Release databricks-sdk-py v0.123 (#1507)

Release databricks-sdk-py v0.123.0

## Summary

<!--
One or two sentences describing what this PR changes and what it
enables.
  Focus on the effect, not the implementation details.

  Example:
    Extracts the credentials chain iteration logic into a reusable
    `NewCredentialsChain` constructor so that internal tools can compose
their own authentication chains from individual credential strategies.
-->

## Why

<!--
Explain the problem that motivated this change. A reviewer who reads
only
this section should understand why the PR exists and what problem it
solves.

- Start with the status quo: how things work today and what limitation
exists.
- Explain who is affected and what they cannot do (or must work around).
  - If alternatives were considered and rejected, briefly mention why.
  - End with how this PR addresses the gap.

  The "why" is the most important part of a PR description — it usually
  cannot be inferred from the code itself.
-->

## What changed

### Interface changes

<!--
New or modified public API surface: types, functions, configuration
options.
Use backticks for code references. Write "None." if there are no
changes.

  Example:
- **`NewCredentialsChain(...CredentialsStrategy) CredentialsStrategy`**
—
Takes an ordered list of credential strategies and returns a strategy
      that tries them in sequence.
-->

### Behavioral changes

<!--
User-visible behavior changes: different defaults, changed error
messages,
new side effects, performance characteristics. Write "None." if this is
a
  pure refactor — this explicitly reassures reviewers.
-->

### Internal changes

<!--
  Refactoring, file moves, implementation details, test infrastructure.
  Things that don't affect the public API or user-visible behavior.
-->

## How is this tested?

<!--
Describe any tests you have done, especially tests that are not part of
the unit tests (e.g. local tests, integration tests, manual
verification).

ALWAYS ANSWER THIS QUESTION: answer with "N/A" if tests are not
applicable
to your PR (e.g. if the PR only modifies comments). Do not be afraid of
answering "Not tested" if the PR has not been tested. Being clear about
what
has been done and not done provides important context to the reviewers.
-->

Author: Divyansh-db

.codegen.json

@@ -0,0 +1,46 @@
+{
+  "mode": "py_v0",
+  "api_changelog": false,
+  "version": {
+    "databricks/sdk/version.py": "__version__ = \"$VERSION\""
+  },
+  "sync": {
+    "skip_from_genkit": [
+      ".github/workflows/next-changelog.yml",
+      "tagging.py",
+      "tagging.py.lock"
+    ],
+    "skip_from_source": [
+      "BUILD.bazel",
+      "CLAUDE.md",
+      "OWNERS",
+      "README.md",
+      "changelog.json",
+      "releases.jsonl",
+      ".cursor",
+      ".cursorrules",
+      ".nextchanges",
+      "changelog_validate",
+      "next-changelog.yml",
+      "tagging.py"
+    ],
+    "rename_on_sync": {
+      "README.public.md": "README.md"
+    }
+  },
+  "toolchain": {
+    "required": ["python3.12", "uv"],
+    "pre_setup": [
+      "python3.12 -m venv .databricks"
+    ],
+    "prepend_path": ".databricks/bin",
+    "setup": [
+      "pip install '.[dev]'"
+    ],
+    "post_generate": [
+      "make fmt",
+      "pip install .",
+      "python3.12 docs/gen-client-docs.py"
+    ]
+  }
+}

.codegen/_last_sha

@@ -1 +1 @@
-9ce12048f2c4a4fefb93fc3f55cba9c9798f9cb6
+6a5609075cc2c497a0a78f7d6ebd599d903b415d

.codegen/_openapi_sha

@@ -1 +1 @@
-9ce12048f2c4a4fefb93fc3f55cba9c9798f9cb6
+bb6433a6fab9f514ae007ef4cb93fb839a83108d

.gitattributes

@@ -1,13 +1,10 @@
 databricks/sdk/__init__.py linguist-generated=true
 databricks/sdk/errors/overrides.py linguist-generated=true
 databricks/sdk/errors/platform.py linguist-generated=true
-databricks/sdk/service/__init__.py linguist-generated=true
-databricks/sdk/service/_internal.py linguist-generated=true
 databricks/sdk/service/agentbricks.py linguist-generated=true
 databricks/sdk/service/aisearch.py linguist-generated=true
 databricks/sdk/service/apps.py linguist-generated=true
 databricks/sdk/service/billing.py linguist-generated=true
-databricks/sdk/service/bundle.py linguist-generated=true
 databricks/sdk/service/bundledeployments.py linguist-generated=true
 databricks/sdk/service/catalog.py linguist-generated=true
 databricks/sdk/service/cleanrooms.py linguist-generated=true

.github/workflows/tagging.yml

@@ -35,15 +35,6 @@ jobs:
       github.repository == 'databricks/databricks-sdk-go' ||
       github.repository == 'databricks/databricks-sdk-py' ||
       github.repository == 'databricks/databricks-sdk-java'
-    env:
-      # Ref the release is cut from. Releasing from a branch is currently
-      # supported in the terraform repository only: for
-      # terraform-provider-databricks this follows the dispatched branch
-      # (``github.ref_name``), so a release can be cut from a branch other
-      # than main; every other synced repo (SDK go/py/java, …) stays pinned
-      # to main, preserving the pre-existing hardcoded-main behaviour and
-      # blocking branch releases even via a manual non-main dispatch.
-      RELEASE_REF: ${{ github.repository == 'databricks/terraform-provider-databricks' && github.ref_name || 'main' }}
     environment: "release-is"
     runs-on:
       group: databricks-protected-runner-group
@@ -59,19 +50,12 @@ jobs:
       - name: Checkout repository
         uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
-          # Check out the release ref. ``RELEASE_REF`` is the dispatched
-          # branch for terraform-provider-databricks and ``main`` for every
-          # other repo (see the job-level ``env``). For scheduled runs and
-          # manual dispatches left at defaults this is the default branch
-          # (main); dispatching terraform with ``--ref <branch>`` cuts the
-          # release from that branch instead.
-          #
-          # Naming the ref explicitly (rather than omitting ``ref:``)
-          # forces re-resolution of the branch head at step time. Without
-          # it, checkout pins to ``github.sha`` — the SHA frozen at
-          # workflow_dispatch time — so re-running a stale dispatch would
-          # check out an older commit even when newer ones exist.
-          ref: ${{ env.RELEASE_REF }}
+          # Force re-resolution of ``main`` at step time. Without
+          # ``ref:``, checkout pins to ``github.sha`` — the SHA frozen
+          # at workflow_dispatch time — which means re-running a stale
+          # dispatch checks out an older main even when newer commits
+          # exist.
+          ref: main
           fetch-depth: 0
           token: ${{ steps.generate-token.outputs.token }}
 
@@ -89,12 +73,6 @@ jobs:
           GITHUB_TOKEN: ${{ steps.generate-token.outputs.token }}
           GITHUB_REPOSITORY: ${{ github.repository }}
           PACKAGES: ${{ inputs.packages }}
-          # Branch the release is cut from. tagging.py commits the
-          # changelog bump and creates the tag on this branch, and its
-          # concurrent-advance guard checks this branch. Matches the ref
-          # checked out above (``RELEASE_REF``): the dispatched branch for
-          # terraform, main everywhere else.
-          DECO_TAGGING_REF: ${{ env.RELEASE_REF }}
         run: |
           if [ -n "$PACKAGES" ]; then
             uv run --locked tagging.py --package "$PACKAGES"

CHANGELOG.md

@@ -1,5 +1,36 @@
 # Version changelog
 
+## Release v0.123.0 (2026-07-30)
+
+### API Changes
+* Add [w.ai_gateway](https://databricks-sdk-py.readthedocs.io/en/latest/workspace/catalog/ai_gateway.html) workspace-level service.
+* Add `update_operation()` method for [w.bundle_deployments](https://databricks-sdk-py.readthedocs.io/en/latest/workspace/bundledeployments/bundle_deployments.html) workspace-level service.
+* Add `patch_telemetry_config()` method for [w.serving_endpoints](https://databricks-sdk-py.readthedocs.io/en/latest/workspace/serving/serving_endpoints.html) workspace-level service.
+* Add `last_successful_version_id` and `updated_by` fields for `databricks.sdk.service.bundledeployments.Deployment`.
+* Add `dashboard_metadata`, `sequence_id` and `update_time` fields for `databricks.sdk.service.bundledeployments.Operation`.
+* Add `dashboard_metadata` field for `databricks.sdk.service.bundledeployments.Resource`.
+* Add `previous_version_id` field for `databricks.sdk.service.bundledeployments.Version`.
+* Add `dependency_mode` field for `databricks.sdk.service.compute.ClusterAttributes`.
+* Add `dependency_mode` field for `databricks.sdk.service.compute.ClusterDetails`.
+* Add `dependency_mode` field for `databricks.sdk.service.compute.ClusterSpec`.
+* Add `dependency_mode` field for `databricks.sdk.service.compute.CreateCluster`.
+* Add `dependency_mode` field for `databricks.sdk.service.compute.EditCluster`.
+* Add `dependency_mode` field for `databricks.sdk.service.compute.EnforcePolicyComplianceForClusterResponseClusterSettings`.
+* Add `dependency_mode` field for `databricks.sdk.service.compute.UpdateClusterResource`.
+* Add `effective_entitlements` field for `databricks.sdk.service.iamv2.WorkspaceAssignmentDetail`.
+* Add `code_source_path` field for `databricks.sdk.service.jobs.AiRuntimeTask`.
+* Add `serverless_compute_id` field for `databricks.sdk.service.jobs.JobCluster`.
+* Add `pause_status` field for `databricks.sdk.service.jobs.TriggerStateProto`.
+* Add `schema_registry_config` field for `databricks.sdk.service.ml.StreamSchemaConfig`.
+* Add `fanout_options` field for `databricks.sdk.service.pipelines.SchemaSpec`.
+* Add `extra_columns` field for `databricks.sdk.service.postgres.SyncedTableSyncedTableSpec`.
+* Add `parameters` field for `databricks.sdk.service.sql.AlertV2`.
+* Add `directory_info` field for `databricks.sdk.service.workspace.ObjectInfo`.
+* Add `special_destination_schema_owner`, `special_destination_table_owner`, `special_destination_volume_owner`, `special_destination_function_owner` and `special_destination_registered_model_owner` enum values for `databricks.sdk.service.catalog.SpecialDestination`.
+* [Breaking] Change `create_deployment()` method for [w.bundle_deployments](https://databricks-sdk-py.readthedocs.io/en/latest/workspace/bundledeployments/bundle_deployments.html) workspace-level service with new required argument order.
+* [Breaking] Remove `deployment_id` field for `databricks.sdk.service.bundledeployments.CreateDeploymentRequest`.
+* [Breaking] Remove `lifetime` field for `databricks.sdk.service.ml.TimeWindow`.
+
 ## Release v0.122.0 (2026-07-21)
 
 ### API Changes

databricks/sdk/__init__.py

@@ -1,8 +1,10 @@
 import io
+import json
 import logging
 import urllib.parse
 from abc import ABC, abstractmethod
-from datetime import timedelta
+from datetime import date, datetime, timedelta
+from enum import Enum
 from types import TracebackType
 from typing import Any, BinaryIO, Callable, Dict, Iterable, Iterator, List, Optional, Type, Union
 
@@ -19,6 +21,27 @@
 logger = logging.getLogger("databricks.sdk")
 
 
+def _json_default(o: Any) -> Any:
+    """``json.dumps`` ``default`` hook coercing SDK values into their wire form.
+
+    Lets an open ``Any`` body field accept an SDK response passed straight into
+    the next call (function chaining), which is otherwise not JSON serializable.
+    """
+    as_dict = getattr(o, "as_dict", None)
+    if callable(as_dict):
+        return as_dict()
+    to_json_string = getattr(o, "ToJsonString", None)  # protobuf Timestamp/Duration, FieldMask
+    if callable(to_json_string):
+        return to_json_string()
+    if isinstance(o, (datetime, date)):
+        return o.isoformat()
+    if isinstance(o, timedelta):
+        return f"{o.total_seconds()}s"
+    if isinstance(o, Enum):
+        return o.value
+    raise TypeError(f"Object of type {type(o).__name__} is not JSON serializable")
+
+
 def _fix_host_if_needed(host: Optional[str]) -> Optional[str]:
     if not host:
         return host
@@ -281,14 +304,18 @@ def _perform(
         data=None,
         auth: Callable[[requests.PreparedRequest], requests.PreparedRequest] = None,
     ):
+        json_data = None
+        if body is not None and data is None and files is None:
+            json_data = json.dumps(body, allow_nan=False, default=_json_default)
+            headers = {"Content-Type": "application/json", **(headers or {})}
         response = self._session.request(
             method,
             url,
             params=self._fix_query_string(query),
-            json=body,
+            json=None if json_data is not None else body,
             headers=headers,
             files=files,
-            data=data,
+            data=json_data if json_data is not None else data,
             auth=auth,
             stream=raw,
             timeout=self._http_timeout_seconds,