vieter-v
/
v
2
0
Fork 0
Code Issues Pull Requests Releases Activity

utf8: fix a panic in utf8.len("pippo"), mark with `unsafe{}` direct pointer manipulations

pull/5941/head
Delyan Angelov 2020-07-22 21:15:07 +03:00
parent 4096432a1d
commit 949ed90b51
1 changed files with 35 additions and 19 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

54
vlib/encoding/utf8/utf8_util.v
View File

@ -23,12 +23,12 @@ pub fn len(s string) int {
mut index := 0
for {
ch_len := utf8util_char_len(s.str[index])
ch_len := utf8util_char_len(s[index])
index += ch_len
if index > s.len {
count++
if index >= s.len {
break
}
count++
}
return count
}
@ -43,14 +43,14 @@ pub fn get_uchar(s string, index int) int {
mut res := 0
mut ch_len := 0
if s.len > 0 {
ch_len = utf8util_char_len(s.str[index])
ch_len = utf8util_char_len(s[index])
if ch_len == 1 {
return u16(s.str[index])
return u16(s[index])
}if ch_len > 1 && ch_len < 5{
mut lword := 0
for i:=0; i < ch_len ; i++ {
lword = (lword << 8 ) | int( s.str[index + i] )
lword = (lword << 8 ) | int( s[index + i] )
}
// 2 byte utf-8
@ -168,20 +168,24 @@ fn up_low(s string, upper_flag bool) string {
mut str_res := malloc(s.len + 1)
for {
ch_len := utf8util_char_len(s.str[index])
ch_len := utf8util_char_len(s[index])
if ch_len == 1 {
if upper_flag==true {
str_res[index] = byte(C.toupper(s.str[index]))
unsafe {
str_res[index] = byte(C.toupper(s.str[index]))
}
}else{
str_res[index] = byte(C.tolower(s.str[index]))
unsafe {
str_res[index] = byte(C.tolower(s.str[index]))
}
}
}
else if ch_len > 1 && ch_len < 5{
mut lword := 0
for i:=0; i < ch_len ; i++ {
lword = (lword << 8 ) | int( s.str[index + i] )
lword = (lword << 8 ) | int( s[index + i] )
}
//C.printf(" #%d (%x) ", index, lword)
@ -215,7 +219,9 @@ fn up_low(s string, upper_flag bool) string {
// char not in table, no need of conversion
if ch_index == 0 {
for i in 0..ch_len {
str_res[index + i] = s.str[index + i]
unsafe {
str_res[index + i] = s[index + i]
}
}
//C.printf("\n")
}else{
@ -227,8 +233,10 @@ fn up_low(s string, upper_flag bool) string {
ch1 := byte( (tab_char >> 0) & 0x3f ) | 0x80 /*10xx xxxx*/
//C.printf("[%02x%02x] \n",ch0,ch1)
str_res[ index + 0 ] = ch0
str_res[ index + 1 ] = ch1
unsafe {
str_res[ index + 0 ] = ch0
str_res[ index + 1 ] = ch1
}
//****************************************************************
// BUG: doesn't compile, workaround use shitf to right of 0 bit
@ -242,16 +250,20 @@ fn up_low(s string, upper_flag bool) string {
ch2 := byte( (tab_char >> 0) & 0x3f ) | 0x80 /*10xx xxxx*/
//C.printf("[%02x%02x%02x] \n",ch0,ch1,ch2)
str_res[index + 0 ] = ch0
str_res[index + 1 ] = ch1
str_res[index + 2 ] = ch2
unsafe {
str_res[index + 0 ] = ch0
str_res[index + 1 ] = ch1
str_res[index + 2 ] = ch2
}
}
// TODO: write if needed
else if ch_len == 4 {
// place holder!!
// at the present time simply copy the utf8 char
for i in 0..ch_len {
str_res[index + i] = s.str[index + i]
unsafe {
str_res[index + i] = s[index + i]
}
}
}
}
@ -260,7 +272,9 @@ fn up_low(s string, upper_flag bool) string {
// other cases, just copy the string
else{
for i in 0..ch_len {
str_res[index + i] = s.str[index + i]
unsafe {
str_res[index + i] = s[index + i]
}
}
}
@ -273,7 +287,9 @@ fn up_low(s string, upper_flag bool) string {
}
// for c compatibility set the ending 0
str_res[index]=0
unsafe {
str_res[index] = 0
}
//C.printf("str_res: %s\n--------------\n",str_res)