refactor(vector): define store contract

This commit is contained in:
2026-07-11 20:24:14 +02:00
parent f6302b31dd
commit ff4d662aba
6 changed files with 314 additions and 22 deletions
+11 -22
View File
@@ -9,8 +9,10 @@ Entrambe mappano i `kind` sulle tabelle per-dominio dello schema `vectors`.
from sqlalchemy import Engine
from tht.adapters.vector.legacy_direct import LegacyDirectVectorStore
from tht.adapters.vector.thoth_http import ThothHttpVectorStore
from tht.vectorstore.rest_client import VectorRestClient
from tht.vectorstore.store import VectorHit, VectorStore, hit_from_metadata
from tht.vectorstore.store import VectorHit
# kind Thoth → tabella dello schema `vectors`.
KIND_TO_TABLE = {
@@ -30,30 +32,19 @@ def tables_for_kinds(kinds: list[str] | None) -> list[str]:
return sorted({KIND_TO_TABLE[k] for k in kinds if k in KIND_TO_TABLE})
def _merge(hits: list[VectorHit], top_n: int) -> list[VectorHit]:
return sorted(hits, key=lambda h: h.similarity, reverse=True)[:top_n]
class RestSearcher:
"""Similarity search via REST: una chiamata `search_similar` per tabella, poi fusione."""
def __init__(self, client: VectorRestClient):
self.client = client
self._store = ThothHttpVectorStore(reader=client, writer=None)
def search(
self, query_vec: list[float], top_n: int = 10, kinds: list[str] | None = None
) -> list[VectorHit]:
hits: list[VectorHit] = []
for table in tables_for_kinds(kinds):
for row in self.client.search_similar(table, query_vec, top_n, kinds=kinds):
hits.append(hit_from_metadata(row.get("similarity", 0.0), row.get("metadata")))
# Il filtro per kind avviene server-side (RPC con `kinds`); il post-filter resta
# come difesa per il fallback legacy (server pre-migrazione: 404 -> query senza
# filtro) e per parita' col path diretto (#25).
if kinds:
allowed = set(kinds)
hits = [h for h in hits if h.kind in allowed]
return _merge(hits, top_n)
return self._store.search(
tables_for_kinds(kinds), query_vec, limit=top_n, kinds=kinds
)
class DirectSearcher:
@@ -63,13 +54,11 @@ class DirectSearcher:
self.engine = engine
self.schema = schema
self.dim = dim
self._store = LegacyDirectVectorStore(engine, schema=schema, dim=dim)
def search(
self, query_vec: list[float], top_n: int = 10, kinds: list[str] | None = None
) -> list[VectorHit]:
hits: list[VectorHit] = []
for table in tables_for_kinds(kinds):
store = VectorStore(self.engine, schema=self.schema, table=table, dim=self.dim)
# passa kinds: dentro schema_records filtra schema_table vs schema_column (#25).
hits.extend(store.search(query_vec, top_n=top_n, kinds=kinds))
return _merge(hits, top_n)
return self._store.search(
tables_for_kinds(kinds), query_vec, limit=top_n, kinds=kinds
)