encoding.utf8: fix is_punct func (fix #22646) (#22647)

vlib/encoding/utf8/utf8_is_punct_test.v

@@ -0,0 +1,33 @@
+module main
+
+import encoding.utf8
+
+fn test_true() {
+	assert utf8.is_punct('!', 0)
+	assert utf8.is_rune_punct(`!`)
+	assert utf8.is_punct('.', 0)
+	assert utf8.is_rune_punct(`.`)
+	assert utf8.is_punct(',', 0)
+	assert utf8.is_rune_punct(`,`)
+	assert utf8.is_punct(';', 0)
+	assert utf8.is_rune_punct(`;`)
+	assert utf8.is_punct("'", 0)
+	assert utf8.is_rune_punct(`'`)
+	assert utf8.is_punct(',', 0)
+	assert utf8.is_rune_punct(`,`)
+	assert utf8.is_punct('/', 0)
+	assert utf8.is_rune_punct(`/`)
+	assert utf8.is_punct('*', 0)
+	assert utf8.is_rune_punct(`*`)
+}
+
+fn test_false() {
+	assert !utf8.is_punct('a', 0)
+	assert !utf8.is_rune_punct(`a`)
+	assert !utf8.is_punct('ç', 0)
+	assert !utf8.is_rune_punct(`ç`)
+	assert !utf8.is_punct('á', 0)
+	assert !utf8.is_rune_punct(`á`)
+	assert !utf8.is_punct('-', 0)
+	assert !utf8.is_rune_punct(`-`)
+}

vlib/encoding/utf8/utf8_util.v

@@ -190,7 +190,7 @@ pub fn is_uchar_punct(uchar int) bool {
 
 // is_rune_punct return true if the input unicode is a western unicode punctuation
 pub fn is_rune_punct(r rune) bool {
-	return find_punct_in_table(r, unicode_punct_western) != 0
+	return find_punct_in_table(r, unicode_punct_western) != rune(-1)
 }
 
 // Global
@@ -209,7 +209,7 @@ pub fn is_uchar_global_punct(uchar int) bool {
 
 // is_rune_global_punct return true if the input unicode is a global unicode punctuation
 pub fn is_rune_global_punct(r rune) bool {
-	return find_punct_in_table(r, unicode_punct) != 0
+	return find_punct_in_table(r, unicode_punct) != rune(-1)
 }
 
 // Private functions
@@ -553,7 +553,6 @@ fn find_punct_in_table(in_code rune, in_table []rune) rune {
 	mut x := rune(0)
 
 	for {
-		index = (first_index + last_index) >> 1
 		x = in_table[index]
 		// C.printf("(%d..%d) index:%d base[%08x]==>[%08x]\n",first_index,last_index,index,in_code,x)
 
@@ -568,9 +567,10 @@ fn find_punct_in_table(in_code rune, in_table []rune) rune {
 		if (last_index - first_index) <= 1 {
 			break
 		}
+		index = (first_index + last_index) >> 1
 	}
 
-	return 0
+	return -1
 }
 
 // Unicode punctuation chars