From b44f0fbfbd997360fa3e21ed59555d2d85767022 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Walter-=E6=A2=81=E6=96=87=E8=B6=85?= <Walter@fanruan.com>
Date: Wed, 5 Aug 2026 17:50:17 +0800
Subject: [PATCH] fix: validate relationship column pair counts

---
 validation/tests/test_validate.py | 60 +++++++++++++++++++++++++++++++
 validation/validate.py            |  7 ++++
 2 files changed, 67 insertions(+)
 create mode 100644 validation/tests/test_validate.py

diff --git a/validation/tests/test_validate.py b/validation/tests/test_validate.py
new file mode 100644
index 00000000..8e4f23da
--- /dev/null
+++ b/validation/tests/test_validate.py
@@ -0,0 +1,60 @@
+from importlib.util import module_from_spec, spec_from_file_location
+from pathlib import Path
+
+
+_VALIDATE_PATH = Path(__file__).parents[1] / "validate.py"
+_SPEC = spec_from_file_location("ossie_validate", _VALIDATE_PATH)
+assert _SPEC is not None and _SPEC.loader is not None
+_VALIDATE = module_from_spec(_SPEC)
+_SPEC.loader.exec_module(_VALIDATE)
+
+validate_references = _VALIDATE.validate_references
+
+
+def _document_with_relationship(from_columns: list[str], to_columns: list[str]) -> dict:
+    return {
+        "version": "0.2.0.dev0",
+        "semantic_model": [
+            {
+                "name": "m",
+                "datasets": [
+                    {"name": "orders", "source": "db.s.orders"},
+                    {"name": "customers", "source": "db.s.customers"},
+                ],
+                "relationships": [
+                    {
+                        "name": "orders_to_customers",
+                        "from": "orders",
+                        "to": "customers",
+                        "from_columns": from_columns,
+                        "to_columns": to_columns,
+                    }
+                ],
+            }
+        ],
+    }
+
+
+def test_validate_references_rejects_mismatched_relationship_column_counts() -> None:
+    errors = validate_references(
+        _document_with_relationship(
+            from_columns=["customer_id", "region_id"],
+            to_columns=["id"],
+        )
+    )
+
+    assert errors == [
+        "[Relationship] Relationship 'orders_to_customers' in model 'm' has "
+        "2 from_columns but 1 to_columns"
+    ]
+
+
+def test_validate_references_accepts_matching_relationship_column_counts() -> None:
+    errors = validate_references(
+        _document_with_relationship(
+            from_columns=["customer_id", "region_id"],
+            to_columns=["id", "region_id"],
+        )
+    )
+
+    assert errors == []
diff --git a/validation/validate.py b/validation/validate.py
index 258d34f1..7c7dbe90 100644
--- a/validation/validate.py
+++ b/validation/validate.py
@@ -140,11 +140,18 @@ def validate_references(data: dict) -> list[str]:
             rel_name = rel.get("name", "<unnamed>")
             from_ds = rel.get("from")
             to_ds = rel.get("to")
+            from_columns = rel.get("from_columns") or []
+            to_columns = rel.get("to_columns") or []
 
             if from_ds and from_ds not in dataset_names:
                 errors.append(f"[Reference] Relationship '{rel_name}' in model '{model_name}' references unknown dataset '{from_ds}'")
             if to_ds and to_ds not in dataset_names:
                 errors.append(f"[Reference] Relationship '{rel_name}' in model '{model_name}' references unknown dataset '{to_ds}'")
+            if len(from_columns) != len(to_columns):
+                errors.append(
+                    f"[Relationship] Relationship '{rel_name}' in model '{model_name}' has "
+                    f"{len(from_columns)} from_columns but {len(to_columns)} to_columns"
+                )
 
     return errors