feat: P3 effective configuration, memory root, and revision-scoped records

harness/tht/config.py

@@ -1,3 +1,4 @@
+import hashlib
 import json
 import os
 import re
@@ -16,6 +17,74 @@ from tht.ports.evidence import canonical_provenance_uri
 
 _ENV_RE = re.compile(r"\$\{([A-Za-z_][A-Za-z0-9_]*)\}")
 
+def _canonical_fingerprint(value: str) -> str:
+    return "sha256:" + hashlib.sha256(value.encode("utf-8")).hexdigest()
+
+
+def canonical_effective_config_document(cfg) -> dict:
+    """Non-secret effective DWH/preprocessing configuration (versioned, P3).
+
+    Mirrors backend/src/workspaces/effective-config.ts: only fields that determine whether a
+    prepared DWH generation is reusable. Deliberately excludes session_storage, runtime_identity,
+    evidence, memory, search, execution, and every credential value.
+    """
+    dwh: dict = {"engine": "postgres"}
+    dwh_cfg = getattr(cfg, "dwh", None)
+    if dwh_cfg is None:
+        raise ConfigError("DWH configuration is unavailable; cannot canonicalize effective config")
+    transport = getattr(dwh_cfg, "type", None)
+    if transport == "postgres_direct":
+        conn = dwh_cfg.connection
+        dwh["database"] = conn.database
+        dwh["schema"] = getattr(conn, "db_schema", None) or getattr(conn, "schema", None) or conn.database
+        dwh["transport"] = "postgres_direct"
+        dwh["host"] = conn.host
+        dwh["port"] = conn.port
+        dwh["user"] = conn.user
+    elif transport == "thoth_rest":
+        dwh["database"] = dwh_cfg.database.database
+        dwh["schema"] = dwh_cfg.database.db_schema
+        dwh["transport"] = "rest_api"
+        dwh["baseUrl"] = dwh_cfg.endpoint.base_url
+    else:
+        raise ConfigError("unsupported DWH transport in canonical effective config")
+    vectors = getattr(cfg, "vectors", None)
+    collection = getattr(vectors, "collection", None) if vectors is not None else None
+    if not collection:
+        raise ConfigError("vector configuration is unavailable; cannot canonicalize effective config")
+    embeddings = getattr(cfg, "embeddings", None)
+    model = getattr(embeddings, "model", None) if embeddings is not None else None
+    embed_dim = getattr(embeddings, "dim", None) if embeddings is not None else None
+    if not model or not embed_dim:
+        raise ConfigError("embedding configuration is unavailable; cannot canonicalize effective config")
+    return {
+        "schemaVersion": 1,
+        "dwh": dwh,
+        "vector": {"collection": collection, "dimensions": 1024, "distance": "cosine"},
+        "embedding": {"model": model, "dimensions": int(embed_dim)},
+        "roots": {
+            "artifacts": str(getattr(cfg.paths, "artifacts", Path("artifacts"))),
+            "indexes": str(getattr(cfg.paths, "indexes", Path("indexes"))),
+        },
+    }
+
+
+def canonical_effective_config_json(cfg) -> str:
+    return json.dumps(canonical_effective_config_document(cfg), separators=(",", ":"), ensure_ascii=False)
+
+
+def effective_config_identity(workspace_id: str, cfg) -> str:
+    digest = hashlib.sha256(canonical_effective_config_json(cfg).encode("utf-8")).hexdigest()
+    return f"workspace://{workspace_id}@v1:{digest}"
+
+
+def effective_config_fingerprint(cfg) -> str:
+    return _canonical_fingerprint(canonical_effective_config_json(cfg))
+
+
+def effective_config_input_fingerprint(workspace_id: str, cfg) -> str:
+    return _canonical_fingerprint(effective_config_identity(workspace_id, cfg))
+
 
 class ConfigError(Exception):
     """Errore di configurazione, con messaggio leggibile per l'utente."""
@@ -252,6 +321,9 @@ class PathsConfig(BaseModel):
     artifacts: Path = Path("artifacts")
     indexes: Path = Path("indexes")
     sessions: Path = Path("sessions")
+    # Explicit workspace-global memory root (P3). When absent, legacy `artifacts/memory` is used
+    # only through the documented migration path.
+    memory: Path | None = None
 
 
 class RuntimeIdentityConfig(BaseModel):