Create a new option to output the schema to a file

.github/workflows/core-ci.yml

@@ -64,6 +64,14 @@ jobs:
         pip install git+https://github.com/sodascience/metasyn-disclosure-control
         pip install .
         metasyn create-meta metasyn/demo/demo_titanic.csv --config examples/config_files/example_config.toml
+    - name: Test GMF file
+      if: ${{ matrix.os == 'ubuntu-latest' }}
+      run: |
+        metasyn schema -o current_schema.json
+        MD5_DATA=(`md5sum current_schema.json`)
+        [[ ${MD5_DATA[0]} == "c2a69330cec7a147ab775f5f0e037d8b" ]]
+
+
 
   build-docs:
     name: Build documentation

.gitignore

@@ -154,3 +154,7 @@ docs/source/api/generated
 
 # uv stuff
 uv.lock
+
+# Schema stuff
+create_schema.py
+schemas/

metasyn/__main__.py

@@ -236,8 +236,14 @@ def schema() -> None:
         action="store_true",
     )
 
+    parser.add_argument(
+        "-o", "--output",
+        help="File to write the schema to.",
+        type=pathlib.Path,
+    )
+
     # parse the args without the subcommand
-    args, _ = parser.parse_known_args()
+    args = parser.parse_args()
 
     # deduplicated list of plugins for schema
     plugins_avail = {entry.name for entry in entry_points(group="metasyn.distribution_provider")}
@@ -257,7 +263,11 @@ def schema() -> None:
         )
         parser.error(errmsg)
     jsonschema = create_schema(list(plugins))
-    print(json.dumps(jsonschema, indent=2))
+    if args.output is None:
+        print(json.dumps(jsonschema, indent=4))
+    else:
+        with open(args.output, "w") as handle:
+            json.dump(jsonschema, handle, indent=4)
 
 
 if __name__ == "__main__":

metasyn/validation.py

@@ -19,7 +19,7 @@
 
 SCHEMA_BASE = {
     "$schema": "https://json-schema.org/draft/2020-12/schema",
-    "$id": "http://sodascience.github.io/generative_metadata_format/core/1.0.0/generative_metadata_format",  # noqa: E501
+    "$id": "http://sodascience.github.io/generative_metadata_format/core/1.1/generative_metadata_format",  # noqa: E501
     "type": "object",
     "properties": {
         "n_rows": {"type": "number"},