objectpool: Verify object IDs when rescuing dangling objects

When we rescue dangling objects in an object pool we execute git-fsck(1)
to find out about any objects which are not currently referenced. We
don't perform any sanity checks though on the output of git-fsck(1): the
only thing we do is to split the string, and then pass whatever data we
have obtained to git-update-ref(1).

Verify that the obtained data is actually a conforming object hash by
parsing the object ID. This also prepares for an upcoming change where
the updateref package will start to accept `git.ObjectID`s instead of
bare strings.

1 files changed, 6 insertions, 1 deletions

diff --git a/internal/git/objectpool/fetch.go b/internal/git/objectpool/fetch.go
index 0ee7d6d0a..cb7516e5d 100644
--- a/internal/git/objectpool/fetch.go
+++ b/internal/git/objectpool/fetch.go
@@ -250,8 +250,13 @@ func (o *ObjectPool) rescueDanglingObjects(ctx context.Context) error {
 			continue
 		}
 
+		danglingObjectID, err := git.ObjectHashSHA1.FromHex(split[2])
+		if err != nil {
+			return fmt.Errorf("parsing object ID %q: %w", split[2], err)
+		}
+
 		ref := git.ReferenceName(danglingObjectNamespace + split[2])
-		if err := updater.Create(ref, split[2]); err != nil {
+		if err := updater.Create(ref, danglingObjectID.String()); err != nil {
 			return err
 		}
 	}