse_file(data: object) -> None | str | Unset:
+        def _parse_file(data: object) -> File | None | Unset:
             if data is None:
                 return data
             if isinstance(data, Unset):
                 return data
-            return cast(None | str | Unset, data)
+            try:
+                if not isinstance(data, bytes):
+                    raise TypeError()
+                return File(payload=BytesIO(data))
+            except:  # noqa: E722
+                pass
+            return cast(File | None | Unset, data)
 
         file = _parse_file(d.pop("file", UNSET))