Fix bug making split deltas a COPIED targets

When FIND_COPIES is used in combination with BREAK_REWRITES for
rename detection, there was a bug where the split MODIFIED delta
was only used as a target for RENAME records and not for COPIED
records.  This fixes that, converting the split into a pair of
DELETED and COPIED deltas when that circumstance arises.

2 files changed, 79 insertions, 16 deletions

diff --git a/src/diff_tform.c b/src/diff_tform.c
index 28a9cc70d..3b5575d28 100644
--- a/src/diff_tform.c
+++ b/src/diff_tform.c
@@ -366,12 +366,28 @@ static int normalize_find_opts(
 	return 0;
 }
 
+static int insert_delete_side_of_split(
+	git_diff *diff, git_vector *onto, const git_diff_delta *delta)
+{
+	/* make new record for DELETED side of split */
+	git_diff_delta *deleted = diff_delta__dup(delta, &diff->pool);
+	GITERR_CHECK_ALLOC(deleted);
+
+	deleted->status = GIT_DELTA_DELETED;
+	deleted->nfiles = 1;
+	memset(&deleted->new_file, 0, sizeof(deleted->new_file));
+	deleted->new_file.path = deleted->old_file.path;
+	deleted->new_file.flags |= GIT_DIFF_FLAG_VALID_OID;
+
+	return git_vector_insert(onto, deleted);
+}
+
 static int apply_splits_and_deletes(
 	git_diff *diff, size_t expected_size, bool actually_split)
 {
 	git_vector onto = GIT_VECTOR_INIT;
 	size_t i;
-	git_diff_delta *delta, *deleted;
+	git_diff_delta *delta;
 
 	if (git_vector_init(&onto, expected_size, git_diff_delta__cmp) < 0)
 		return -1;
@@ -384,17 +400,7 @@ static int apply_splits_and_deletes(
 		if ((delta->flags & GIT_DIFF_FLAG__TO_SPLIT) != 0 && actually_split) {
 			delta->similarity = 0;
 
-			/* make new record for DELETED side of split */
-			if (!(deleted = diff_delta__dup(delta, &diff->pool)))
-				goto on_error;
-
-			deleted->status = GIT_DELTA_DELETED;
-			deleted->nfiles = 1;
-			memset(&deleted->new_file, 0, sizeof(deleted->new_file));
-			deleted->new_file.path = deleted->old_file.path;
-			deleted->new_file.flags |= GIT_DIFF_FLAG_VALID_OID;
-
-			if (git_vector_insert(&onto, deleted) < 0)
+			if (insert_delete_side_of_split(diff, &onto, delta) < 0)
 				goto on_error;
 
 			if (diff->new_src == GIT_ITERATOR_TYPE_WORKDIR)
@@ -1058,10 +1064,7 @@ find_best_matches:
 			}
 		}
 
-		else if (delta_is_new_only(tgt)) {
-			if (!FLAG_SET(&opts, GIT_DIFF_FIND_COPIES))
-				continue;
-
+		else if (FLAG_SET(&opts, GIT_DIFF_FIND_COPIES)) {
 			if (tgt2src_copy[t].similarity < opts.copy_threshold)
 				continue;
 
@@ -1069,10 +1072,21 @@ find_best_matches:
 			best_match = &tgt2src_copy[t];
 			src = GIT_VECTOR_GET(&diff->deltas, best_match->idx);
 
+			if (delta_is_split(tgt)) {
+				error = insert_delete_side_of_split(diff, &diff->deltas, tgt);
+				if (error < 0)
+					goto cleanup;
+				num_rewrites--;
+			}
+
+			if (!delta_is_split(tgt) && !delta_is_new_only(tgt))
+				continue;
+
 			tgt->status     = GIT_DELTA_COPIED;
 			tgt->similarity = best_match->similarity;
 			tgt->nfiles     = 2;
 			memcpy(&tgt->old_file, &src->old_file, sizeof(tgt->old_file));
+			tgt->flags &= ~GIT_DIFF_FLAG__TO_SPLIT;
 
 			num_updates++;
 		}
diff --git a/tests/diff/rename.c b/tests/diff/rename.c
index 42bb65aa8..7f033936e 100644
--- a/tests/diff/rename.c
+++ b/tests/diff/rename.c
@@ -1284,3 +1284,52 @@ void test_diff_rename__rewrite_on_single_file(void)
 	git_diff_free(diff);
 	git_index_free(index);
 }
+
+void test_diff_rename__can_find_copy_to_split(void)
+{
+	git_buf c1 = GIT_BUF_INIT;
+	git_index *index;
+	git_tree *tree;
+	git_diff *diff;
+	git_diff_options diffopts = GIT_DIFF_OPTIONS_INIT;
+	git_diff_find_options opts = GIT_DIFF_FIND_OPTIONS_INIT;
+	diff_expects exp;
+
+	cl_git_pass(git_futils_readbuffer(&c1, "renames/songof7cities.txt"));
+	cl_git_pass(git_futils_writebuffer(&c1, "renames/untimely.txt", 0, 0));
+
+	cl_git_pass(
+		git_revparse_single((git_object **)&tree, g_repo, "HEAD^{tree}"));
+
+	cl_git_pass(git_repository_index(&index, g_repo));
+	cl_git_pass(git_index_read_tree(index, tree));
+	cl_git_pass(git_index_add_bypath(index, "untimely.txt"));
+
+	diffopts.flags = GIT_DIFF_INCLUDE_UNMODIFIED;
+
+	cl_git_pass(git_diff_tree_to_index(&diff, g_repo, tree, index, &diffopts));
+
+	memset(&exp, 0, sizeof(exp));
+	cl_git_pass(git_diff_foreach(
+		diff, diff_file_cb, diff_hunk_cb, diff_line_cb, &exp));
+	cl_assert_equal_i(4, exp.files);
+	cl_assert_equal_i(1, exp.file_status[GIT_DELTA_MODIFIED]);
+	cl_assert_equal_i(3, exp.file_status[GIT_DELTA_UNMODIFIED]);
+
+	opts.flags = GIT_DIFF_FIND_ALL;
+	cl_git_pass(git_diff_find_similar(diff, &opts));
+
+	memset(&exp, 0, sizeof(exp));
+	cl_git_pass(git_diff_foreach(
+		diff, diff_file_cb, diff_hunk_cb, diff_line_cb, &exp));
+	cl_assert_equal_i(5, exp.files);
+	cl_assert_equal_i(1, exp.file_status[GIT_DELTA_DELETED]);
+	cl_assert_equal_i(1, exp.file_status[GIT_DELTA_COPIED]);
+	cl_assert_equal_i(3, exp.file_status[GIT_DELTA_UNMODIFIED]);
+
+	git_diff_free(diff);
+	git_tree_free(tree);
+	git_index_free(index);
+
+	git_buf_free(&c1);
+}