fix: support qdrant-only vector maintenance

This commit is contained in:
2026-08-08 18:14:03 +02:00
parent 5e39cfa347
commit 32cd2165eb
9 changed files with 261 additions and 28 deletions
+27 -2
View File
@@ -3,8 +3,10 @@
import json
import re
from psycopg2 import Error as PsycopgError
from psycopg2 import sql
from sqlalchemy import Engine
from sqlalchemy.exc import SQLAlchemyError
from tht.config import DatabaseConfig
from tht.db.connection import make_engine
@@ -19,7 +21,6 @@ from tht.ports.vector import (
)
from tht.vectorstore.store import VectorHit, hit_from_metadata
COLLECTION_KINDS = {
"schema_records": {"schema_table", "schema_column"},
"evidence": {"evidence"},
@@ -243,7 +244,7 @@ class PgVectorStore:
return True, None, dimensions
finally:
raw.close()
except Exception as exc:
except (AttributeError, TypeError, ValueError, PsycopgError, SQLAlchemyError) as exc:
return False, f"vector database probe failed: {type(exc).__name__}", set()
def health(self) -> VectorHealth:
@@ -471,6 +472,30 @@ class PgVectorStore:
if raw is not None:
raw.close()
def delete_kinds(self, collection: str, kinds: list[str]) -> int:
_collection(self._schema, collection)
_validate_collection_kinds(collection, kinds)
raw = None
try:
raw = self._require_writer().raw_connection()
with raw.cursor() as cursor:
cursor.execute(
sql.SQL("DELETE FROM {} WHERE kind = ANY(%s)").format(
_collection(self._schema, collection)
),
(kinds,),
)
count = cursor.rowcount
raw.commit()
return count
except Exception as exc:
if raw is not None:
raw.rollback()
raise VectorWriteUnavailable("Vector kind cleanup unavailable") from exc
finally:
if raw is not None:
raw.close()
def list_evidence_generations(self, collection: str, workspace_id: str) -> list[str]:
if collection != "evidence":
raise VectorStoreError("Only exact Evidence generations may be listed")
+15
View File
@@ -212,6 +212,21 @@ class QdrantVectorStore:
)
return len(records)
def delete_kinds(self, collection: str, kinds: list[str]) -> int:
_collection("vectors", collection)
_validate_collection_kinds(collection, kinds)
must = [
*self._workspace_filter(),
{"key": "record_kind", "match": {"any": sorted(kinds)}},
]
before = len(self._scroll(must))
self._call(
"POST",
f"/collections/{self._collection}/points/delete?wait=true",
{"filter": {"must": must}},
)
return before
def delete_generation(self, collection: str, generation: str, workspace_id: str) -> int:
if collection != "evidence" or _GENERATION.fullmatch(generation) is None:
raise VectorStoreError("Only exact Evidence generations may be deleted")
+14 -6
View File
@@ -2,6 +2,11 @@
import re
from tht.adapters.vector.pgvector import (
_collection,
_validate_collection_kinds,
_validate_known_kinds,
)
from tht.ports.vector import (
VectorCapabilities,
VectorHealth,
@@ -14,11 +19,6 @@ from tht.ports.vector import (
)
from tht.vectorstore.rest_client import VectorRestClient, VectorRestError
from tht.vectorstore.store import hit_from_metadata
from tht.adapters.vector.pgvector import (
_collection,
_validate_collection_kinds,
_validate_known_kinds,
)
def _merge(hits: list[VectorHit], limit: int) -> list[VectorHit]:
@@ -85,7 +85,7 @@ class ThothHttpVectorStore:
return None, None, []
try:
return True, None, client.list_tables()
except Exception as exc:
except (RuntimeError, VectorRestError) as exc:
return False, str(exc), []
def search(
@@ -155,6 +155,14 @@ class ThothHttpVectorStore:
except VectorRestError as exc:
raise VectorStoreError(str(exc)) from exc
def delete_kinds(self, collection: str, kinds: list[str]) -> int:
_collection("vectors", collection)
_validate_collection_kinds(collection, kinds)
try:
return self._require_writer().delete_kinds(collection, kinds)
except VectorRestError as exc:
raise VectorStoreError(str(exc)) from exc
def delete_generation(self, collection: str, generation: str, workspace_id: str) -> int:
if collection != "evidence" or re.fullmatch(r"gen:[0-9a-f]{32}", generation) is None:
raise VectorStoreError("Only exact Evidence generations may be deleted")
+7 -5
View File
@@ -43,6 +43,12 @@ def _resync_memory(cfg):
)
def clear_memory_index(cfg):
from tht.adapters.factory import build_vector_store
return build_vector_store(cfg, require_write=True).delete_kinds("memory", ["memory"])
@memory_app.command("promote")
def promote_cmd(
session: str = typer.Option(..., "--session"),
@@ -192,7 +198,6 @@ def clear_cmd(
config: Path = CONFIG_OPT,
) -> None:
"""Cancella TUTTA la review memory: registro canonico + indice pgvector (kind=memory)."""
from tht.cli.vector_cmd import make_embedder, open_store, require_direct_vector_cfg
from tht.memory import load_registry
cfg = _load_config_or_exit(config)
@@ -209,10 +214,7 @@ def clear_cmd(
typer.secho("Annullato.", fg=typer.colors.YELLOW)
raise typer.Exit(code=1)
# Indice pgvector: rimuove i record kind=memory (sync con insieme vuoto).
require_direct_vector_cfg(cfg)
store = open_store(cfg, "memory")
store.sync([], make_embedder(cfg.embeddings), kinds={"memory"})
clear_memory_index(cfg)
# Registro canonico.
registry.unlink()
+2 -2
View File
@@ -26,8 +26,8 @@ def require_vector_cfg(cfg):
missing = []
if cfg.embeddings is None:
missing.append("embeddings")
if cfg.vector_db is None and not has_vector_write_rest(cfg):
missing.append("vector_db o vector_write_rest")
if cfg.vectors is None and cfg.vector_db is None and not has_vector_write_rest(cfg):
missing.append("vectors o vector_db o vector_write_rest")
if missing:
typer.secho(
f"ERRORE: sezioni mancanti nel workspace yaml: {', '.join(missing)}.",
+2
View File
@@ -80,6 +80,8 @@ class VectorStore(Protocol):
def upsert(self, collection: str, records: list[VectorWriteRecord]) -> int: ...
def delete_kinds(self, collection: str, kinds: list[str]) -> int: ...
def delete_generation(self, collection: str, generation: str, workspace_id: str) -> int: ...
def list_evidence_generations(self, collection: str, workspace_id: str) -> list[str]: ...
+19 -2
View File
@@ -5,9 +5,10 @@ Endpoint dedicato (es. https://host/vector/v1/), distinto dal DWH. La lettura us
Errori in italiano e azionabili, stile `rest/client.py`.
"""
import requests
import re
import requests
from tht.config import RestConfig
@@ -46,7 +47,7 @@ class VectorRestClient:
try:
body = resp.json()
detail = body.get("message") or body.get("details") or resp.text
except Exception:
except ValueError:
detail = resp.text
return f"Vector REST rpc {fn} → HTTP {resp.status_code}: {detail}"
@@ -149,6 +150,22 @@ class VectorRestClient:
return int(payload.get("deleted", 0))
return 0
def delete_kinds(self, table_name: str, kinds: list[str]) -> int:
try:
payload = self._call(
"delete_vector_kinds",
{"table_name": table_name, "kinds": kinds},
)
except VectorRestError as error:
if "HTTP 404" in str(error):
raise VectorRestError(
"delete_vector_kinds RPC is unavailable; deploy the cleanup migration"
) from None
raise
if isinstance(payload, dict):
return int(payload.get("deleted", 0))
return 0
def list_evidence_generations(self, table_name: str, workspace_id: str) -> list[str]:
if re.fullmatch(r"[a-z][a-z0-9_-]{0,63}", workspace_id) is None:
raise ValueError("workspace namespace must be canonical")