♻️🔨 Is3517/refactor service io and diagnostics tool concept (ITISFoundation#3537)

Author: pcrespov

packages/models-library/Makefile

@@ -81,7 +81,7 @@ docker_rest_api.py: ## auto-generates pydantic models for Docker REST API models
 #  - NOTE: these tools require graphviz!
 #  - SEE https://erdantic.drivendata.org/stable/
 #
-
+.PHONY: _erdantic
 _erdantic: _check_venv_active
 	# ensures erdantic installed
 	@python3 -c "import erdantic" 2>/dev/null || pip install erdantic
@@ -101,3 +101,23 @@ erd-Node.svg: _erdantic
 erd-ServiceInput.svg: _erdantic
 	erdantic models_library.services.ServiceInput \
 		--out $@
+
+
+#
+# Test data
+#
+
+DOWNLOADED_TEST_DATA_DIR = "$(CURDIR)/tests/data/.downloaded-ignore"
+
+.PHONY: _httpx
+_httpx: _check_venv_active
+	# ensures requirements installed
+	@python3 -c "import httpx" 2>/dev/null || pip install httpx
+
+PHONY: pull_test_data
+pull_test_data: $(DOT_ENV_FILE) _httpx ## downloads tests data from registry (this can take some time!)
+	# downloading all metadata files
+	@set -o allexport; \
+	source $<; \
+	set +o allexport; \
+	python3 "$(PACKAGES_DIR)/pytest-simcore/src/pytest_simcore/helpers/utils_docker_registry.py" $(DOWNLOADED_TEST_DATA_DIR)

packages/models-library/setup.cfg

@@ -16,3 +16,6 @@ test = pytest
 
 [tool:pytest]
 asyncio_mode = auto
+
+markers =
+  diagnostics: "can be used to run diagnostics against deployed data (e.g. database, registry etc)"

packages/models-library/src/models_library/utils/services_io.py

@@ -1,3 +1,4 @@
+import mimetypes
 from copy import deepcopy
 from typing import Any, Literal, Optional, Union
 
@@ -14,6 +15,25 @@
 }
 
 
+def guess_media_type(io: Union[ServiceInput, ServiceOutput]) -> str:
+    # SEE https://docs.python.org/3/library/mimetypes.html
+    # SEE https://www.iana.org/assignments/media-types/media-types.xhtml
+    media_type = io.property_type.removeprefix("data:")
+    if media_type == "*/*" and io.file_to_key_map:
+        filename = list(io.file_to_key_map.keys())[0]
+        media_type, _ = mimetypes.guess_type(filename)
+        if media_type is None:
+            media_type = "*/*"
+    return media_type
+
+
+def update_schema_doc(schema: dict[str, Any], port: Union[ServiceInput, ServiceOutput]):
+    schema["title"] = port.label
+    if port.label != port.description:
+        schema["description"] = port.description
+    return schema
+
+
 def get_service_io_json_schema(
     port: Union[ServiceInput, ServiceOutput]
 ) -> Optional[dict[str, Any]]:
@@ -25,16 +45,15 @@ def get_service_io_json_schema(
     of BaseServiceIO once we proceed to a full deprecation of legacy fields like units, etc
     """
     if port.content_schema:
+        # NOTE this schema was already validated in BaseServiceIOModel
         return deepcopy(port.content_schema)
 
     # converts legacy
     if schema := _PROPERTY_TYPE_TO_SCHEMAS.get(port.property_type):
         schema = deepcopy(schema)
 
         # updates schema-doc, i.e description and title
-        schema["title"] = port.label
-        if port.label != port.description:
-            schema["description"] = port.description
+        update_schema_doc(schema=schema, port=port)
 
         # new x_unit custom field in json-schema
         if port.unit:

packages/models-library/tests/conftest.py

@@ -9,10 +9,10 @@
 import pytest
 
 pytest_plugins = [
-    "pytest_simcore.repository_paths",
-    "pytest_simcore.schemas",
     "pytest_simcore.pydantic_models",
     "pytest_simcore.pytest_global_environs",
+    "pytest_simcore.repository_paths",
+    "pytest_simcore.schemas",
 ]
 
 CURRENT_DIR = Path(sys.argv[0] if __name__ == "__main__" else __file__).resolve().parent

packages/models-library/tests/data/metadata-ascent-1.3.2.json

@@ -0,0 +1,139 @@
+{
+  "authors": [
+    {
+      "name": "Katie Zhuang",
+      "email": "[email protected]",
+      "affiliation": "IT'IS Foundation"
+    }
+  ],
+  "contact": "[email protected]",
+  "description": "ASCENT is an open source platform for simulating peripheral nerve stimulation. To download the software, visit the [ASCENT GitHub repository](https://github.com/wmglab-duke/ascent). This implementation uses a limited set of user-defined parameters to run the pipeline and will output the results of the NEURON simulations * Musselman ED, Cariello JE, Grill WM, Pelot NA. ASCENT (Automated Simulations to Characterize Electrical Nerve Thresholds): A Pipeline for Sample-Specific Computational Modeling of Electrical Stimulation of Peripheral Nerves. PLoS Comput Biol [Internet]. 2021; Available from: https://doi.org/10.1371/journal.pcbi.1009285 * Musselman ED, Cariello JE, Grill WM, Pelot NA. ASCENT (Automated Simulations to Characterize Electrical Nerve Thresholds): A Pipeline for Sample-Specific Computational Modeling of Electrical Stimulation of Peripheral Nerves. PLoS Comput Biol [Internet]. 2021, DOI: 10.5281/zenodo.5500260\n",
+  "inputs": {
+    "input_1": {
+      "displayOrder": 1,
+      "label": "Nerve Morphology",
+      "description": "Choice of Vagus nerve morphology - either Rat VN or Human VN",
+      "type": "ref_contentSchema",
+      "contentSchema": {
+        "title": "Nerve Morphology",
+        "default": "Rat",
+        "enum": [
+          "Rat",
+          "Human"
+        ]
+      }
+    },
+    "input_2": {
+      "displayOrder": 2,
+      "label": "Cuff geometry",
+      "description": "45\u00b0 contact wrap monopolar cuff or 360\u00b0 contact wrap monopolar cuff or 45\u00b0 contact wrap, bipolar cuff or 360\u00b0 contact wrap, bipolar cuff \n",
+      "type": "ref_contentSchema",
+      "contentSchema": {
+        "title": "Cuff Geometry",
+        "default": "45\u00b0 monopolar",
+        "enum": [
+          "45\u00b0 monopolar",
+          "360\u00b0 monopolar",
+          "45\u00b0 bipolar",
+          "360\u00b0 bipolar"
+        ]
+      }
+    },
+    "input_3": {
+      "displayOrder": 3,
+      "label": "Fiber Locations",
+      "description": "centroid: one fiber location at the centroid of each fascicle wheel: 6 spokes with 2 fibers per spoke, plus the centroid; 13 fiber locations per fascicle\n",
+      "type": "ref_contentSchema",
+      "contentSchema": {
+        "title": "Fiber Locations",
+        "default": "wheel",
+        "enum": [
+          "centroid",
+          "wheel"
+        ]
+      }
+    },
+    "fibers": {
+      "displayOrder": 4,
+      "label": "List of Fiber Diameters",
+      "description": "Comma-separated list of fiber diameters (between 2 and 10), defaults units are \u03bcm. At least one value must be provided, maximum 5 values",
+      "type": "ref_contentSchema",
+      "contentSchema": {
+        "title": "Fiber Diameters",
+        "type": "array",
+        "minItems": 1,
+        "maxItems": 5,
+        "default": [
+          10
+        ],
+        "x_unit": "micro-meter",
+        "items": {
+          "minimum": 2,
+          "maximum": 10,
+          "type": "number"
+        }
+      }
+    },
+    "input_4": {
+      "displayOrder": 5,
+      "label": "Waveform",
+      "description": "monophasic: pulse (cathodic for  monopolar  cuff;  cathodic phase  first  on  contact  closest to  recording  site  for bipolar cuff) biphasic: symmetric biphasic pulse\n",
+      "type": "ref_contentSchema",
+      "contentSchema": {
+        "title": "Waveform",
+        "default": "biphasic",
+        "enum": [
+          "monophasic",
+          "biphasic"
+        ]
+      }
+    },
+    "durations": {
+      "displayOrder": 6,
+      "label": "List of Waveform Durations",
+      "description": "Comma-separated list of pulse widths (between 0.05 and 2), default units are ms. At least one value must be provided, maximum 5 values.",
+      "type": "ref_contentSchema",
+      "contentSchema": {
+        "title": "Waveform Durations",
+        "type": "array",
+        "minItems": 1,
+        "maxItems": 5,
+        "default": [
+          0.5
+        ],
+        "x_unit": "milli-second",
+        "items": {
+          "minimum": 0.05,
+          "maximum": 2.0,
+          "type": "number"
+        }
+      }
+    }
+  },
+  "integration-version": "1.0.0",
+  "key": "simcore/services/comp/ascent-runner",
+  "name": "ascent-runner",
+  "outputs": {
+    "output_1": {
+      "displayOrder": 1,
+      "label": "Simulation Outputs",
+      "description": "Resulting thresholds from simulations.",
+      "type": "data:*/*",
+      "fileToKeyMap": {
+        "ascent_results.zip": "output_1"
+      }
+    },
+    "output_2": {
+      "displayOrder": 1,
+      "label": "Environment Variables",
+      "description": "Sample number, model number and simulation number for postprocessing.",
+      "type": "data:*/*",
+      "fileToKeyMap": {
+        "envs": "output_2"
+      }
+    }
+  },
+  "thumbnail": "https://wmglab-duke-ascent.readthedocs.io/en/latest/_images/ascent_media_release_v2.png",
+  "type": "computational",
+  "version": "1.3.2"
+}

packages/models-library/tests/data/image-meta.yaml renamed to packages/models-library/tests/data/metadata-sleeper-2.0.2.yaml

@@ -4,16 +4,16 @@ type: computational
 integration-version: 1.0.0
 version: 2.0.2
 description: A service which awaits for time to pass.
-contact: neagu@itis.swiss
+contact: neagu@test.it
 authors:
   - name: "Manuel Guidon"
-    email: guidon@itis.swiss
+    email: guidon@test.it
     affiliation: "IT'IS Foundation"
   - name: "Odei Maiz"
-    email: maiz@itis.swiss
+    email: maiz@test.it
     affiliation: "IT'IS Foundation"
   - name: "Andrei Neagu"
-    email: neagu@itis.swiss
+    email: neagu@test.it
     affiliation: "IT'IS Foundation"
 inputs:
   input_1:

packages/models-library/tests/test_services_io.py

@@ -12,7 +12,9 @@
 def test_service_port_units(project_tests_dir: Path):
     ureg = UnitRegistry()
 
-    data = yaml.safe_load((project_tests_dir / "data" / "image-meta.yaml").read_text())
+    data = yaml.safe_load(
+        (project_tests_dir / "data" / "metadata-sleeper-2.0.2.yaml").read_text()
+    )
     print(ServiceDockerData.schema_json(indent=2))
 
     service_meta = ServiceDockerData.parse_obj(data)

packages/models-library/tests/test_utils_service_io.py

@@ -4,12 +4,17 @@
 # pylint: disable=too-many-arguments
 
 
+import itertools
+import json
+import sys
+from pathlib import Path
 from typing import Union
 
 import pytest
-from models_library.services import ServiceInput, ServiceOutput
+from models_library.services import ServiceInput, ServiceOutput, ServicePortKey
 from models_library.utils.json_schema import jsonschema_validate_schema
 from models_library.utils.services_io import get_service_io_json_schema
+from pydantic import parse_obj_as
 
 example_inputs_labels = [
     e for e in ServiceInput.Config.schema_extra["examples"] if e["label"]
@@ -20,7 +25,7 @@
 
 
 @pytest.fixture(params=example_inputs_labels + example_outputs_labels)
-def service_io(request: pytest.FixtureRequest) -> Union[ServiceInput, ServiceOutput]:
+def service_port(request: pytest.FixtureRequest) -> Union[ServiceInput, ServiceOutput]:
     try:
         index = example_inputs_labels.index(request.param)
         example = ServiceInput.Config.schema_extra["examples"][index]
@@ -31,17 +36,49 @@ def service_io(request: pytest.FixtureRequest) -> Union[ServiceInput, ServiceOut
         return ServiceOutput.parse_obj(example)
 
 
-def test_it(service_io: Union[ServiceInput, ServiceOutput]):
-    print(service_io.json(indent=2))
+def test_get_schema_from_port(service_port: Union[ServiceInput, ServiceOutput]):
+    print(service_port.json(indent=2))
 
     # get
-    schema = get_service_io_json_schema(service_io)
+    schema = get_service_io_json_schema(service_port)
     print(schema)
 
-    if service_io.property_type.startswith("data"):
+    if service_port.property_type.startswith("data"):
         assert not schema
     else:
         assert schema
-
         # check valid jsons-schema
         jsonschema_validate_schema(schema)
+
+
+CURRENT_DIR = Path(sys.argv[0] if __name__ == "__main__" else __file__).resolve().parent
+TEST_DATA_FOLDER = CURRENT_DIR / "data"
+
+
+@pytest.mark.diagnostics
+@pytest.mark.parametrize(
+    "metadata_path",
+    TEST_DATA_FOLDER.rglob("metadata*.json"),
+    ids=lambda p: f"{p.parent.name}/{p.name}",
+)
+def test_against_service_metadata_configs(metadata_path: Path):
+    """
+    This tests can be used as well to validate all metadata in a given registry
+
+    SEE make pull_test_data to pull data from the registry specified in .env
+    """
+
+    meta = json.loads(metadata_path.read_text())
+
+    inputs = parse_obj_as(dict[ServicePortKey, ServiceInput], meta["inputs"])
+    outputs = parse_obj_as(dict[ServicePortKey, ServiceOutput], meta["outputs"])
+
+    for port in itertools.chain(inputs.values(), outputs.values()):
+        schema = get_service_io_json_schema(port)
+
+        if port.property_type.startswith("data"):
+            assert not schema
+        else:
+            assert schema
+            # check valid jsons-schema
+            jsonschema_validate_schema(schema)

packages/pytest-simcore/src/pytest_simcore/environment_configs.py

@@ -3,27 +3,26 @@
 # pylint: disable=unused-variable
 
 
-from copy import deepcopy
 from pathlib import Path
 
-import dotenv
 import pytest
 from pytest import MonkeyPatch
 
+from .helpers.typing_env import EnvVarsDict
+from .helpers.utils_envs import load_dotenv, setenvs_from_dict
+
 
 @pytest.fixture(scope="session")
-def env_devel_dict(env_devel_file: Path) -> dict[str, str]:
+def env_devel_dict(env_devel_file: Path) -> EnvVarsDict:
     assert env_devel_file.exists()
     assert env_devel_file.name == ".env-devel"
-    environ = dotenv.dotenv_values(env_devel_file, verbose=True, interpolate=True)
-    assert all(v is not None for v in environ.values())
-    return environ  # type: ignore
+    envs = load_dotenv(env_devel_file, verbose=True, interpolate=True)
+    return envs
 
 
 @pytest.fixture(scope="function")
 def mock_env_devel_environment(
     env_devel_dict: dict[str, str], monkeypatch: MonkeyPatch
-) -> dict[str, str]:
-    for key, value in env_devel_dict.items():
-        monkeypatch.setenv(key, str(value))
-    return deepcopy(env_devel_dict)
+) -> EnvVarsDict:
+    envs = setenvs_from_dict(monkeypatch, env_devel_dict)
+    return envs