Stop snapshot purge when a snapshot cannot leave the local index (closes #337)

snapshot purge, and the purge step of snapshot create --prune, logged a
failed local index delete, then deleted that snapshot's metadata from
the destination store anyway and exited 0, leaving a local record whose
metadata was gone. It now stops at that snapshot with an error, before
touching its metadata, so the command exits non-zero; snapshots already
purged stay purged. This matches how the same loop already handles a
failed metadata delete.

Model: opus-5-5

internal/cli/entry_snapshot_purge_test.go

@@ -0,0 +1,90 @@
+package cli //nolint:testpackage // shares runEntry and the hermetic config helpers
+
+import (
+	"context"
+	"database/sql"
+	"os"
+	"path/filepath"
+	"testing"
+	"time"
+
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+	"sneak.berlin/go/vaultik/internal/database"
+	"sneak.berlin/go/vaultik/internal/snapshot"
+	"sneak.berlin/go/vaultik/internal/types"
+)
+
+// The `snapshot purge` command word and flags used below.
+const (
+	cmdPurge       = "purge"
+	flagKeepLatest = "--keep-latest"
+	flagForce      = "--force"
+)
+
+// TestEntrySnapshotPurgeKeepsMetadataWhenLocalIndexDeleteFails makes
+// removing a snapshot from the local index fail and checks that
+// `snapshot purge` exits non-zero and leaves that snapshot's metadata on
+// the destination store. It used to log the error, delete the metadata
+// anyway and exit 0.
+//
+//nolint:paralleltest // replaces os.Args, os.Stdout, os.Stderr and the xdg globals
+func TestEntrySnapshotPurgeKeepsMetadataWhenLocalIndexDeleteFails(t *testing.T) {
+	ctx := context.Background()
+	configPath := writeHermeticPruneConfig(t, false)
+
+	// writeHermeticPruneConfig puts the index and the store directory
+	// beside the config file.
+	dir := filepath.Dir(configPath)
+
+	// Two snapshots of the same name: --keep-latest purges the older.
+	olderStart := time.Date(2026, 5, 1, 9, 0, 0, 0, time.UTC)
+	newerStart := olderStart.Add(time.Hour)
+	olderID := "test-host_test_" + olderStart.Format(time.RFC3339)
+	newerID := "test-host_test_" + newerStart.Format(time.RFC3339)
+
+	db, err := database.New(ctx, filepath.Join(dir, "index.sqlite"))
+	require.NoError(t, err)
+
+	repos := database.NewRepositories(db)
+
+	manifestPaths := make(map[string]string)
+
+	for id, startedAt := range map[string]time.Time{
+		olderID: olderStart,
+		newerID: newerStart,
+	} {
+		completedAt := startedAt.Add(time.Minute)
+		snap := &database.Snapshot{
+			ID:             types.SnapshotID(id),
+			Hostname:       "test-host",
+			VaultikVersion: "test",
+			StartedAt:      startedAt,
+			CompletedAt:    &completedAt,
+		}
+
+		err = repos.WithTx(ctx, func(ctx context.Context, tx *sql.Tx) error {
+			return repos.Snapshots.Create(ctx, tx, snap)
+		})
+		require.NoError(t, err)
+
+		manifestPaths[id] = filepath.Join(dir, "store", "metadata",
+			snapshot.RemoteSnapshotKey(id), "manifest.json.zst")
+		require.NoError(t, os.MkdirAll(filepath.Dir(manifestPaths[id]), 0o750))
+		require.NoError(t, os.WriteFile(manifestPaths[id], []byte("stub"), 0o600))
+	}
+
+	_, err = db.Conn().ExecContext(ctx, `
+		CREATE TRIGGER refuse_snapshot_delete BEFORE DELETE ON snapshots
+		BEGIN SELECT RAISE(ABORT, 'simulated index error'); END`)
+	require.NoError(t, err)
+	require.NoError(t, db.Close())
+
+	code, _, stderr := runEntry(t,
+		flagConfig, configPath, cmdSnapshot, cmdPurge, flagKeepLatest, flagForce)
+
+	assert.Equal(t, 1, code)
+	assert.Contains(t, stderr, "simulated index error")
+	assert.FileExists(t, manifestPaths[olderID],
+		"a snapshot still in the local index keeps its metadata")
+}

internal/vaultik/snapshot.go

@@ -627,7 +627,8 @@ func selectSnapshotsToPurge(
 }
 
 // confirmAndExecutePurge shows deletion candidates, confirms with the
-// user, and deletes snapshots.
+// user, and deletes snapshots. It stops with an error at the first
+// snapshot it cannot delete.
 func (v *Vaultik) confirmAndExecutePurge(
 	toDelete []SnapshotInfo, force, quiet bool,
 ) error {
@@ -669,10 +670,11 @@ func (v *Vaultik) confirmAndExecutePurge(
 		snapshotID := snap.ID.String()
 		log.Info("Deleting snapshot", "id", snapshotID)
 
+		// Stop before the remote delete: a snapshot still in the local
+		// index must keep its metadata on the destination store.
 		err := v.deleteSnapshotFromLocalDB(snapshotID)
 		if err != nil {
-			log.Error("Failed to delete from local database",
-				"snapshot_id", snapshotID, "error", err)
+			return fmt.Errorf("deleting snapshot %s from local index: %w", snapshotID, err)
 		}
 
 		_, err = v.deleteRemoteSnapshotByKey(snapshot.RemoteSnapshotKey(snapshotID))