pkg/vcs: fix repetitive change handling in BaseForDiff

The current implementation runs into problems if we modify the same file
several times in the patch series since the sequential blob hashes will
not be found in the reproducer.

Fix it by moving the check for already processed files up.

2 files changed, 31 insertions, 6 deletions

diff --git a/pkg/vcs/git.go b/pkg/vcs/git.go
index f1311e312..efcb64470 100644
--- a/pkg/vcs/git.go
+++ b/pkg/vcs/git.go
@@ -805,6 +805,10 @@ func (git Git) BaseForDiff(diff []byte, tracer debugtracer.DebugTracer) ([]*Base
 		if _, ignore := ignoreFiles[file.Name]; ignore {
 			continue
 		}
+		if _, ok := nameToHash[file.Name]; ok {
+			// We only care about the first occurrence of a file in the diff series.
+			continue
+		}
 		if ok, err := git.verifyHash(file.LeftHash); err != nil {
 			return nil, fmt.Errorf("hash verification failed: %w", err)
 		} else if !ok {
@@ -812,12 +816,8 @@ func (git Git) BaseForDiff(diff []byte, tracer debugtracer.DebugTracer) ([]*Base
 			tracer.Logf("unknown object %s, stopping base commit search", file.LeftHash)
 			return nil, nil
 		}
-		if _, ok := nameToHash[file.Name]; !ok {
-			// If the diff is actually a concatenation of several diffs, we only
-			// want to remember the first left side hash for each file.
-			fileNames = append(fileNames, file.Name)
-			nameToHash[file.Name] = file.LeftHash
-		}
+		fileNames = append(fileNames, file.Name)
+		nameToHash[file.Name] = file.LeftHash
 		args = append(args, "--find-object="+file.LeftHash)
 	}
 	tracer.Logf("extracted %d left blob hashes", len(nameToHash))
diff --git a/pkg/vcs/git_test.go b/pkg/vcs/git_test.go
index f2fd9e059..d867ed692 100644
--- a/pkg/vcs/git_test.go
+++ b/pkg/vcs/git_test.go
@@ -599,6 +599,31 @@ index fa49b07..01c887f 100644
 		require.NoError(t, err)
 		require.Nil(t, base)
 	})
+	t.Run("multiple modifications", func(t *testing.T) {
+		map1, _ := repo.repo.fileHashes(commit3.Hash, []string{"a.txt"})
+
+		twoDiffs := []byte(fmt.Sprintf(`diff --git a/a.txt b/a.txt
+index %s..1111111 100644
+--- a/a.txt
++++ b/a.txt
+@@ -1 +1 @@
+-update a.txt
++update a.txt again
+diff --git a/a.txt b/a.txt
+index 1111111..2222222 100644
+--- a/a.txt
++++ b/a.txt
+@@ -1 +1 @@
+-update a.txt again
++update a.txt again and again
+`, map1["a.txt"]))
+
+		base, err := repo.repo.BaseForDiff(twoDiffs, &debugtracer.TestTracer{T: t})
+		require.NoError(t, err)
+		require.Len(t, base, 1)
+
+		assert.Equal(t, commit3.Hash, base[0].Hash)
+	})
 }
 
 func TestBaseForDiffMerge(t *testing.T) {