More diff rename tests; better split swap handling

This adds a couple more tests of different rename scenarios.

Also, this fixes a problem with the case where you have two
"split" deltas and the left half of one matches the right half of
the other.  That case was already being handled, but in the wrong
order in a way that could result in bad output.  Also, if the swap
also happened to put the other two halves into the correct place
(i.e. two files exchanged places with each other), then the second
delta was left with the SPLIT flag set when it really should be
cleared.

6 files changed, 139 insertions, 24 deletions

diff --git a/include/git2/diff.h b/include/git2/diff.h
index 6939f6a2e..0d4875b43 100644
--- a/include/git2/diff.h
+++ b/include/git2/diff.h
@@ -484,7 +484,7 @@ typedef struct {
 	unsigned int version;
 
 	/** Combination of git_diff_find_t values (default FIND_RENAMES) */
-	unsigned int flags;
+	uint32_t flags;
 
 	/** Similarity to consider a file renamed (default 50) */
 	uint16_t rename_threshold;
diff --git a/src/diff.h b/src/diff.h
index a9a543ecd..ac8ab2aed 100644
--- a/src/diff.h
+++ b/src/diff.h
@@ -44,6 +44,8 @@ enum {
 
 #define GIT_DIFF_FLAG__CLEAR_INTERNAL(F) (F) = ((F) & 0x00FFFF)
 
+#define GIT_DIFF__VERBOSE  (1 << 30)
+
 struct git_diff_list {
 	git_refcount     rc;
 	git_repository   *repo;
diff --git a/src/diff_tform.c b/src/diff_tform.c
index b481e64ce..0f4ecc7b5 100644
--- a/src/diff_tform.c
+++ b/src/diff_tform.c
@@ -828,22 +828,37 @@ int git_diff_find_similar(
 				if (similarity < (int)opts.rename_from_rewrite_threshold)
 					continue;
 
-				memcpy(&swap, &from->new_file, sizeof(swap));
+				memcpy(&swap, &to->new_file, sizeof(swap));
 
-				from->status = GIT_DELTA_RENAMED;
-				from->similarity = (uint32_t)similarity;
-				memcpy(&from->new_file, &to->new_file, sizeof(from->new_file));
-				if ((from->flags & GIT_DIFF_FLAG__TO_SPLIT) != 0) {
-					from->flags &= ~GIT_DIFF_FLAG__TO_SPLIT;
+				to->status = GIT_DELTA_RENAMED;
+				to->similarity = (uint32_t)similarity;
+				memcpy(&to->new_file, &from->new_file, sizeof(to->new_file));
+				if ((to->flags & GIT_DIFF_FLAG__TO_SPLIT) != 0) {
+					to->flags &= ~GIT_DIFF_FLAG__TO_SPLIT;
 					num_rewrites--;
 				}
 
-				memcpy(&to->new_file, &swap, sizeof(to->new_file));
-				if ((to->flags & GIT_DIFF_FLAG__TO_SPLIT) == 0) {
-					to->flags |= GIT_DIFF_FLAG__TO_SPLIT;
+				memcpy(&from->new_file, &swap, sizeof(from->new_file));
+				if ((from->flags & GIT_DIFF_FLAG__TO_SPLIT) == 0) {
+					from->flags |= GIT_DIFF_FLAG__TO_SPLIT;
 					num_rewrites++;
 				}
 
+				/* in the off chance that we've just swapped the new
+				 * element into the correct place, clear the SPLIT flag
+				 */
+				if (matches[matches[i].idx].idx == i &&
+					matches[matches[i].idx].similarity >
+					opts.rename_from_rewrite_threshold) {
+
+					from->status = GIT_DELTA_RENAMED;
+					from->similarity =
+						(uint32_t)matches[matches[i].idx].similarity;
+					matches[matches[i].idx].similarity = 0;
+					from->flags &= ~GIT_DIFF_FLAG__TO_SPLIT;
+					num_rewrites--;
+				}
+
 				num_updates++;
 			}
 		}
diff --git a/tests-clar/diff/diff_helpers.c b/tests-clar/diff/diff_helpers.c
index 75eda0520..4e23792a6 100644
--- a/tests-clar/diff/diff_helpers.c
+++ b/tests-clar/diff/diff_helpers.c
@@ -213,3 +213,8 @@ void diff_print(FILE *fp, git_diff_list *diff)
 {
 	cl_git_pass(git_diff_print_patch(diff, diff_print_cb, fp ? fp : stderr));
 }
+
+void diff_print_raw(FILE *fp, git_diff_list *diff)
+{
+	cl_git_pass(git_diff_print_raw(diff, diff_print_cb, fp ? fp : stderr));
+}
diff --git a/tests-clar/diff/diff_helpers.h b/tests-clar/diff/diff_helpers.h
index b39a69d1d..bb76d0076 100644
--- a/tests-clar/diff/diff_helpers.h
+++ b/tests-clar/diff/diff_helpers.h
@@ -65,4 +65,4 @@ extern int diff_foreach_via_iterator(
 	void *data);
 
 extern void diff_print(FILE *fp, git_diff_list *diff);
-
+extern void diff_print_raw(FILE *fp, git_diff_list *diff);
diff --git a/tests-clar/diff/rename.c b/tests-clar/diff/rename.c
index b4f9df713..1edf66e95 100644
--- a/tests-clar/diff/rename.c
+++ b/tests-clar/diff/rename.c
@@ -15,18 +15,6 @@ void test_diff_rename__cleanup(void)
 }
 
 /*
-static int debug_print(
-    const git_diff_delta *delta, const git_diff_range *range, char usage,
-    const char *line, size_t line_len, void *data)
-{
-    GIT_UNUSED(delta); GIT_UNUSED(range); GIT_UNUSED(usage);
-    GIT_UNUSED(line_len); GIT_UNUSED(data);
-    fputs(line, stderr);
-    return 0;
-}
-*/
-
-/*
  * Renames repo has:
  *
  * commit 31e47d8c1fa36d7f8d537b96158e3f024de0a9f2 -
@@ -539,7 +527,7 @@ void test_diff_rename__working_directory_changes(void)
 
 	/*
 	fprintf(stderr, "\n\n");
-    cl_git_pass(git_diff_print_raw(diff, debug_print, NULL));
+    diff_print_raw(stderr, diff);
 	*/
 
 	memset(&exp, 0, sizeof(exp));
@@ -613,3 +601,108 @@ void test_diff_rename__patch(void)
 	git_tree_free(old_tree);
 	git_tree_free(new_tree);
 }
+
+void test_diff_rename__file_exchange(void)
+{
+	git_buf c1 = GIT_BUF_INIT, c2 = GIT_BUF_INIT;
+	git_index *index;
+	git_tree *tree;
+	git_diff_list *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/untimely.txt"));
+	cl_git_pass(git_futils_readbuffer(&c2, "renames/songof7cities.txt"));
+	cl_git_pass(git_futils_writebuffer(&c1, "renames/songof7cities.txt", 0, 0));
+	cl_git_pass(git_futils_writebuffer(&c2, "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, "songof7cities.txt"));
+	cl_git_pass(git_index_add_bypath(index, "untimely.txt"));
+
+	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(2, exp.files);
+	cl_assert_equal_i(2, exp.file_status[GIT_DELTA_MODIFIED]);
+
+	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(2, exp.files);
+	cl_assert_equal_i(2, exp.file_status[GIT_DELTA_RENAMED]);
+
+	git_diff_list_free(diff);
+	git_tree_free(tree);
+	git_index_free(index);
+
+	git_buf_free(&c1);
+	git_buf_free(&c2);
+}
+
+void test_diff_rename__file_split(void)
+{
+	git_buf c1 = GIT_BUF_INIT, c2 = GIT_BUF_INIT;
+	git_index *index;
+	git_tree *tree;
+	git_diff_list *diff;
+	git_diff_options diffopts = GIT_DIFF_OPTIONS_INIT;
+	git_diff_find_options opts = GIT_DIFF_FIND_OPTIONS_INIT;
+	diff_expects exp;
+
+	/* put the first 2/3 of file into one new place
+	 * and the second 2/3 of file into another new place
+	 */
+	cl_git_pass(git_futils_readbuffer(&c1, "renames/songof7cities.txt"));
+	cl_git_pass(git_buf_set(&c2, c1.ptr, c1.size));
+	git_buf_truncate(&c1, c1.size * 2 / 3);
+	git_buf_consume(&c2, ((char *)c2.ptr) + (c2.size / 3));
+	cl_git_pass(git_futils_writebuffer(&c1, "renames/song_a.txt", 0, 0));
+	cl_git_pass(git_futils_writebuffer(&c2, "renames/song_b.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, "song_a.txt"));
+	cl_git_pass(git_index_add_bypath(index, "song_b.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(6, exp.files);
+	cl_assert_equal_i(2, exp.file_status[GIT_DELTA_ADDED]);
+	cl_assert_equal_i(4, 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(6, exp.files);
+	cl_assert_equal_i(2, exp.file_status[GIT_DELTA_COPIED]);
+	cl_assert_equal_i(4, exp.file_status[GIT_DELTA_UNMODIFIED]);
+
+	git_diff_list_free(diff);
+	git_tree_free(tree);
+	git_index_free(index);
+
+	git_buf_free(&c1);
+	git_buf_free(&c2);
+}