feat: index semantic records in qdrant

harness/tht/config.py

@@ -152,8 +152,14 @@ class ThothVectorHttpConfig(BaseModel):
     direct: DatabaseConfig | None = None
 
 
+class QdrantConfig(BaseModel):
+    type: Literal["qdrant"]
+    base_url: str
+    collection: str = Field(min_length=1)
+
+
 VectorResourceConfig = Annotated[
-    PgvectorDirectConfig | ThothVectorHttpConfig,
+    PgvectorDirectConfig | ThothVectorHttpConfig | QdrantConfig,
     Field(discriminator="type"),
 ]
 
@@ -397,6 +403,7 @@ def load_config(path: Path) -> Config:
         raise ConfigError(f"Configurazione non valida (atteso un mapping YAML): {path}")
     expanded = _resolve_secret_files(_expand_env(raw))
     _validate_internal_embedding_contract(expanded, path)
+    _validate_internal_vector_contract(expanded, path)
     translated, used_legacy = translate_legacy_config(expanded)
     _populate_legacy_views(translated)
     try:
@@ -455,6 +462,7 @@ def load_config(path: Path) -> Config:
         else path.resolve().as_posix()
     )
     _validate_active_embeddings_config(cfg.embeddings, path)
+    _validate_active_vector_config(cfg.vectors, path)
     return cfg
 
 
@@ -500,6 +508,42 @@ def _validate_internal_embedding_contract(raw: dict[str, Any], path: Path) -> No
         )
 
 
+def _validate_internal_vector_contract(raw: dict[str, Any], path: Path) -> None:
+    resources = raw.get("resources")
+    if not isinstance(resources, dict):
+        return
+    vector = resources.get("vector")
+    if not isinstance(vector, dict):
+        return
+
+    engine = vector.get("engine")
+    base_url = vector.get("base_url")
+    collection = vector.get("collection")
+    allowed = {"engine", "base_url", "collection"}
+    unexpected = sorted(set(vector) - allowed)
+    if unexpected:
+        raise ConfigError(
+            f"Configurazione non valida in {path}:\n"
+            f"resources.vector non supporta: {', '.join(unexpected)}"
+        )
+    if engine != "qdrant":
+        raise ConfigError(
+            f"Configurazione non valida in {path}:\n"
+            "resources.vector.engine deve essere 'qdrant'"
+        )
+    if not isinstance(collection, str) or not collection:
+        raise ConfigError(
+            f"Configurazione non valida in {path}:\n"
+            "resources.vector.collection deve essere valorizzato"
+        )
+    if not _is_allowed_internal_qdrant_url(base_url):
+        raise ConfigError(
+            f"Configurazione non valida in {path}:\n"
+            "resources.vector.base_url deve usare http://qdrant:6333 "
+            "oppure un endpoint loopback di sviluppo su porta 6333"
+        )
+
+
 def _validate_active_embeddings_config(
     embeddings: "EmbeddingsConfig | None",
     path: Path,
@@ -529,6 +573,20 @@ def _validate_active_embeddings_config(
         )
 
 
+def _validate_active_vector_config(
+    vectors: "VectorResourceConfig | None",
+    path: Path,
+) -> None:
+    if vectors is None or vectors.type != "qdrant":
+        return
+    if not _is_allowed_internal_qdrant_url(vectors.base_url):
+        raise ConfigError(
+            f"Configurazione non valida in {path}:\n"
+            "vectors.base_url deve usare http://qdrant:6333 "
+            "oppure un endpoint loopback di sviluppo su porta 6333"
+        )
+
+
 def _is_allowed_internal_embedding_url(value: Any) -> bool:
     if not isinstance(value, str):
         return False
@@ -548,6 +606,25 @@ def _is_allowed_internal_embedding_url(value: Any) -> bool:
     return host.is_loopback
 
 
+def _is_allowed_internal_qdrant_url(value: Any) -> bool:
+    if not isinstance(value, str):
+        return False
+    parsed = urlparse(value)
+    if parsed.scheme != "http" or not parsed.hostname or parsed.port != 6333:
+        return False
+    if parsed.params or parsed.query or parsed.fragment:
+        return False
+    if parsed.path not in ("", "/"):
+        return False
+    if parsed.hostname == "qdrant":
+        return True
+    try:
+        host = ip_address(parsed.hostname)
+    except ValueError:
+        return parsed.hostname == "localhost"
+    return host.is_loopback
+
+
 def _populate_legacy_views(raw: dict[str, Any]) -> None:
     """Populate old Config attributes for command compatibility during migration."""
     dwh = raw.get("dwh")