Add function queryPathFromHashPart()

To implement binary caches efficiently, Hydra needs to be able to map
the hash part of a store path (e.g. "gbg...zr7") to the full store
path (e.g. "/nix/store/gbg...kzr7-subversion-1.7.5").  (The binary
cache mechanism uses hash parts as a key for looking up store paths to
ensure privacy.)  However, doing a search in the Nix store for
/nix/store/<hash>* is expensive since it requires reading the entire
directory.  queryPathFromHashPart() prevents this by doing a cheap
database lookup.

1 files changed, 9 insertions, 0 deletions

diff --git a/src/nix-worker/nix-worker.cc b/src/nix-worker/nix-worker.cc
index 2f0a2ab209c5..74a619c71d0a 100644
--- a/src/nix-worker/nix-worker.cc
+++ b/src/nix-worker/nix-worker.cc
@@ -350,6 +350,15 @@ static void performOp(unsigned int clientVersion,
         break;
     }
 
+    case wopQueryPathFromHashPart: {
+        string hashPart = readString(from);
+        startWork();
+        Path path = store->queryPathFromHashPart(hashPart);
+        stopWork();
+        writeString(path, to);
+        break;
+    }
+
     case wopAddToStore: {
         string baseName = readString(from);
         bool fixed = readInt(from) == 1; /* obsolete */