Add diff option to update index stat cache

When diff is scanning the working directory, if it finds a file
where it is not sure if the index entry matches the working dir,
it will recalculate the OID (which is pretty expensive).  This
adds a new flag to diff so that if the OID calculation finds that
the file actually has not changed (i.e. just the modified time was
altered or such), then it will refresh the stat cache in the index
so that future calls to diff will not have to check the oid again.

5 files changed, 128 insertions, 6 deletions

diff --git a/include/git2/diff.h b/include/git2/diff.h
index 273f471b6..afdb2c981 100644
--- a/include/git2/diff.h
+++ b/include/git2/diff.h
@@ -145,6 +145,13 @@ typedef enum {
 	 */
 	GIT_DIFF_ENABLE_FAST_UNTRACKED_DIRS = (1u << 14),
 
+	/** When diff finds a file in the working directory with stat
+	 * information different from the index, but the OID ends up being the
+	 * same, write the correct stat information into the index.  Note:
+	 * without this flag, diff will always leave the index untouched.
+	 */
+	GIT_DIFF_UPDATE_INDEX = (1u << 15),
+
 	/*
 	 * Options controlling how output will be generated
 	 */
diff --git a/src/checkout.c b/src/checkout.c
index d94cb0c7d..727911694 100644
--- a/src/checkout.c
+++ b/src/checkout.c
@@ -184,7 +184,7 @@ static bool checkout_is_workdir_modified(
 	if (baseitem->size && wditem->file_size != baseitem->size)
 		return true;
 
-	if (git_diff__oid_for_entry(&oid, data->diff, wditem) < 0)
+	if (git_diff__oid_for_entry(&oid, data->diff, wditem, NULL) < 0)
 		return false;
 
 	return (git_oid__cmp(&baseitem->id, &oid) != 0);
diff --git a/src/diff.c b/src/diff.c
index eae4543fc..caed8bf40 100644
--- a/src/diff.c
+++ b/src/diff.c
@@ -442,6 +442,14 @@ static int diff_list_apply_options(
 		diff->new_src = tmp_src;
 	}
 
+	/* Unset UPDATE_INDEX unless diffing workdir and index */
+	if (DIFF_FLAG_IS_SET(diff, GIT_DIFF_UPDATE_INDEX) &&
+		(!(diff->old_src == GIT_ITERATOR_TYPE_WORKDIR ||
+		   diff->new_src == GIT_ITERATOR_TYPE_WORKDIR) ||
+		 !(diff->old_src == GIT_ITERATOR_TYPE_INDEX ||
+		   diff->new_src == GIT_ITERATOR_TYPE_INDEX)))
+		diff->opts.flags &= ~GIT_DIFF_UPDATE_INDEX;
+
 	/* if ignore_submodules not explicitly set, check diff config */
 	if (diff->opts.ignore_submodules <= 0) {
 		const git_config_entry *entry;
@@ -523,11 +531,14 @@ int git_diff__oid_for_file(
 	entry.file_size = size;
 	entry.path = (char *)path;
 
-	return git_diff__oid_for_entry(out, diff, &entry);
+	return git_diff__oid_for_entry(out, diff, &entry, NULL);
 }
 
 int git_diff__oid_for_entry(
-	git_oid *out, git_diff *diff, const git_index_entry *src)
+	git_oid *out,
+	git_diff *diff,
+	const git_index_entry *src,
+	const git_oid *update_match)
 {
 	int error = 0;
 	git_buf full_path = GIT_BUF_INIT;
@@ -595,7 +606,16 @@ int git_diff__oid_for_entry(
 		git_filter_list_free(fl);
 	}
 
-	/* TODO: update index for entry if requested */
+	/* update index for entry if requested */
+	if (!error && update_match && git_oid_equal(out, update_match)) {
+		git_index *idx;
+
+		if (!(error = git_repository_index(&idx, diff->repo))) {
+			memcpy(&entry.id, out, sizeof(entry.id));
+			error = git_index_add(idx, &entry);
+			git_index_free(idx);
+		}
+ 	}
 
 	git_buf_free(&full_path);
 	return error;
@@ -776,7 +796,12 @@ static int maybe_modified(
 	 */
 	if (modified_uncertain && git_oid_iszero(&nitem->id)) {
 		if (git_oid_iszero(&noid)) {
-			if ((error = git_diff__oid_for_entry(&noid, diff, nitem)) < 0)
+			const git_oid *update_check =
+				DIFF_FLAG_IS_SET(diff, GIT_DIFF_UPDATE_INDEX) ?
+				&oitem->id : NULL;
+
+			if ((error = git_diff__oid_for_entry(
+					&noid, diff, nitem, update_check)) < 0)
 				return error;
 		}
 
@@ -1208,6 +1233,9 @@ int git_diff_index_to_workdir(
 			&b, repo, GIT_ITERATOR_DONT_AUTOEXPAND, pfx, pfx)
 	);
 
+	if (!error && DIFF_FLAG_IS_SET(*diff, GIT_DIFF_UPDATE_INDEX))
+		error = git_index_write(index);
+
 	return error;
 }
 
diff --git a/src/diff.h b/src/diff.h
index 8fa3c9b7b..b2b7dba70 100644
--- a/src/diff.h
+++ b/src/diff.h
@@ -97,7 +97,7 @@ extern int git_diff_delta__format_file_header(
 extern int git_diff__oid_for_file(
 	git_oid *out, git_diff *, const char *, uint16_t, git_off_t);
 extern int git_diff__oid_for_entry(
-	git_oid *out, git_diff *, const git_index_entry *entry);
+	git_oid *out, git_diff *, const git_index_entry *, const git_oid *update);
 
 extern int git_diff__from_iterators(
 	git_diff **diff_ptr,
diff --git a/tests/diff/workdir.c b/tests/diff/workdir.c
index 84c8866e0..9e4608e9d 100644
--- a/tests/diff/workdir.c
+++ b/tests/diff/workdir.c
@@ -1502,3 +1502,90 @@ void test_diff_workdir__with_stale_index(void)
 
 	git_index_free(idx);
 }
+
+static int touch_file(void *payload, git_buf *path)
+{
+	int fd;
+	char b;
+
+	GIT_UNUSED(payload);
+	if (git_path_isdir(path->ptr))
+		return 0;
+
+	cl_assert((fd = p_open(path->ptr, O_RDWR)) >= 0);
+	cl_assert_equal_i(1, p_read(fd, &b, 1));
+	cl_must_pass(p_lseek(fd, 0, SEEK_SET));
+	cl_must_pass(p_write(fd, &b, 1));
+	cl_must_pass(p_close(fd));
+
+	return 0;
+}
+
+static void basic_diff_status(git_diff **out, const git_diff_options *opts)
+{
+	diff_expects exp;
+
+	cl_git_pass(git_diff_index_to_workdir(out, g_repo, NULL, opts));
+
+	memset(&exp, 0, sizeof(exp));
+
+	cl_git_pass(git_diff_foreach(
+		*out, diff_file_cb, diff_hunk_cb, diff_line_cb, &exp));
+
+	cl_assert_equal_i(13, exp.files);
+	cl_assert_equal_i(0, exp.file_status[GIT_DELTA_ADDED]);
+	cl_assert_equal_i(4, exp.file_status[GIT_DELTA_DELETED]);
+	cl_assert_equal_i(4, exp.file_status[GIT_DELTA_MODIFIED]);
+	cl_assert_equal_i(1, exp.file_status[GIT_DELTA_IGNORED]);
+	cl_assert_equal_i(4, exp.file_status[GIT_DELTA_UNTRACKED]);
+}
+
+void test_diff_workdir__can_update_index(void)
+{
+	git_diff_options opts = GIT_DIFF_OPTIONS_INIT;
+	git_diff *diff = NULL;
+
+	g_repo = cl_git_sandbox_init("status");
+
+	/* touch all the files so stat times are different */
+	{
+		git_buf path = GIT_BUF_INIT;
+		cl_git_pass(git_buf_sets(&path, "status"));
+		cl_git_pass(git_path_direach(&path, 0, touch_file, NULL));
+		git_buf_free(&path);
+	}
+
+	opts.flags |= GIT_DIFF_INCLUDE_IGNORED | GIT_DIFF_INCLUDE_UNTRACKED;
+
+	basic_diff_status(&diff, &opts);
+#ifdef GIT_PERF
+	cl_assert_equal_sz(diff->stat_calls, 13 + 3);
+	cl_assert_equal_sz(diff->oid_calculations, 5);
+	cl_assert_equal_sz(diff->submodule_lookups, 1);
+#endif
+
+	git_diff_free(diff);
+
+	/* now allow diff to update stat cache */
+	opts.flags |= GIT_DIFF_UPDATE_INDEX;
+
+	basic_diff_status(&diff, &opts);
+#ifdef GIT_PERF
+	cl_assert_equal_sz(diff->stat_calls, 13 + 3);
+	cl_assert_equal_sz(diff->oid_calculations, 5);
+	cl_assert_equal_sz(diff->submodule_lookups, 1);
+#endif
+
+	git_diff_free(diff);
+
+	/* now if we do it again, we should see fewer OID calculations */
+
+	basic_diff_status(&diff, &opts);
+#ifdef GIT_PERF
+	cl_assert_equal_sz(diff->stat_calls, 13 + 3);
+	cl_assert_equal_sz(diff->oid_calculations, 0); /* Yay */
+	cl_assert_equal_sz(diff->submodule_lookups, 1);
+#endif
+
+	git_diff_free(diff);
+}