fix: harden task2 preprocessing and vector guards

harness/tht/cli/preprocess_cmd.py

@@ -4,6 +4,7 @@ from __future__ import annotations
 
 import hashlib
 import json
+import logging
 import re
 from pathlib import Path
 
@@ -12,8 +13,17 @@ import typer
 from tht.cli.config_cmd import CONFIG_OPT
 from tht.cli.schema_cmd import _load_config_or_exit
 from tht.config import workspace_id_for_config
+from tht.ports.evidence import EvidenceSourceError
+from tht.ports.vector import VectorStoreError
+from tht.vectorstore.embeddings import EmbeddingsError
 
 preprocess_app = typer.Typer(help="Materialize versioned preprocessing artifacts")
+logger = logging.getLogger(__name__)
+
+_PREPROCESS_EXPECTED_ERRORS = (
+    OSError, RuntimeError, ValueError, TypeError, KeyError,
+    EvidenceSourceError, VectorStoreError, EmbeddingsError,
+)
 
 
 def run_dwh_from_config(
@@ -132,7 +142,16 @@ def evidence_cmd(
     if action == "gc":
         try:
             payload = gc_from_config(config, dry_run=dry_run)
-        except (OSError, RuntimeError, ValueError, TypeError, KeyError):
+        except _PREPROCESS_EXPECTED_ERRORS:
+            payload = {"status": "failed", "error": "evidence cleanup failed"}
+            if json_output:
+                typer.echo(json.dumps(payload, sort_keys=True))
+            else:
+                typer.secho("ERRORE: evidence cleanup failed", fg=typer.colors.RED, err=True)
+            raise typer.Exit(code=1) from None
+        except Exception:
+            if not json_output:
+                logger.exception("Evidence cleanup failed")
             payload = {"status": "failed", "error": "evidence cleanup failed"}
             if json_output:
                 typer.echo(json.dumps(payload, sort_keys=True))
@@ -153,7 +172,16 @@ def evidence_cmd(
         raise typer.Exit(code=2)
     try:
         result = run_from_config(config, dry_run=dry_run, resume=resume)
-    except (OSError, RuntimeError, ValueError, TypeError, KeyError):
+    except _PREPROCESS_EXPECTED_ERRORS:
+        payload = {"status": "failed", "error": "preprocessing failed"}
+        if json_output:
+            typer.echo(json.dumps(payload, sort_keys=True))
+        else:
+            typer.secho("ERRORE: preprocessing failed", fg=typer.colors.RED, err=True)
+        raise typer.Exit(code=1) from None
+    except Exception:
+        if not json_output:
+            logger.exception("Evidence preprocessing failed")
         payload = {"status": "failed", "error": "preprocessing failed"}
         if json_output:
             typer.echo(json.dumps(payload, sort_keys=True))

harness/tht/cli/schema_cmd.py

@@ -448,10 +448,10 @@ def check_cmd(
             raise typer.Exit(code=1) from None
         _render_schema_machine_error(error, cfg)
     except Exception:
-        logger.exception("Schema check failed")
         if json_output:
             _schema_json({"status": "failed", "code": "schema_check_failed"})
             raise typer.Exit(code=1) from None
+        logger.exception("Schema check failed")
         typer.secho("ERRORE: impossibile verificare le annotazioni.", fg=typer.colors.RED, err=True)
         raise typer.Exit(code=1) from None
     if json_output:
@@ -531,10 +531,10 @@ def suggest_fks_cmd(
             raise typer.Exit(code=1) from None
         _render_schema_machine_error(error, cfg)
     except Exception:
-        logger.exception("Schema suggestion failed")
         if json_output:
             _schema_json({"status": "failed", "code": "schema_suggestion_failed"})
             raise typer.Exit(code=1) from None
+        logger.exception("Schema suggestion failed")
         typer.secho("ERRORE: impossibile elaborare il catalogo.", fg=typer.colors.RED, err=True)
         raise typer.Exit(code=1) from None
     if json_output:

harness/tht/cli/vector_cmd.py

@@ -208,9 +208,10 @@ def index_schema_data(
         cfg = config
     _vector_write_or_error(cfg)
     _vector_cfg_or_error(cfg)
-    phys_file = physical_path(cfg)
-    if not phys_file.exists():
-        raise _MachineVectorError("physical_schema_missing")
+    if physical is None:
+        phys_file = physical_path(cfg)
+        if not phys_file.exists():
+            raise _MachineVectorError("physical_schema_missing")
     import yaml
     from pydantic import ValidationError
 
@@ -259,6 +260,9 @@ def index_schema_cmd(
     else:
         cfg = _load_config_or_exit(config)
     try:
+        # Authorization and configuration are checked before touching any artifacts.
+        _vector_write_or_error(cfg)
+        _vector_cfg_or_error(cfg)
         physical, annotations = _load_schema_artifacts(cfg)
         payload = index_schema_data(cfg, physical=physical, annotations=annotations)
     except _MachineVectorError as error:
@@ -267,10 +271,10 @@ def index_schema_cmd(
             raise typer.Exit(code=1) from None
         _render_index_schema_error(error, cfg)
     except Exception:
-        logger.exception("Schema indexing failed")
         if json_output:
             typer.echo(json.dumps({"status": "failed", "code": "schema_index_failed"}, sort_keys=True, separators=(",", ":")))
             raise typer.Exit(code=1) from None
+        logger.exception("Schema indexing failed")
         typer.secho("ERRORE: impossibile indicizzare lo schema.", fg=typer.colors.RED, err=True)
         raise typer.Exit(code=1) from None
     if json_output: