Add support for ClassVar variables (#227)

docs/source/advanced.ipynb

@@ -139,6 +139,38 @@
     "df.dtypes"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# ClassVar variables\n",
+    "\n",
+    "Variables annotated with `typing.ClassVar` variables are not included in the schema, so these can be used for example to store metadata about the DataSet."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from typing import ClassVar\n",
+    "\n",
+    "class Schema:\n",
+    "    id: int\n",
+    "    name: str\n",
+    "    file_name: ClassVar[str] = \"schema_data.csv\"\n",
+    "\n",
+    "df1 = DataSet[Schema](\n",
+    "    {\n",
+    "        \"id\": [1, 2, 3],\n",
+    "        \"name\": [\"John\", \"Jane\", \"Jack\"],\n",
+    "    }\n",
+    ")\n",
+    "\n",
+    "print(Schema.file_name)"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},

strictly_typed_pandas/validate_schema.py

@@ -1,4 +1,4 @@
-from typing import Any, Dict, Set
+from typing import Any, ClassVar, Dict, Set, get_origin
 
 import numpy as np  # type: ignore
 from pandas.api.extensions import ExtensionDtype
@@ -14,7 +14,15 @@ def check_for_duplicate_columns(names_index: Set[str], names_data: Set[str]) ->
         raise TypeError(msg.format(intersection))
 
 
+def remove_classvars(schema_expected: Dict[str, Any]) -> Dict[str, Any]:
+    return {
+        key: value
+        for key, value in schema_expected.items() if get_origin(value) is not ClassVar
+    }
+
+
 def validate_schema(schema_expected: Dict[str, Any], schema_observed: Dict[str, Any]):
+    schema_expected = remove_classvars(schema_expected)
     _check_names(set(schema_expected.keys()), set(schema_observed.keys()))
     _check_dtypes(schema_expected, schema_observed)
 

tests/test_dataset.py

@@ -1,5 +1,6 @@
 import pickle
 import tempfile
+from typing import ClassVar
 
 import numpy as np  # type: ignore
 import pandas as pd
@@ -18,6 +19,11 @@ class AlternativeSchema:
     a: int
 
 
+class SchemaWithClassVar:
+    a: int
+    b: ClassVar[str] = "abc"
+
+
 dictionary = {"a": [1, 2, 3], "b": ["a", "b", "c"]}
 
 
@@ -111,6 +117,15 @@ def test_pickle():
     assert (df == loaded).all().all()
 
 
+def test_classvar_colum_not_allowed():
+    with pytest.raises(TypeError):
+        DataSet[SchemaWithClassVar](dictionary)
+
+
+def test_classvar_colum_not_required():
+    DataSet[SchemaWithClassVar]({"a": [1, 2, 3]})
+
+
 class A:
     a: int