Python DSC Adapter and Test Resource implementation - #1520
Conversation
4c04457 to
6ccefc6
Compare
|
@microsoft-github-policy-service agree company="Microsoft" |
e3b036f to
bb93ae9
Compare
|
Copilot Review this |
There was a problem hiding this comment.
Pull request overview
Adds a new Python DSC adapter implementation with test fixtures and Pester/Python tests, intended to let DSC invoke Python resource classes through adapted resource manifests.
Changes:
- Adds Python adapter manifest and
pyDscAdapterCLI/runtime modules. - Adds Python test resources, manifests, pyproject mapping, and unit/component/integration tests.
- Updates ignore rules and one PowerShell adapter test stderr redirection.
Reviewed changes
Copilot reviewed 21 out of 27 changed files in this pull request and generated 46 comments.
Show a summary per file
| File | Description |
|---|---|
.gitignore |
Adds Visual Studio and Python cache ignores. |
adapters/__init__.py |
Adds package marker. |
adapters/powershell/Tests/powershellgroup.resource.tests.ps1 |
Redirects stderr for one adapted-resource test. |
adapters/python/.project.data.json |
Defines Python adapter build project metadata. |
adapters/python/__init__.py |
Adds package marker. |
adapters/python/pythonadapter.dsc.resource.json |
Adds DSC adapter manifest for Python operations. |
adapters/python/pyDscAdapter/__init__.py |
Adds adapter package marker. |
adapters/python/pyDscAdapter/__main__.py |
Adds Python adapter entry point. |
adapters/python/pyDscAdapter/adapter.py |
Implements operation routing, class loading, logging, and profiling. |
adapters/python/pyDscAdapter/cli.py |
Adds command-line parser and stdin/input handling. |
adapters/python/pyDscAdapter/discovery.py |
Adds pyproject class-map parsing and dynamic file import. |
adapters/python/pyDscAdapter/dsc_logging.py |
Adds structured DSC-style logging. |
adapters/python/pyDscAdapter/utils.py |
Adds JSON parsing helper. |
adapters/python/tests/.project.data.json |
Defines test project copy metadata. |
adapters/python/tests/TEST_LAYERS.md |
Documents unit/component/integration test layers. |
adapters/python/tests/__init__.py |
Adds test package marker. |
adapters/python/tests/pythoncomponent.tests.ps1 |
Adds adapter subprocess component tests. |
adapters/python/tests/pythonintegration.tests.ps1 |
Adds DSC CLI integration tests. |
adapters/python/tests/pythonunit.tests.py |
Adds Python unittest coverage for adapter internals. |
adapters/python/tests/pyproject.toml |
Maps PythonTest resource types to fixture classes. |
adapters/python/tests/pythontest.dsc.manifests.json |
Adds adapted resource manifests for Python test resources. |
adapters/python/tests/src/.project.data.json |
Defines fixture resource copy metadata. |
adapters/python/tests/src/__init__.py |
Adds fixture package marker. |
adapters/python/tests/src/export.py |
Adds export-only fixture resource. |
adapters/python/tests/src/get.py |
Adds get-only fixture resource. |
adapters/python/tests/src/set.py |
Adds set fixture resource. |
adapters/python/tests/src/test.py |
Adds test-only fixture resource. |
88dca5c to
cb13bdd
Compare
63a7654 to
7029eb3
Compare
There was a problem hiding this comment.
Pull request overview
Copilot reviewed 22 out of 28 changed files in this pull request and generated 7 comments.
Comments suppressed due to low confidence (1)
adapters/python/tests/pythonintegration.tests.ps1:25
- This comment claims the Python adapter manifest has a condition on python3 availability, but the manifest added in this PR does not include a
conditionfield. This is misleading for future maintainers and for interpreting failure expectations in this helper.
# Python adapter manifest now has a condition on python3 availability.
# If unavailable, resource discovery should fail cleanly rather than execution failing.
There was a problem hiding this comment.
Pull request overview
Copilot reviewed 22 out of 28 changed files in this pull request and generated no new comments.
Suppressed comments (4)
adapters/python/pyDscAdapter/adapter.py:164
- _instantiate_resource checks only for attribute existence with hasattr(cls, "from_json"), but doesn’t ensure it’s callable. If a resource defines a non-callable from_json attribute (e.g., a class variable/property), this will raise a TypeError and fail the operation unexpectedly. Prefer a callable() check on getattr().
if hasattr(cls, "from_json"):
return cls.from_json(json_input, operation=operation)
# Fallback: direct init from dict if needed
data = json.loads(json_input or "{}")
return cls(**data)
adapters/python/pyDscAdapter/adapter.py:197
- When resource introspection fails in _build_list_entry, capabilities is forced to ["get","set","test"]. This can advertise operations the resource does not implement, causing DSC to attempt unsupported operations later. It’s safer to return an empty capabilities list (or only those confirmed) when introspection fails.
if not capabilities:
capabilities = ["get", "set", "test"] # remove this
adapters/python/pyDscAdapter/adapter.py:243
- _discover_python_resources appends list entries even when it can’t find a corresponding .py file for a discovered resource type. This can produce list results with non-existent paths and causes avoidable import/introspection failures. Skip the resource when no file can be resolved.
resource_file = pyproject_path.parent / (discovered_type.split("/")[-1].lower() + ".py")
if not resource_file.exists():
resource_file = pyproject_path.parent / (class_name.lower() + ".py")
adapters/python/pyDscAdapter/discovery.py:152
- The non-tomllib fallback parser in get_class_map_from_pyproject only supports a flat [tool.dsc.resources] key/value form, but this repo’s pyproject.toml uses per-resource tables like [tool.dsc.resources."PythonTest/Get"]. On Python < 3.11 (no tomllib), discovery will return an empty class map and the adapter won’t resolve any resources.
class_map = {}
in_section = False
for line in content.splitlines():
stripped = line.strip()
if stripped == "[tool.dsc.resources]":
PR Summary
The Python adapter (Microsoft.DSC.Adapters/Python) enables DSC v3 to dynamically execute Python resource classes. It acts as a bridge between the DSC framework and Python-based resource implementations, handling all DSC operations (get, set, test, export, list, validate).
Key Features
Dynamic Resource Discovery: Reads resource mappings from pyproject.toml under [tool.dsc.resources]
Operation Routing: Routes DSC operations to appropriate Python resource class methods
Structured Logging: Full trace/debug/info/error logging with optional profiling capabilities
State Management: Handles JSON serialization and properly emits state + diffs per DSC specification
Error Handling: Graceful error handling with proper exit codes and stderr logging
Core Components
adapter.py: Main ResourceAdapter class managing operation routing, class loading, and resource discovery
discovery.py: Parses pyproject.toml and dynamically imports resource classes
cli.py: Command-line interface handling arguments and stdin input
dsc_logging.py: Environment variable-based logging configuration
Test Coverage
Three comprehensive test layers:
Unit Tests (Python/unittest) — Tests adapter internals: initialization, manifest discovery, class resolution, operation routing
Component Tests (Pester) — Validates adapter subprocess behavior directly: output shapes, exit codes, error handling
Integration Tests (Pester) — End-to-end DSC CLI integration testing through the full DSC framework
PR Context
DSC already supports adapters for non-native resource ecosystems, and this PR adds a Python adapter so Python-based resources can participate in the same DSC execution model as built-in and PowerShell resources.
This adapter is needed to: