Initial SQL-only 1C adapter baseline

tests/1c/test_semantic_cache_embedding_worker.py

@@ -0,0 +1,75 @@
+from __future__ import annotations
+
+import json
+import sys
+from pathlib import Path
+
+
+ROOT = Path(__file__).resolve().parents[2]
+sys.path.insert(0, str(ROOT / "scripts"))
+
+import embed_1c_semantic_cache as worker  # noqa: E402
+
+
+def test_embed_pending_semantic_cache_upserts_with_precondition(monkeypatch) -> None:
+    calls: list[tuple[str, dict]] = []
+
+    def fake_adapter_call(adapter_url: str, method: str, payload: dict, *, timeout_seconds: int = 180) -> dict:
+        calls.append((method, payload))
+        if method == "semantic.cache.pending":
+            return {
+                "status": "ok",
+                "documents": [
+                    {
+                        "document_id": "doc-1",
+                        "content_sha1": "a" * 40,
+                        "text": "ОбластьШапка макета",
+                    }
+                ],
+            }
+        if method == "semantic.cache.embedding.upsert":
+            assert payload["document_id"] == "doc-1"
+            assert payload["content_sha1"] == "a" * 40
+            assert payload["embedding_model"] == "local-hashing-v1"
+            assert isinstance(payload["embedding"], list)
+            assert len(payload["embedding"]) == 16
+            return {"status": "ok", "dimensions": len(payload["embedding"])}
+        raise AssertionError(method)
+
+    monkeypatch.setattr(worker, "adapter_call", fake_adapter_call)
+
+    result = worker.embed_pending_semantic_cache(
+        adapter_url="http://adapter/rpc",
+        base_id="upo_test",
+        limit=1,
+        dimensions=16,
+    )
+
+    assert result["status"] == "ok"
+    assert result["counts"] == {"pending": 1, "processed": 1, "stored": 1, "conflicts": 0, "skipped": 0, "errors": 0}
+    assert [method for method, _ in calls] == ["semantic.cache.pending", "semantic.cache.embedding.upsert"]
+
+
+def test_embed_pending_semantic_cache_dry_run_does_not_upsert(monkeypatch) -> None:
+    calls: list[str] = []
+
+    def fake_adapter_call(adapter_url: str, method: str, payload: dict, *, timeout_seconds: int = 180) -> dict:
+        calls.append(method)
+        return {
+            "status": "ok",
+            "documents": [{"document_id": "doc-1", "content_sha1": "b" * 40, "text": "Реквизиты"}],
+        }
+
+    monkeypatch.setattr(worker, "adapter_call", fake_adapter_call)
+
+    result = worker.embed_pending_semantic_cache(
+        adapter_url="http://adapter/rpc",
+        base_id="upo_test",
+        dimensions=8,
+        dry_run=True,
+    )
+
+    assert result["counts"]["processed"] == 1
+    assert result["counts"]["stored"] == 0
+    assert result["upserts"][0]["status"] == "dry_run"
+    assert calls == ["semantic.cache.pending"]