PARROT_WARN_UNUSED_RESULT
PARROT_CANNOT_RETURN_NULL
static STRING *
utf16_to_encoding(PARROT_INTERP, ARGIN(const STRING *src))
{
ASSERT_ARGS(utf16_to_encoding)
STRING *result;
UINTVAL src_len;
src_len = STRING_length(src);
if (STRING_max_bytes_per_codepoint(src) == 1) {
result = Parrot_gc_new_string_header(interp, 0);
result->encoding = Parrot_ucs2_encoding_ptr;
result->bufused = 2 * src_len;
result->strlen = src_len;
if (src_len) {
UINTVAL i;
Parrot_UInt2 *p;
Parrot_gc_allocate_string_storage(interp, result, 2 * src_len);
p = (Parrot_UInt2 *)result->strstart;
for (i = 0; i < src_len; ++i) {
p[i] = (unsigned char)src->strstart[i];
}
}
}
else if (src->encoding == Parrot_utf16_encoding_ptr
|| src->encoding == Parrot_ucs2_encoding_ptr) {
/* we have to use clone instead of copy because the Unicode upcase
* and downcase functions assume to get an unshared buffer */
result = Parrot_str_clone(interp, src);
}
else {
result = encoding_to_encoding(interp, src, Parrot_utf16_encoding_ptr, 2.2);
/* downgrade if possible */
if (result->bufused == result->strlen << 1)
result->encoding = Parrot_ucs2_encoding_ptr;
}
return result;
}
PARROT_CANNOT_RETURN_NULL
static STRING *
ascii_to_encoding(PARROT_INTERP, ARGIN(const STRING *src))
{
ASSERT_ARGS(ascii_to_encoding)
STRING *dest;
if (STRING_max_bytes_per_codepoint(src) == 1) {
unsigned char * const src_buf = (unsigned char *)src->strstart;
UINTVAL offs;
for (offs = 0; offs < src->strlen; ++offs) {
UINTVAL c = src_buf[offs];
if (c >= 0x80)
Parrot_ex_throw_from_c_args(interp, NULL, EXCEPTION_LOSSY_CONVERSION,
"lossy conversion to ascii");
}
dest = Parrot_str_clone(interp, src);
dest->encoding = Parrot_ascii_encoding_ptr;
}
else {
String_iter iter;
unsigned char *p;
const UINTVAL len = src->strlen;
dest = Parrot_str_new_init(interp, NULL, len,
Parrot_ascii_encoding_ptr, 0);
p = (unsigned char *)dest->strstart;
STRING_ITER_INIT(interp, &iter);
while (iter.charpos < len) {
const UINTVAL c = STRING_iter_get_and_advance(interp, src, &iter);
if (c >= 0x80)
Parrot_ex_throw_from_c_args(interp, NULL, EXCEPTION_LOSSY_CONVERSION,
"can't convert unicode string to ascii");
*p++ = c;
}
dest->bufused = len;
dest->strlen = len;
}
return dest;
}
PARROT_WARN_UNUSED_RESULT
PARROT_CANNOT_RETURN_NULL
static STRING *
ucs4_to_encoding(PARROT_INTERP, ARGIN(const STRING *src))
{
ASSERT_ARGS(ucs4_to_encoding)
const UINTVAL len = src->strlen;
UINTVAL i;
STRING *res;
utf32_t *ptr;
if (src->encoding == Parrot_ucs4_encoding_ptr)
return Parrot_str_copy(interp, src);
res = Parrot_str_new_init(interp, NULL, len * 4,
Parrot_ucs4_encoding_ptr, 0);
ptr = (utf32_t *)res->strstart;
if (STRING_max_bytes_per_codepoint(src) == 1) {
const unsigned char *s = (unsigned char *)src->strstart;
for (i = 0; i < len; i++) {
ptr[i] = s[i];
}
}
else {
String_iter iter;
STRING_ITER_INIT(interp, &iter);
while (iter.charpos < len) {
i = iter.charpos;
ptr[i] = STRING_iter_get_and_advance(interp, src, &iter);
}
}
res->strlen = len;
res->bufused = len * 4;
return res;
}
