hare

commit 2042af21047d33c551a7f85c240fdf50750ef6f7
parent ea5dfc8e136e4222fdeabfa3a65f998a7466dc41
Author: Bor Grošelj Simić <bgs@turminal.net>
Date:   Tue, 19 Apr 2022 03:57:10 +0200

strings::{l,r,}trim: trim whitespace with bytes::

by doing so we avoid utf8 decoding for this common case

Signed-off-by: Bor Grošelj Simić <bgs@turminal.net>

Diffstat:
M
strings/trim.ha
 | 
9
++++++---

1 file changed, 6 insertions(+), 3 deletions(-)
diff --git a/strings/trim.ha b/strings/trim.ha
@@ -3,16 +3,18 @@
 // (c) 2021 Eyal Sawady <ecs@d2evs.net>
 // (c) 2021 Miccah Castorina <contact@miccah.io>
 // (c) 2021 Sudipto Mallick <smlckz@disroot.org>
+use bytes;
 use encoding::utf8;
 
-const whitespace: [_]rune = [' ', '\n', '\t', '\r'];
+const whitespace: [_]u8 = [' ', '\n', '\t', '\r'];
 
 // Returns a string (borrowed from given input string) after trimming off of
 // the start of the input string the characters in the given list of runes. If
 // no runes are given, returns the string with leading whitespace stripped off.
 export fn ltrim(input: str, trim: rune...) str = {
 	if (len(trim) == 0) {
-		trim = whitespace;
+		let input = toutf8(input);
+		return fromutf8_unsafe(bytes::ltrim(input, whitespace...));
 	};
 	let it = iter(input);
 	for (true) {
@@ -42,7 +44,8 @@ export fn ltrim(input: str, trim: rune...) str = {
 // runes are given, returns the string with trailing whitespace stripped off.
 export fn rtrim(input: str, trim: rune...) str = {
 	if (len(trim) == 0) {
-		trim = whitespace;
+		let input = toutf8(input);
+		return fromutf8_unsafe(bytes::rtrim(input, whitespace...));
 	};
 	let it = riter(input);
 	for (true) {