From 4012954b596b725dd61d49668691a69d491120c3 Mon Sep 17 00:00:00 2001
From: Eelco Dolstra <edolstra@gmail.com>
Date: Mon, 14 Oct 2024 13:53:54 +0200
Subject: [PATCH 1/2] Handle tarballs where directory entries are not
 contiguous

I.e. when not all entries underneath a directory X follow eachother,
but there is some entry Y that isn't a child of X in between.

Fixes #11656.
---
 src/libfetchers/git-utils.cc | 18 +++++++++++++++++-
 1 file changed, 17 insertions(+), 1 deletion(-)

diff --git a/src/libfetchers/git-utils.cc b/src/libfetchers/git-utils.cc
index d13daf887..95ee33089 100644
--- a/src/libfetchers/git-utils.cc
+++ b/src/libfetchers/git-utils.cc
@@ -977,8 +977,24 @@ struct GitFileSystemObjectSinkImpl : GitFileSystemObjectSink
 
     void pushBuilder(std::string name)
     {
+        const git_tree_entry * entry;
+        Tree prevTree = nullptr;
+
+        if (!pendingDirs.empty() &&
+            (entry = git_treebuilder_get(pendingDirs.back().builder.get(), name.c_str())))
+        {
+            /* Clone a tree that we've already finished. This happens
+               if a tarball has directory entries that are not
+               contiguous. */
+            if (git_tree_entry_type(entry) != GIT_OBJECT_TREE)
+                throw Error("parent of '%s' is not a directory", name);
+
+            if (git_tree_entry_to_object((git_object * *) (git_tree * *) Setter(prevTree), *repo, entry))
+                throw Error("looking up parent of '%s': %s", name, git_error_last()->message);
+        }
+
         git_treebuilder * b;
-        if (git_treebuilder_new(&b, *repo, nullptr))
+        if (git_treebuilder_new(&b, *repo, prevTree.get()))
             throw Error("creating a tree builder: %s", git_error_last()->message);
         pendingDirs.push_back({ .name = std::move(name), .builder = TreeBuilder(b) });
     };

From a7b9877da9d1bdafcc9b2f4681ecb3a1b83de7fc Mon Sep 17 00:00:00 2001
From: Eelco Dolstra <edolstra@gmail.com>
Date: Mon, 14 Oct 2024 14:10:36 +0200
Subject: [PATCH 2/2] Add a test

---
 tests/functional/tarball.sh | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/tests/functional/tarball.sh b/tests/functional/tarball.sh
index 0202037aa..0682869b2 100755
--- a/tests/functional/tarball.sh
+++ b/tests/functional/tarball.sh
@@ -97,3 +97,17 @@ chmod +x "$TEST_ROOT/tar_root/foo"
 tar cvf "$TEST_ROOT/tar.tar" -C "$TEST_ROOT/tar_root" .
 path="$(nix flake prefetch --refresh --json "tarball+file://$TEST_ROOT/tar.tar" | jq -r .storePath)"
 [[ $(cat "$path/foo") = bar ]]
+
+# Test a tarball with non-contiguous directory entries.
+rm -rf "$TEST_ROOT/tar_root"
+mkdir -p "$TEST_ROOT/tar_root/a/b"
+echo foo > "$TEST_ROOT/tar_root/a/b/foo"
+echo bla > "$TEST_ROOT/tar_root/bla"
+tar cvf "$TEST_ROOT/tar.tar" -C "$TEST_ROOT/tar_root" .
+echo abc > "$TEST_ROOT/tar_root/bla"
+echo xyzzy > "$TEST_ROOT/tar_root/a/b/xyzzy"
+tar rvf "$TEST_ROOT/tar.tar" -C "$TEST_ROOT/tar_root" ./a/b/xyzzy ./bla
+path="$(nix flake prefetch --refresh --json "tarball+file://$TEST_ROOT/tar.tar" | jq -r .storePath)"
+[[ $(cat "$path/a/b/xyzzy") = xyzzy ]]
+[[ $(cat "$path/a/b/foo") = foo ]]
+[[ $(cat "$path/bla") = abc ]]