encoding.utf8: add is_space (#15847)

vlib/encoding/utf8/utf8_tables.v

@@ -1132,6 +1132,20 @@ const letter_table = RangeTable{
 	latin_offset: 6
 }
 
+const white_space_table = RangeTable{
+	r16: [
+		Range16{0x0009, 0x000d, 1},
+		Range16{0x0020, 0x0085, 101},
+		Range16{0x00a0, 0x1680, 5600},
+		Range16{0x2000, 0x200a, 1},
+		Range16{0x2028, 0x2029, 1},
+		Range16{0x202f, 0x205f, 48},
+		Range16{0x3000, 0x3000, 1},
+	]
+	r32: []
+	latin_offset: 2
+}
+
 struct RangeTable {
 pub:
 	r16          []Range16

vlib/encoding/utf8/utf8_util.v

@@ -160,6 +160,24 @@ pub fn is_letter(r rune) bool {
 	return is_excluding_latin(letter_table, r)
 }
 
+// is_space returns true if the rune is character in unicode category Z with property white space or the following character set:
+// ```
+// `\t`, `\n`, `\v`, `\f`, `\r`, ` `, 0x85 (NEL), 0xA0 (NBSP)
+// ```
+pub fn is_space(r rune) bool {
+	if r <= max_latin_1 {
+		match r {
+			`\t`, `\n`, `\v`, `\f`, `\r`, ` `, 0x85, 0xA0 {
+				return true
+			}
+			else {
+				return false
+			}
+		}
+	}
+	return is_excluding_latin(white_space_table, r)
+}
+
 // is_uchar_punct return true if the input unicode is a western unicode punctuation
 pub fn is_uchar_punct(uchar int) bool {
 	return find_punct_in_table(uchar, utf8.unicode_punct_western) != 0

vlib/encoding/utf8/utf8_util_test.v

@@ -91,3 +91,17 @@ fn test_is_letter() {
 	assert utf8.is_letter(`ȶ`) == true
 	assert utf8.is_letter(`ȹ`) == true
 }
+
+fn test_is_space() {
+	for ra in `a` .. `z` {
+		assert utf8.is_space(ra) == false
+	}
+
+	for ra in `A` .. `Z` {
+		assert utf8.is_space(ra) == false
+	}
+
+	assert utf8.is_space(`\u202f`) == true
+	assert utf8.is_space(`\u2009`) == true
+	assert utf8.is_space(`\u00A0`) == true
+}