More rename detection tests

This includes tests for crlf changes, whitespace changes with the
default comparison and with the ignore whitespace comparison, and
more sensitivity checking for the comparison code.

tests-clar/diff/rename.c

@@ -153,11 +153,11 @@ void test_diff_rename__not_exact_match(void)
 	git_diff_find_options opts = GIT_DIFF_FIND_OPTIONS_INIT;
 	diff_expects exp;
 
-	/* Changes:
-	 *   songofseven.txt -> songofseven.txt (major rewrite, <20% match - split)
-	 *   sixserving.txt  -> sixserving.txt  (indentation change)
-	 *   sixserving.txt  -> ikeepsix.txt    (copy, add title, >80% match)
-	 *   sevencities.txt                    (no change)
+	/* == Changes =====================================================
+	 * songofseven.txt -> songofseven.txt (major rewrite, <20% match - split)
+	 * sixserving.txt  -> sixserving.txt  (indentation change)
+	 * sixserving.txt  -> ikeepsix.txt    (copy, add title, >80% match)
+	 * sevencities.txt                    (no change)
 	 */
 
 	old_tree = resolve_commit_oid_to_tree(g_repo, sha0);
@@ -204,11 +204,9 @@ void test_diff_rename__not_exact_match(void)
 
 	git_diff_list_free(diff);
 
-	/* git diff -M -C 2bc7f351d20b53f1c72c16c4b036e491c478c49a \
+	/* git diff -M -C \
+	 *          2bc7f351d20b53f1c72c16c4b036e491c478c49a \
 	 *          1c068dee5790ef1580cfc4cd670915b48d790084
-	 *
-	 * must not pass NULL for opts because it will pick up environment
-	 * values for "diff.renames" and test won't be consistent.
 	 */
 	cl_git_pass(git_diff_tree_to_tree(
 		&diff, g_repo, old_tree, new_tree, &diffopts));
@@ -250,18 +248,104 @@ void test_diff_rename__not_exact_match(void)
 
 	git_diff_list_free(diff);
 
-	/* Changes:
-	 *   songofseven.txt -> untimely.txt    (rename, convert to crlf)
-	 *   ikeepsix.txt    -> ikeepsix.txt    (reorder sections in file)
-	 *   sixserving.txt  -> sixserving.txt  (whitespace - not just indent)
-	 *   sevencities.txt -> songof7cities.txt (rename, small text changes)
+	/* == Changes =====================================================
+	 * songofseven.txt -> untimely.txt    (rename, convert to crlf)
+	 * ikeepsix.txt    -> ikeepsix.txt    (reorder sections in file)
+	 * sixserving.txt  -> sixserving.txt  (whitespace - not just indent)
+	 * sevencities.txt -> songof7cities.txt (rename, small text changes)
 	 */
 
 	git_tree_free(old_tree);
 	old_tree = new_tree;
 	new_tree = resolve_commit_oid_to_tree(g_repo, sha2);
 
-	/* moar tests needed */
+	cl_git_pass(git_diff_tree_to_tree(
+		&diff, g_repo, old_tree, new_tree, &diffopts));
+
+	/* git diff --no-renames \
+	 *          1c068dee5790ef1580cfc4cd670915b48d790084 \
+	 *          19dd32dfb1520a64e5bbaae8dce6ef423dfa2f13
+	 */
+	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_MODIFIED]);
+	cl_assert_equal_i(2, exp.file_status[GIT_DELTA_ADDED]);
+	cl_assert_equal_i(2, exp.file_status[GIT_DELTA_DELETED]);
+
+	/* git diff -M -C \
+	 *          1c068dee5790ef1580cfc4cd670915b48d790084 \
+	 *          19dd32dfb1520a64e5bbaae8dce6ef423dfa2f13
+	 */
+	cl_git_pass(git_diff_tree_to_tree(
+		&diff, g_repo, old_tree, new_tree, &diffopts));
+
+	opts.flags = GIT_DIFF_FIND_RENAMES | GIT_DIFF_FIND_COPIES;
+	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(4, exp.files);
+	cl_assert_equal_i(2, exp.file_status[GIT_DELTA_MODIFIED]);
+	cl_assert_equal_i(2, exp.file_status[GIT_DELTA_RENAMED]);
+
+	git_diff_list_free(diff);
+
+	/* git diff -M -C --find-copies-harder --break-rewrites \
+	 *          1c068dee5790ef1580cfc4cd670915b48d790084 \
+	 *          19dd32dfb1520a64e5bbaae8dce6ef423dfa2f13
+	 * with libgit2 default similarity comparison...
+	 */
+	cl_git_pass(git_diff_tree_to_tree(
+		&diff, g_repo, old_tree, new_tree, &diffopts));
+
+	opts.flags = GIT_DIFF_FIND_ALL;
+	cl_git_pass(git_diff_find_similar(diff, &opts));
+
+	/* the default match algorithm is going to find the internal
+	 * whitespace differences in the lines of sixserving.txt to be
+	 * significant enough that this will decide to split it into
+	 * an ADD and a DELETE
+	 */
+
+	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_MODIFIED]);
+	cl_assert_equal_i(1, exp.file_status[GIT_DELTA_ADDED]);
+	cl_assert_equal_i(1, exp.file_status[GIT_DELTA_DELETED]);
+	cl_assert_equal_i(2, exp.file_status[GIT_DELTA_RENAMED]);
+
+	git_diff_list_free(diff);
+
+	/* git diff -M -C --find-copies-harder --break-rewrites \
+	 *          1c068dee5790ef1580cfc4cd670915b48d790084 \
+	 *          19dd32dfb1520a64e5bbaae8dce6ef423dfa2f13
+	 * with ignore_space whitespace comparision
+	 */
+	cl_git_pass(git_diff_tree_to_tree(
+		&diff, g_repo, old_tree, new_tree, &diffopts));
+
+	opts.flags = GIT_DIFF_FIND_ALL | GIT_DIFF_FIND_IGNORE_WHITESPACE;
+	cl_git_pass(git_diff_find_similar(diff, &opts));
+
+	/* Ignoring whitespace, this should no longer split sixserver.txt */
+
+	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(2, exp.file_status[GIT_DELTA_MODIFIED]);
+	cl_assert_equal_i(2, exp.file_status[GIT_DELTA_RENAMED]);
+
+	git_diff_list_free(diff);
 
 	git_tree_free(old_tree);
 	git_tree_free(new_tree);