Add tests for remote YAML ref loading and caching

With any non-`.json` suffix, e.g., `.yaml`, we need the cache to
preserve extensions in order to be certain that the cache loader will
choose the same parser when loading the cached ref as the one it used
when the ref was remote. The `.yaml` ref + cache population trickery
allows us to test this codepath in an acceptance test.

Author: sirosen

tests/acceptance/test_nonjson_schema_handling.py

@@ -1,8 +1,10 @@
 import json
 
 import pytest
+import responses
 
 from check_jsonschema.parsers.json5 import ENABLED as JSON5_ENABLED
+from check_jsonschema.schema_loader.resolver import ref_url_to_cache_filename
 
 SIMPLE_SCHEMA = {
     "$schema": "http://json-schema.org/draft-07/schema",
@@ -87,3 +89,79 @@ def test_can_load_json5_schema(run_line, tmp_path, passing_data):
         ["check-jsonschema", "--schemafile", str(main_schemafile), str(doc)]
     )
     assert result.exit_code == (0 if passing_data else 1)
+
+
+@pytest.mark.parametrize("passing_data", [True, False])
+def test_can_load_remote_yaml_schema(run_line, tmp_path, passing_data):
+    retrieval_uri = "https://example.org/retrieval/schemas/main.yaml"
+    responses.add(
+        "GET",
+        retrieval_uri,
+        body="""\
+"$schema": "http://json-schema.org/draft-07/schema"
+properties:
+  title: {"type": "string"}
+additionalProperties: false
+""",
+    )
+
+    doc = tmp_path / "doc.json"
+    doc.write_text(json.dumps(PASSING_DOCUMENT if passing_data else FAILING_DOCUMENT))
+
+    result = run_line(["check-jsonschema", "--schemafile", retrieval_uri, str(doc)])
+    assert result.exit_code == (0 if passing_data else 1)
+
+
+@pytest.mark.parametrize("passing_data", [True, False])
+def test_can_load_remote_yaml_schema_ref(run_line, tmp_path, passing_data):
+    retrieval_uri = "https://example.org/retrieval/schemas/main.yaml"
+    responses.add(
+        "GET",
+        retrieval_uri,
+        body="""\
+"$schema": "http://json-schema.org/draft-07/schema"
+properties:
+  "title": {"$ref": "./title_schema.yaml"}
+additionalProperties: false
+""",
+    )
+    responses.add(
+        "GET",
+        "https://example.org/retrieval/schemas/title_schema.yaml",
+        body="type: string",
+    )
+
+    doc = tmp_path / "doc.json"
+    doc.write_text(json.dumps(PASSING_DOCUMENT if passing_data else FAILING_DOCUMENT))
+
+    result = run_line(["check-jsonschema", "--schemafile", retrieval_uri, str(doc)])
+    assert result.exit_code == (0 if passing_data else 1)
+
+
+def test_can_load_remote_yaml_schema_ref_from_cache(run_line, cache_dir, tmp_path):
+    retrieval_uri = "https://example.org/retrieval/schemas/main.yaml"
+    responses.add(
+        "GET",
+        retrieval_uri,
+        body="""\
+"$schema": "http://json-schema.org/draft-07/schema"
+properties:
+  "title": {"$ref": "./title_schema.yaml"}
+additionalProperties: false
+""",
+    )
+
+    ref_loc = "https://example.org/retrieval/schemas/title_schema.yaml"
+    # populate a bad schema, but then "override" that with a good cache value
+    # this can only pass (in the success case) if the cache loading really works
+    responses.add("GET", ref_loc, body="false")
+    ref_cache_dir = cache_dir / "check_jsonschema" / "refs"
+    ref_cache_dir.mkdir(parents=True)
+    ref_path = ref_cache_dir / ref_url_to_cache_filename(ref_loc)
+    ref_path.write_text("type: string")
+
+    doc = tmp_path / "doc.json"
+    doc.write_text(json.dumps(PASSING_DOCUMENT))
+
+    result = run_line(["check-jsonschema", "--schemafile", retrieval_uri, str(doc)])
+    assert result.exit_code == 0