@pentatonic-ai/ai-agent-sdk 0.10.0 → 0.10.2

package/packages/memory-engine-v2/tests/test_entity_id_parity.py

@@ -0,0 +1,57 @@
+"""Parity guard: extractor-sync/entity_id.py and extractor-async/entity_id.py MUST
+be byte-identical.
+
+The two extractors run as separate Docker services with per-service build contexts
+(docker-compose `context: ./extractor-sync` / `./extractor-async`), so they can't
+import a single shared module — the file is duplicated. If the copies ever drift,
+the two passes would key entities differently again and re-introduce the exact
+fragmentation RFC-entity-reconciliation.md (step 1) fixes. This test fails loudly if
+that happens. When you change one copy, copy it to the other.
+
+Run: pytest packages/memory-engine-v2/tests/test_entity_id_parity.py
+"""
+
+from __future__ import annotations
+
+import importlib.util
+from pathlib import Path
+
+_PKG = Path(__file__).resolve().parent.parent
+_SYNC = _PKG / "extractor-sync" / "entity_id.py"
+_ASYNC = _PKG / "extractor-async" / "entity_id.py"
+
+
+def test_both_copies_exist():
+    assert _SYNC.exists(), f"missing {_SYNC}"
+    assert _ASYNC.exists(), f"missing {_ASYNC}"
+
+
+def test_byte_identical():
+    assert _SYNC.read_bytes() == _ASYNC.read_bytes(), (
+        "extractor-sync/entity_id.py and extractor-async/entity_id.py have drifted — "
+        "they must be byte-identical so the two extractors key entities the same way. "
+        "Copy one over the other."
+    )
+
+
+def _load(path: Path, name: str):
+    spec = importlib.util.spec_from_file_location(name, path)
+    assert spec and spec.loader
+    mod = importlib.util.module_from_spec(spec)
+    spec.loader.exec_module(mod)
+    return mod
+
+
+def test_identical_output_across_copies():
+    a = _load(_SYNC, "entity_id_sync")
+    b = _load(_ASYNC, "entity_id_async")
+    cases = [
+        ("arena1", "person", "Carly Snider"),
+        ("arena1", "person", "carly@pactcollective.org"),
+        ("arena1", "person", "  Ben   Gordon "),
+        ("arena2", "org", "Pact Collective"),
+        ("arena1", "person", ""),
+    ]
+    for arena, etype, name in cases:
+        assert a.entity_id(arena, etype, name) == b.entity_id(arena, etype, name)
+        assert a.normalize_surface_form(name) == b.normalize_surface_form(name)