pipeloom-databricks 0.5.0__tar.gz → 0.6.0__tar.gz

{pipeloom_databricks-0.5.0 → pipeloom_databricks-0.6.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: pipeloom-databricks
-Version: 0.5.0
+Version: 0.6.0
 Summary: Databricks renderer for pipeloom: a compiled plan as Jobs settings JSON and bundle YAML.
 Project-URL: Repository, https://github.com/emanuel-luis/pipeloom
 Project-URL: Documentation, https://emanuel-luis.github.io/pipeloom/
@@ -21,8 +21,8 @@ Classifier: Programming Language :: Python :: 3.14
 Classifier: Typing :: Typed
 Requires-Python: >=3.11
 Requires-Dist: databricks-sdk<1,>=0.50
-Requires-Dist: pipeloom-spark~=0.5.0
-Requires-Dist: pipeloom~=0.5.0
+Requires-Dist: pipeloom-spark~=0.6.0
+Requires-Dist: pipeloom~=0.6.0
 Description-Content-Type: text/markdown
 
 # pipeloom-databricks

{pipeloom_databricks-0.5.0 → pipeloom_databricks-0.6.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "pipeloom-databricks"
-version = "0.5.0"
+version = "0.6.0"
 description = "Databricks renderer for pipeloom: a compiled plan as Jobs settings JSON and bundle YAML."
 readme = "README.md"
 requires-python = ">=3.11"
@@ -23,7 +23,7 @@ classifiers = [
     "Programming Language :: Python :: 3.14",
     "Typing :: Typed",
 ]
-dependencies = ["pipeloom~=0.5.0", "pipeloom-spark~=0.5.0", "databricks-sdk>=0.50,<1"]
+dependencies = ["pipeloom~=0.6.0", "pipeloom-spark~=0.6.0", "databricks-sdk>=0.50,<1"]
 
 [project.urls]
 Repository = "https://github.com/emanuel-luis/pipeloom"

pipeloom_databricks-0.6.0/src/pipeloom_databricks/api.py

@@ -0,0 +1,77 @@
+"""The stable surface other pipeloom distributions (such as an Airflow renderer) may call.
+
+Importing it loads no SDK: `connect` and `connect_explicit` import `databricks.sdk` only when they
+run.
+"""
+
+from __future__ import annotations
+
+from typing import Literal
+
+from pipeloom_databricks.bundle import Runtime
+from pipeloom_databricks.client import (
+    ApiError,
+    ConnectionNotConfigured,
+    DatabricksClient,
+    JobSummary,
+    RunInfo,
+    TaskRunInfo,
+    connect,
+)
+from pipeloom_databricks.deploy import DeployResult, DuplicateJobs, api_errors, deploy, find_job
+from pipeloom_databricks.finalize_states import task_states
+from pipeloom_databricks.grouping import task_key_map
+from pipeloom_databricks.keys import FINALIZE_KEY
+from pipeloom_databricks.render import deploy_options, job_name, render_job
+from pipeloom_databricks.repair import repair, rerun_set
+from pipeloom_databricks.run import TERMINAL, RunStarted, SelectionError, find_run, run_job
+
+__all__ = [
+    "FINALIZE_KEY",
+    "TERMINAL",
+    "ApiError",
+    "ConnectionNotConfigured",
+    "DatabricksClient",
+    "DeployResult",
+    "DuplicateJobs",
+    "JobSummary",
+    "RunInfo",
+    "RunStarted",
+    "Runtime",
+    "SelectionError",
+    "TaskRunInfo",
+    "api_errors",
+    "connect",
+    "connect_explicit",
+    "deploy",
+    "deploy_options",
+    "find_job",
+    "find_run",
+    "job_name",
+    "render_job",
+    "repair",
+    "rerun_set",
+    "run_job",
+    "task_key_map",
+    "task_states",
+]
+
+
+def connect_explicit(
+    *,
+    host: str,
+    auth_type: Literal["pat", "oauth-m2m"],
+    token: str | None = None,
+    client_id: str | None = None,
+    client_secret: str | None = None,
+) -> DatabricksClient:
+    """A client from explicit credentials only; see `client.connect.connect_explicit`."""
+    from pipeloom_databricks.client.connect import connect_explicit as _connect
+
+    return _connect(
+        host=host,
+        auth_type=auth_type,
+        token=token,
+        client_id=client_id,
+        client_secret=client_secret,
+    )

{pipeloom_databricks-0.5.0 → pipeloom_databricks-0.6.0}/src/pipeloom_databricks/client/connect.py

@@ -13,7 +13,7 @@ import importlib
 import tempfile
 from collections.abc import Mapping
 from pathlib import Path
-from typing import Any, Final
+from typing import Any, Final, Literal
 
 from pipeloom.secrets.redaction import Redactor
 from pipeloom_databricks.client import ConnectionNotConfigured, DatabricksClient
@@ -61,11 +61,36 @@ def connect(
             "DATABRICKS_HOST with credentials, or run on a Databricks runtime"
         )
 
+    return _build(kwargs, with_file=bool(profile))
+
+
+def connect_explicit(
+    *,
+    host: str,
+    auth_type: Literal["pat", "oauth-m2m"],
+    token: str | None = None,
+    client_id: str | None = None,
+    client_secret: str | None = None,
+) -> DatabricksClient:
+    """A client from explicit credentials only (an orchestrator's connection, ADR-0039).
+
+    The SDK gets the auth type and a `config_file` that does not exist, so neither
+    `~/.databrickscfg` nor a `DATABRICKS_*` variable of the process can change the auth.
+    """
+    kwargs: dict[str, Any] = {"host": host, "auth_type": auth_type}
+    if auth_type == "pat":
+        kwargs["token"] = token
+    else:
+        kwargs.update(client_id=client_id, client_secret=client_secret)
+    return _build(kwargs, with_file=False)
+
+
+def _build(kwargs: dict[str, Any], *, with_file: bool) -> DatabricksClient:
     from databricks.sdk import WorkspaceClient
     from databricks.sdk.core import Config
 
     try:
-        if profile:
+        if with_file:
             config = Config(**kwargs)
         else:
             with tempfile.TemporaryDirectory(prefix="pipeloom-") as folder:

{pipeloom_databricks-0.5.0 → pipeloom_databricks-0.6.0}/tests/test_connect.py

@@ -154,3 +154,28 @@ def test_factory_skips_every_rule(sdk: Recorder) -> None:
 def test_a_bad_factory_is_pld010(spec: str) -> None:
     with pytest.raises(ConnectionNotConfigured):
         connect(profile=None, env={}, factory=spec)
+
+
+def test_explicit_credentials_ignore_profiles_and_variables(
+    sdk: Recorder, monkeypatch: pytest.MonkeyPatch
+) -> None:
+    from pipeloom_databricks.api import connect_explicit
+
+    monkeypatch.setenv("DATABRICKS_CONFIG_PROFILE", "someone-elses")
+    monkeypatch.setenv("DATABRICKS_TOKEN", "env-token")
+    client = connect_explicit(host="https://h.example", auth_type="pat", token="t")
+    connect_explicit(
+        host="https://h.example", auth_type="oauth-m2m", client_id="id", client_secret="s"
+    )
+    first, second = sdk.configs
+    for kwargs in (first, second):
+        cfg = Path(kwargs.pop("config_file"))
+        assert not cfg.exists() and not cfg.parent.exists()
+    assert first == {"host": "https://h.example", "auth_type": "pat", "token": "t"}
+    assert second == {
+        "host": "https://h.example",
+        "auth_type": "oauth-m2m",
+        "client_id": "id",
+        "client_secret": "s",
+    }
+    assert isinstance(client, SdkClient)

pipeloom_databricks-0.5.0/src/pipeloom_databricks/api.py

@@ -1,45 +0,0 @@
-"""The stable surface other pipeloom distributions (such as an Airflow renderer) may call.
-
-Importing it loads no SDK: `connect` imports `databricks.sdk` only when it runs.
-"""
-
-from __future__ import annotations
-
-from pipeloom_databricks.client import (
-    ApiError,
-    ConnectionNotConfigured,
-    DatabricksClient,
-    JobSummary,
-    RunInfo,
-    TaskRunInfo,
-    connect,
-)
-from pipeloom_databricks.deploy import DeployResult, DuplicateJobs, deploy, find_job
-from pipeloom_databricks.finalize_states import task_states
-from pipeloom_databricks.grouping import task_key_map
-from pipeloom_databricks.render import deploy_options, render_job
-from pipeloom_databricks.repair import repair
-from pipeloom_databricks.run import RunStarted, SelectionError, find_run, run_job
-
-__all__ = [
-    "ApiError",
-    "ConnectionNotConfigured",
-    "DatabricksClient",
-    "DeployResult",
-    "DuplicateJobs",
-    "JobSummary",
-    "RunInfo",
-    "RunStarted",
-    "SelectionError",
-    "TaskRunInfo",
-    "connect",
-    "deploy",
-    "deploy_options",
-    "find_job",
-    "find_run",
-    "render_job",
-    "repair",
-    "run_job",
-    "task_key_map",
-    "task_states",
-]