static void TestNextPrevNonCharacters() {
/* test non-characters */
static const uint8_t nonChars[]={
0xef, 0xb7, 0x90, /* U+fdd0 */
0xef, 0xbf, 0xbf, /* U+feff */
0xf0, 0x9f, 0xbf, 0xbe, /* U+1fffe */
0xf0, 0xbf, 0xbf, 0xbf, /* U+3ffff */
0xf4, 0x8f, 0xbf, 0xbe /* U+10fffe */
};
UChar32 ch;
int32_t idx;
for(idx=0; idx<(int32_t)sizeof(nonChars);) {
U8_NEXT(nonChars, idx, sizeof(nonChars), ch);
if(!U_IS_UNICODE_NONCHAR(ch)) {
log_err("U8_NEXT(before %d) failed to read a non-character\n", idx);
}
}
for(idx=(int32_t)sizeof(nonChars); idx>0;) {
U8_PREV(nonChars, 0, idx, ch);
if(!U_IS_UNICODE_NONCHAR(ch)) {
log_err("U8_PREV(at %d) failed to read a non-character\n", idx);
}
}
}
/*
* Handle the non-inline part of the U8_NEXT() and U8_NEXT_FFFD() macros
* and their obsolete sibling UTF8_NEXT_CHAR_SAFE().
*
* U8_NEXT() supports NUL-terminated strings indicated via length<0.
*
* The "strict" parameter controls the error behavior:
* <0 "Safe" behavior of U8_NEXT():
* -1: All illegal byte sequences yield U_SENTINEL=-1.
* -2: Same as -1, except for lenient treatment of surrogate code points as legal.
* Some implementations use this for roundtripping of
* Unicode 16-bit strings that are not well-formed UTF-16, that is, they
* contain unpaired surrogates.
* -3: All illegal byte sequences yield U+FFFD.
* 0 Obsolete "safe" behavior of UTF8_NEXT_CHAR_SAFE(..., FALSE):
* All illegal byte sequences yield a positive code point such that this
* result code point would be encoded with the same number of bytes as
* the illegal sequence.
* >0 Obsolete "strict" behavior of UTF8_NEXT_CHAR_SAFE(..., TRUE):
* Same as the obsolete "safe" behavior, but non-characters are also treated
* like illegal sequences.
*
* Note that a UBool is the same as an int8_t.
*/
U_CAPI UChar32 U_EXPORT2
utf8_nextCharSafeBody(const uint8_t *s, int32_t *pi, int32_t length, UChar32 c, UBool strict) {
int32_t i=*pi;
uint8_t count=U8_COUNT_TRAIL_BYTES(c);
U_ASSERT(count <= 5); /* U8_COUNT_TRAIL_BYTES returns value 0...5 */
if(i+count<=length || length<0) {
uint8_t trail;
U8_MASK_LEAD_BYTE(c, count);
/* support NUL-terminated strings: do not read beyond the first non-trail byte */
switch(count) {
/* each branch falls through to the next one */
case 0:
/* count==0 for illegally leading trail bytes and the illegal bytes 0xfe and 0xff */
case 5:
case 4:
/* count>=4 is always illegal: no more than 3 trail bytes in Unicode's UTF-8 */
break;
case 3:
trail=s[i++]-0x80;
c=(c<<6)|trail;
/* c>=0x110 would result in code point>0x10ffff, outside Unicode */
if(c>=0x110 || trail>0x3f) { break; }
case 2:
trail=s[i++]-0x80;
c=(c<<6)|trail;
/*
* test for a surrogate d800..dfff unless we are lenient:
* before the last (c<<6), a surrogate is c=360..37f
*/
if(((c&0xffe0)==0x360 && strict!=-2) || trail>0x3f) { break; }
case 1:
trail=s[i++]-0x80;
c=(c<<6)|trail;
if(trail>0x3f) { break; }
/* correct sequence - all trail bytes have (b7..b6)==(10) */
if(c>=utf8_minLegal[count] &&
/* strict: forbid non-characters like U+fffe */
(strict<=0 || !U_IS_UNICODE_NONCHAR(c))) {
*pi=i;
return c;
}
/* no default branch to optimize switch() - all values are covered */
}
} else {
/* too few bytes left */
count=length-i;
}
/* error handling */
i=*pi;
while(count>0 && U8_IS_TRAIL(s[i])) {
++i;
--count;
}
c=errorValue(i-*pi, strict);
*pi=i;
return c;
}
static void
getAgeIfAssigned(UChar32 c, UVersionInfo age) {
if(u_isdefined(c)) {
u_charAge(c, age);
} else if(U_IS_UNICODE_NONCHAR(c)) {
age[0]=0;
age[1]=0;
age[2]=0;
age[3]=1;
} else {
memset(age, 0, 4);
}
}
static void TestNextPrevChar(){
static const uint8_t input[]={0x61, 0xf0, 0x90, 0x90, 0x81, 0xc0, 0x80, 0xfd, 0xbe, 0xc2, 0x61, 0x81, 0x90, 0x90, 0xf0, 0x00};
static const UChar32 result[]={
/*next_unsafe next_safe_ns next_safe_s prev_unsafe prev_safe_ns prev_safe_s*/
0x0061, 0x0061, 0x0061, 0x0000, 0x0000, 0x0000,
0x10401, 0x10401, 0x10401, 0xf0, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1,
0x90, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1, 0x2841410, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1,
0x90, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1, 0xa1050, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1,
0x81, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1, 0x2841, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1,
0x00, UTF8_ERROR_VALUE_2, UTF8_ERROR_VALUE_2, 0x61, 0x61, 0x61,
0x80, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1, 0xc2, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1,
0xfd, UTF8_ERROR_VALUE_2, UTF8_ERROR_VALUE_2, 0x77e, UTF8_ERROR_VALUE_2, UTF8_ERROR_VALUE_2,
0xbe, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1, 0xfd, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1,
0xa1, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1, 0x00, UTF8_ERROR_VALUE_2, UTF8_ERROR_VALUE_2,
0x61, 0x61, 0x61, 0xc0, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1,
0x81, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1, 0x10401, 0x10401, 0x10401,
0x90, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1, 0x410, UTF_ERROR_VALUE, UTF_ERROR_VALUE,
0x90, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1, 0x410, UTF8_ERROR_VALUE_2, UTF8_ERROR_VALUE_2,
0x0840, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1, 0xf0, UTF8_ERROR_VALUE_1, UTF8_ERROR_VALUE_1,
0x0000, 0x0000, 0x0000, 0x0061, 0x0061, 0x0061
};
static const int32_t movedOffset[]={
/*next_unsafe next_safe_ns next_safe_s prev_unsafe prev_safe_ns prev_safe_s*/
1, 1, 1, 15, 15, 15,
5, 5, 5, 14, 14 , 14,
3, 3, 3, 9, 13, 13,
4, 4, 4, 9, 12, 12,
5, 5, 5, 9, 11, 11,
7, 7, 7, 10, 10, 10,
7, 7, 7, 9, 9, 9,
8, 9, 9, 7, 7, 7,
9, 9, 9, 7, 7, 7,
11, 10, 10, 5, 5, 5,
11, 11, 11, 5, 5, 5,
12, 12, 12, 1, 1, 1,
13, 13, 13, 1, 1, 1,
14, 14, 14, 1, 1, 1,
14, 15, 15, 1, 1, 1,
14, 16, 16, 0, 0, 0,
};
UChar32 c=0x0000;
uint32_t i=0;
uint32_t offset=0;
int32_t setOffset=0;
for(offset=0; offset<sizeof(input); offset++){
if (offset < sizeof(input) - 2) { /* Can't have it go off the end of the array based on input */
setOffset=offset;
UTF8_NEXT_CHAR_UNSAFE(input, setOffset, c);
if(setOffset != movedOffset[i]){
log_err("ERROR: UTF8_NEXT_CHAR_UNSAFE failed to move the offset correctly at %d\n ExpectedOffset:%d Got %d\n",
offset, movedOffset[i], setOffset);
}
if(c != result[i]){
log_err("ERROR: UTF8_NEXT_CHAR_UNSAFE failed for offset=%ld. Expected:%lx Got:%lx\n", offset, result[i], c);
}
setOffset=offset;
U8_NEXT_UNSAFE(input, setOffset, c);
if(setOffset != movedOffset[i]){
log_err("ERROR: U8_NEXT_UNSAFE failed to move the offset correctly at %d\n ExpectedOffset:%d Got %d\n",
offset, movedOffset[i], setOffset);
}
if(c != result[i]){
log_err("ERROR: U8_NEXT_UNSAFE failed for offset=%ld. Expected:%lx Got:%lx\n", offset, result[i], c);
}
}
setOffset=offset;
UTF8_NEXT_CHAR_SAFE(input, setOffset, sizeof(input), c, FALSE);
if(setOffset != movedOffset[i+1]){
log_err("ERROR: UTF8_NEXT_CHAR_SAFE failed to move the offset correctly at %d\n ExpectedOffset:%d Got %d\n",
offset, movedOffset[i+1], setOffset);
}
if(c != result[i+1]){
log_err("ERROR: UTF8_NEXT_CHAR_SAFE failed for input=%ld. Expected:%lx Got:%lx\n", offset, result[i+1], c);
}
setOffset=offset;
U8_NEXT(input, setOffset, sizeof(input), c);
if(setOffset != movedOffset[i+1]){
log_err("ERROR: U8_NEXT failed to move the offset correctly at %d\n ExpectedOffset:%d Got %d\n",
offset, movedOffset[i+1], setOffset);
}
if(UTF_IS_ERROR(result[i+1]) ? c >= 0 : c != result[i+1]){
log_err("ERROR: U8_NEXT failed for input=%ld. Expected:%lx Got:%lx\n", offset, result[i+1], c);
}
setOffset=offset;
UTF8_NEXT_CHAR_SAFE(input, setOffset, sizeof(input), c, TRUE);
if(setOffset != movedOffset[i+1]){
log_err("ERROR: UTF8_NEXT_CHAR_SAFE(strict) failed to move the offset correctly at %d\n ExpectedOffset:%d Got %d\n",
offset, movedOffset[i+2], setOffset);
}
if(c != result[i+2]){
log_err("ERROR: UTF8_NEXT_CHAR_SAFE(strict) failed for input=%ld. Expected:%lx Got:%lx\n", offset, result[i+2], c);
}
i=i+6;
}
i=0;
for(offset=sizeof(input); offset > 0; --offset){
setOffset=offset;
UTF8_PREV_CHAR_UNSAFE(input, setOffset, c);
if(setOffset != movedOffset[i+3]){
log_err("ERROR: UTF8_PREV_CHAR_UNSAFE failed to move the offset correctly at %d\n ExpectedOffset:%d Got %d\n",
offset, movedOffset[i+3], setOffset);
}
if(c != result[i+3]){
log_err("ERROR: UTF8_PREV_CHAR_UNSAFE failed for offset=%ld. Expected:%lx Got:%lx\n", offset, result[i+3], c);
}
setOffset=offset;
UTF8_PREV_CHAR_SAFE(input, 0, setOffset, c, FALSE);
if(setOffset != movedOffset[i+4]){
log_err("ERROR: UTF8_PREV_CHAR_SAFE failed to move the offset correctly at %d\n ExpectedOffset:%d Got %d\n",
offset, movedOffset[i+4], setOffset);
}
if(c != result[i+4]){
log_err("ERROR: UTF8_PREV_CHAR_SAFE failed for input=%ld. Expected:%lx Got:%lx\n", offset, result[i+4], c);
}
setOffset=offset;
U8_PREV(input, 0, setOffset, c);
if(setOffset != movedOffset[i+4]){
log_err("ERROR: U8_PREV failed to move the offset correctly at %d\n ExpectedOffset:%d Got %d\n",
offset, movedOffset[i+4], setOffset);
}
if(UTF_IS_ERROR(result[i+4]) ? c >= 0 : c != result[i+4]){
log_err("ERROR: U8_PREV failed for input=%ld. Expected:%lx Got:%lx\n", offset, result[i+4], c);
}
setOffset=offset;
UTF8_PREV_CHAR_SAFE(input, 0, setOffset, c, TRUE);
if(setOffset != movedOffset[i+5]){
log_err("ERROR: UTF8_PREV_CHAR_SAFE(strict) failed to move the offset correctly at %d\n ExpectedOffset:%d Got %d\n",
offset, movedOffset[i+5], setOffset);
}
if(c != result[i+5]){
log_err("ERROR: UTF8_PREV_CHAR_SAFE(strict) failed for input=%ld. Expected:%lx Got:%lx\n", offset, result[i+5], c);
}
i=i+6;
}
{
/* test non-characters */
static const uint8_t nonChars[]={
0xef, 0xb7, 0x90, /* U+fdd0 */
0xef, 0xbf, 0xbf, /* U+feff */
0xf0, 0x9f, 0xbf, 0xbe, /* U+1fffe */
0xf0, 0xbf, 0xbf, 0xbf, /* U+3ffff */
0xf4, 0x8f, 0xbf, 0xbe /* U+10fffe */
};
UChar32 ch;
int32_t idx;
for(idx=0; idx<(int32_t)sizeof(nonChars);) {
U8_NEXT(nonChars, idx, sizeof(nonChars), ch);
if(!U_IS_UNICODE_NONCHAR(ch)) {
log_err("U8_NEXT(before %d) failed to read a non-character\n", idx);
}
}
for(idx=(int32_t)sizeof(nonChars); idx>0;) {
U8_PREV(nonChars, 0, idx, ch);
if(!U_IS_UNICODE_NONCHAR(ch)) {
log_err("U8_PREV(at %d) failed to read a non-character\n", idx);
}
}
}
}