feat: remove v from ontology_info.json (#196)

## Reason for Change Census currently uses x.x.x for [encoding the CxG dataset schema version](https://github.com/chanzuckerberg/cellxgene-census/blob/main/docs/cellxgene_census_schema.md#census-metadata--census_objcensus_infosummary--somadataframe), e.g. 5.0.0 whereas the service uses vx.x.x e.g. v5.0.0 I think we should probably update the ontology service to adopt the Census semantics since there are plenty of Census data releases out there (which are immutable). ## Changes - update the ontology_info.json to not require the leading V in the cxg_schema version. - Use get_latest_schema_version from the API in all_ontology_generator.py. - updated tests and notebooks. ## Testing steps - All unit tests pass - run all_ontology_generate.py locally using the update API and it works. ## Notes for Reviewer *Updates to Ontology Files / ontology-processing (push) * will fail until we release the API. It also doesn't need to run for this PR.
chanzuckerberg · Apr 11, 2024 · a1292a3 · a1292a3
1 parent 96c7213
commit a1292a3
Show file tree

Hide file tree

Showing 8 changed files with 54 additions and 75 deletions.
diff --git a/api/python/notebooks/Python API.ipynb b/api/python/notebooks/Python API.ipynb
@@ -92,7 +92,7 @@
    "outputs": [],
    "source": [
     "# Init a CXGSchema object to track what ontology versions are supported\n",
-    "cxg_schema = CXGSchema(version=\"v5.0.0\")"
+    "cxg_schema = CXGSchema(version=\"5.0.0\")"
    ]
   },
   {
@@ -104,10 +104,7 @@
      "end_time": "2024-04-11T18:01:55.813250Z",
      "start_time": "2024-04-11T18:01:55.808657Z"
     },
-    "collapsed": false,
-    "jupyter": {
-     "outputs_hidden": false
-    }
+    "collapsed": false
    },
    "outputs": [
     {
@@ -135,10 +132,7 @@
      "end_time": "2024-04-11T18:02:03.459302Z",
      "start_time": "2024-04-11T18:02:03.454016Z"
     },
-    "collapsed": false,
-    "jupyter": {
-     "outputs_hidden": false
-    }
+    "collapsed": false
    },
    "outputs": [
     {
@@ -165,10 +159,7 @@
      "end_time": "2024-04-11T18:02:04.680816Z",
      "start_time": "2024-04-11T18:02:04.677350Z"
     },
-    "collapsed": false,
-    "jupyter": {
-     "outputs_hidden": false
-    }
+    "collapsed": false
    },
    "outputs": [],
    "source": [
@@ -567,10 +558,7 @@
      "end_time": "2024-04-11T18:29:24.902116Z",
      "start_time": "2024-04-11T18:29:24.897976Z"
     },
-    "collapsed": false,
-    "jupyter": {
-     "outputs_hidden": false
-    }
+    "collapsed": false
    },
    "outputs": [
     {
@@ -597,10 +585,7 @@
      "end_time": "2024-04-11T18:29:49.816735Z",
      "start_time": "2024-04-11T18:29:49.812574Z"
     },
-    "collapsed": false,
-    "jupyter": {
-     "outputs_hidden": false
-    }
+    "collapsed": false
    },
    "outputs": [
     {
@@ -627,10 +612,7 @@
      "end_time": "2024-04-11T18:34:30.054568Z",
      "start_time": "2024-04-11T18:34:29.903171Z"
     },
-    "collapsed": false,
-    "jupyter": {
-     "outputs_hidden": false
-    }
+    "collapsed": false
    },
    "outputs": [
     {
@@ -970,10 +952,7 @@
    "execution_count": null,
    "id": "6056479e1dc03237",
    "metadata": {
-    "collapsed": false,
-    "jupyter": {
-     "outputs_hidden": false
-    }
+    "collapsed": false
    },
    "outputs": [],
    "source": []

diff --git a/api/python/src/cellxgene_ontology_guide/supported_versions.py b/api/python/src/cellxgene_ontology_guide/supported_versions.py
@@ -28,13 +28,20 @@ def get_latest_schema_version(versions: List[str]) -> str:
     """Given a list of schema versions, return the latest version.
 
     :param versions: List[str] list of schema versions. Versions can be in the format "v5.0.0" or "5.0.0"
-    :return: str latest version with a "v" prefix
+    :return: str latest version without the leading "v"
     """
 
-    def _coerce(v: str) -> Version:
-        return Version.coerce(v[1:]) if v[0] == "v" else Version.coerce(v)
+    return str(sorted([coerce_version(version) for version in versions])[-1])
 
-    return "v" + str(sorted([_coerce(version) for version in versions])[-1])
+
+def coerce_version(version: str) -> Version:
+    """Coerce a version string into a semantic_version.Version object.
+
+    :param version: str version string to coerce
+    :return: Version coerced version object
+    """
+    v = version[1:] if version[0] == "v" else version
+    return Version.coerce(v)
 
 
 def load_supported_versions() -> Any:
@@ -60,18 +67,20 @@ def __init__(self, version: Optional[str] = None):
         """
         ontology_info = load_supported_versions()
         if version is None:
-            version = get_latest_schema_version(ontology_info.keys())
-        elif version not in ontology_info:
-            raise ValueError(f"Schema version {version} is not supported in this package version.")
-
-        self.version = version
-        self.supported_ontologies = ontology_info[version]["ontologies"]
+            _version = get_latest_schema_version(ontology_info.keys())
+        else:
+            _version = str(coerce_version(version))
+            if str(_version) not in ontology_info:
+                raise ValueError(f"Schema version {_version} is not supported in this package version.")
+
+        self.version = _version
+        self.supported_ontologies = ontology_info[_version]["ontologies"]
         self.ontology_file_names: Dict[str, str] = {}
-        self.deprecated_on = ontology_info[version].get("deprecated_on")
+        self.deprecated_on = ontology_info[_version].get("deprecated_on")
         if self.deprecated_on:
             parsed_date = datetime.strptime(self.deprecated_on, "%Y-%m-%d")
             warnings.warn(
-                f"Schema version {version} is deprecated as of {parsed_date}. It will be removed in a future version.",
+                f"Schema version {_version} is deprecated as of {parsed_date}. It will be removed in a future version.",
                 DeprecationWarning,
                 stacklevel=1,
             )

diff --git a/api/python/tests/test_ontology_parser.py b/api/python/tests/test_ontology_parser.py
@@ -43,9 +43,7 @@ def ontology_dict():
 @pytest.fixture
 def mock_CXGSchema(ontology_dict, mock_load_supported_versions, mock_load_ontology_file):
     mock_load_supported_versions.return_value = {
-        "v5.0.0": {
-            "ontologies": {"CL": {"version": "2024-01-01", "source": "http://example.com", "filename": "cl.owl"}}
-        }
+        "5.0.0": {"ontologies": {"CL": {"version": "2024-01-01", "source": "http://example.com", "filename": "cl.owl"}}}
     }
     cxg_schema = CXGSchema()
     cxg_schema.ontology_file_names = {"CL": "CL-ontology-2024-01-01.json.gz"}

diff --git a/api/python/tests/test_supported_versions.py b/api/python/tests/test_supported_versions.py
@@ -17,7 +17,7 @@
 @pytest.fixture
 def initialized_CXGSchemaInfo(mock_load_supported_versions):
     mock_load_supported_versions.return_value = {
-        "v5.0.0": {
+        "5.0.0": {
             "ontologies": {"CL": {"version": "v2024-01-01", "source": "http://example.com", "filename": "cl.owl"}}
         }
     }
@@ -26,7 +26,7 @@ def initialized_CXGSchemaInfo(mock_load_supported_versions):
 
 @pytest.mark.parametrize("versions, expected", [(["v5.0.0", "v0.0.1"], "v5.0.0"), (["5.0.0", "0.0.1"], "v5.0.0")])
 def test__get_latest_schema_version__OK(versions, expected):
-    assert get_latest_schema_version(versions) == "v5.0.0"
+    assert get_latest_schema_version(versions) == "5.0.0"
 
 
 @pytest.fixture
@@ -71,36 +71,42 @@ def test__load_supported_versions__OK(tmpdir):
         assert load_supported_versions() == file_contents
 
 
+@pytest.mark.parametrize("version, expected", [("v5.0.0", "5.0.0"), ("5.0.0", "5.0.0")])
+def test_coerce_version(version, expected):
+    assert get_latest_schema_version([version]) == expected
+
+
 class TestCXGSchema:
     def test__init__defaults(self, mock_load_supported_versions):
-        support_versions = {"v5.0.0": {"ontologies": {}}, "v0.0.1": {"ontologies": {}}}
+        support_versions = {"5.0.0": {"ontologies": {}}, "0.0.1": {"ontologies": {}}}
         mock_load_supported_versions.return_value = support_versions
         cxgs = CXGSchema()
-        assert cxgs.version == "v5.0.0"
-        assert cxgs.supported_ontologies == support_versions["v5.0.0"]["ontologies"]
+        assert cxgs.version == "5.0.0"
+        assert cxgs.supported_ontologies == support_versions["5.0.0"]["ontologies"]
 
-    def test__init__specific_version(self, mock_load_supported_versions):
-        support_versions = {"v5.0.0": {"ontologies": {}}, "v0.0.1": {"ontologies": {}}}
+    @pytest.mark.parametrize("version", ["v0.0.1", "0.0.1"])
+    def test__init__specific_version(self, version, mock_load_supported_versions):
+        support_versions = {"5.0.0": {"ontologies": {}}, "0.0.1": {"ontologies": {}}}
         mock_load_supported_versions.return_value = support_versions
-        cxgs = CXGSchema(version="v0.0.1")
-        assert cxgs.version == "v0.0.1"
-        assert cxgs.supported_ontologies == support_versions["v0.0.1"]["ontologies"]
+        cxgs = CXGSchema(version=version)
+        assert cxgs.version == "0.0.1"
+        assert cxgs.supported_ontologies == support_versions["0.0.1"]["ontologies"]
 
     def test__init__deprecated_version(self, mock_load_supported_versions):
-        support_versions = {"v5.0.0": {"ontologies": {}}, "v0.0.1": {"ontologies": {}, "deprecated_on": "2024-01-01"}}
+        support_versions = {"5.0.0": {"ontologies": {}}, "0.0.1": {"ontologies": {}, "deprecated_on": "2024-01-01"}}
         mock_load_supported_versions.return_value = support_versions
         # catch the deprecation warning
         with pytest.warns(DeprecationWarning) as record:
-            CXGSchema(version="v0.0.1")
+            CXGSchema(version="0.0.1")
         warning = record.pop()
         assert warning.message.args[0] == (
-            "Schema version v0.0.1 is deprecated as of 2024-01-01 00:00:00. It will be removed in a " "future version."
+            "Schema version 0.0.1 is deprecated as of 2024-01-01 00:00:00. It will be removed in a " "future version."
         )
 
     def test__init__unsupported_version(self, mock_load_supported_versions):
         mock_load_supported_versions.return_value = {}
         with pytest.raises(ValueError):
-            CXGSchema(version="v5.0.1")
+            CXGSchema(version="5.0.1")
 
     def test__ontology__unsupported_ontology_by_package(self, initialized_CXGSchemaInfo, mock_load_ontology_file):
         with pytest.raises(ValueError):

diff --git a/asset-schemas/ontology_info_schema.json b/asset-schemas/ontology_info_schema.json
@@ -4,7 +4,7 @@
   "description": "A schema for the set of valid ontology reference files mapping to a CZ CellXGene Dataset Schema Versions",
   "type": "object",
   "patternProperties": {
-    "^v[0-9]+\\.[0-9]+\\.[0-9]+$": {
+    "^[0-9]+\\.[0-9]+\\.[0-9]+$": {
       "description": "The version of CellxGene schema that maps to this set of ontology versions",
       "type": "object",
       "properties": {

diff --git a/ontology-assets/ontology_info.json b/ontology-assets/ontology_info.json
@@ -1,5 +1,5 @@
 {
-  "v5.0.0": {
+  "5.0.0": {
     "ontologies": {
       "CL": {
         "version": "v2024-01-04",

diff --git a/tools/ontology-builder/src/all_ontology_generator.py b/tools/ontology-builder/src/all_ontology_generator.py
@@ -12,14 +12,10 @@
 
 import env
 import owlready2
-import semantic_version
+from cellxgene_ontology_guide.supported_versions import get_latest_schema_version
 from validate_json_schemas import register_schemas, verify_json
 
 
-def _get_latest_version(versions: List[str]) -> str:
-    return "v" + str(sorted([semantic_version.Version.coerce(version[1:]) for version in versions])[-1])
-
-
 def get_ontology_info_file(ontology_info_file: str = env.ONTOLOGY_INFO_FILE) -> Any:
     """
     Read ontology information from file
@@ -345,7 +341,7 @@ def list_expired_cellxgene_schema_version(ontology_info: Dict[str, Any]) -> List
 if __name__ == "__main__":
     logging.basicConfig(level=logging.INFO)
     ontology_info = get_ontology_info_file()
-    current_version = _get_latest_version(ontology_info.keys())
+    current_version = get_latest_schema_version(ontology_info.keys())
     latest_ontology_version = ontology_info[current_version]
     latest_ontologies = latest_ontology_version["ontologies"]
     _download_ontologies(latest_ontologies)

diff --git a/tools/ontology-builder/tests/test_all_ontology_generator.py b/tools/ontology-builder/tests/test_all_ontology_generator.py
@@ -7,7 +7,6 @@
 import pytest
 from all_ontology_generator import (
     _download_ontologies,
-    _get_latest_version,
     _parse_ontologies,
     deprecate_previous_cellxgene_schema_versions,
     get_ontology_info_file,
@@ -46,14 +45,6 @@ def mock_raw_ontology_dir(tmpdir):
     return str(sub_dir)
 
 
-def test_get_latest_version():
-    # Call the function
-    latest_version = _get_latest_version(versions=["v1", "v2.0", "v3.0.0", "v3.0.1", "v3.1.0"])
-
-    # Assertion
-    assert latest_version == "v3.1.0"
-
-
 def test_get_ontology_info_file_default(mock_ontology_info_file):
     # Call the function
     ontology_info = get_ontology_info_file(ontology_info_file=mock_ontology_info_file)