/**
* @param str the string to be scrubbed
* @param repl the replacement character
* @return If given string is invalid, returns a new string. Otherwise, returns Qnil.
*/
static VALUE
str_scrub0(int argc, VALUE *argv, VALUE str)
{
int cr = ENC_CODERANGE(str);
rb_encoding *enc;
int encidx;
VALUE repl;
if (cr == ENC_CODERANGE_7BIT || cr == ENC_CODERANGE_VALID)
return Qnil;
enc = STR_ENC_GET(str);
rb_scan_args(argc, argv, "01", &repl);
if (argc != 0) {
repl = str_compat_and_valid(repl, enc);
}
if (rb_enc_dummy_p(enc)) {
return Qnil;
}
encidx = rb_enc_to_index(enc);
#define DEFAULT_REPLACE_CHAR(str) do { \
static const char replace[sizeof(str)-1] = str; \
rep = replace; replen = (int)sizeof(replace); \
} while (0)
if (rb_enc_asciicompat(enc)) {
const char *p = RSTRING_PTR(str);
const char *e = RSTRING_END(str);
const char *p1 = p;
const char *rep;
long replen;
int rep7bit_p;
VALUE buf = Qnil;
if (rb_block_given_p()) {
rep = NULL;
replen = 0;
rep7bit_p = FALSE;
}
else if (!NIL_P(repl)) {
rep = RSTRING_PTR(repl);
replen = RSTRING_LEN(repl);
rep7bit_p = (ENC_CODERANGE(repl) == ENC_CODERANGE_7BIT);
}
else if (encidx == rb_utf8_encindex()) {
DEFAULT_REPLACE_CHAR("\xEF\xBF\xBD");
rep7bit_p = FALSE;
}
else {
DEFAULT_REPLACE_CHAR("?");
rep7bit_p = TRUE;
}
cr = ENC_CODERANGE_7BIT;
p = search_nonascii(p, e);
if (!p) {
p = e;
}
while (p < e) {
int ret = rb_enc_precise_mbclen(p, e, enc);
if (MBCLEN_NEEDMORE_P(ret)) {
break;
}
else if (MBCLEN_CHARFOUND_P(ret)) {
cr = ENC_CODERANGE_VALID;
p += MBCLEN_CHARFOUND_LEN(ret);
}
else if (MBCLEN_INVALID_P(ret)) {
/*
* p1~p: valid ascii/multibyte chars
* p ~e: invalid bytes + unknown bytes
*/
long clen = rb_enc_mbmaxlen(enc);
if (NIL_P(buf)) buf = rb_str_buf_new(RSTRING_LEN(str));
if (p > p1) {
rb_str_buf_cat(buf, p1, p - p1);
}
if (e - p < clen) clen = e - p;
if (clen <= 2) {
clen = 1;
}
else {
const char *q = p;
clen--;
for (; clen > 1; clen--) {
ret = rb_enc_precise_mbclen(q, q + clen, enc);
if (MBCLEN_NEEDMORE_P(ret)) break;
if (MBCLEN_INVALID_P(ret)) continue;
UNREACHABLE;
}
}
if (rep) {
rb_str_buf_cat(buf, rep, replen);
if (!rep7bit_p) cr = ENC_CODERANGE_VALID;
}
else {
repl = rb_yield(rb_enc_str_new(p, clen, enc));
repl = str_compat_and_valid(repl, enc);
rb_str_buf_cat(buf, RSTRING_PTR(repl), RSTRING_LEN(repl));
if (ENC_CODERANGE(repl) == ENC_CODERANGE_VALID)
cr = ENC_CODERANGE_VALID;
}
p += clen;
p1 = p;
p = search_nonascii(p, e);
if (!p) {
p = e;
break;
}
}
else {
UNREACHABLE;
}
}
if (NIL_P(buf)) {
if (p == e) {
ENC_CODERANGE_SET(str, cr);
return Qnil;
}
buf = rb_str_buf_new(RSTRING_LEN(str));
}
if (p1 < p) {
rb_str_buf_cat(buf, p1, p - p1);
}
if (p < e) {
if (rep) {
rb_str_buf_cat(buf, rep, replen);
if (!rep7bit_p) cr = ENC_CODERANGE_VALID;
}
else {
repl = rb_yield(rb_enc_str_new(p, e-p, enc));
repl = str_compat_and_valid(repl, enc);
rb_str_buf_cat(buf, RSTRING_PTR(repl), RSTRING_LEN(repl));
if (ENC_CODERANGE(repl) == ENC_CODERANGE_VALID)
cr = ENC_CODERANGE_VALID;
}
}
ENCODING_CODERANGE_SET(buf, rb_enc_to_index(enc), cr);
return buf;
}
else {
/* ASCII incompatible */
const char *p = RSTRING_PTR(str);
const char *e = RSTRING_END(str);
const char *p1 = p;
VALUE buf = Qnil;
const char *rep;
long replen;
long mbminlen = rb_enc_mbminlen(enc);
if (!NIL_P(repl)) {
rep = RSTRING_PTR(repl);
replen = RSTRING_LEN(repl);
}
else if (!strcasecmp(rb_enc_name(enc), "UTF-16BE")) {
DEFAULT_REPLACE_CHAR("\xFF\xFD");
}
else if (!strcasecmp(rb_enc_name(enc), "UTF-16LE")) {
DEFAULT_REPLACE_CHAR("\xFD\xFF");
}
else if (!strcasecmp(rb_enc_name(enc), "UTF-32BE")) {
DEFAULT_REPLACE_CHAR("\x00\x00\xFF\xFD");
}
else if (!strcasecmp(rb_enc_name(enc), "UTF-32lE")) {
DEFAULT_REPLACE_CHAR("\xFD\xFF\x00\x00");
}
else {
DEFAULT_REPLACE_CHAR("?");
}
while (p < e) {
int ret = rb_enc_precise_mbclen(p, e, enc);
if (MBCLEN_NEEDMORE_P(ret)) {
break;
}
else if (MBCLEN_CHARFOUND_P(ret)) {
p += MBCLEN_CHARFOUND_LEN(ret);
}
else if (MBCLEN_INVALID_P(ret)) {
const char *q = p;
long clen = rb_enc_mbmaxlen(enc);
if (NIL_P(buf)) buf = rb_str_buf_new(RSTRING_LEN(str));
if (p > p1) rb_str_buf_cat(buf, p1, p - p1);
if (e - p < clen) clen = e - p;
if (clen <= mbminlen * 2) {
clen = mbminlen;
}
else {
clen -= mbminlen;
for (; clen > mbminlen; clen-=mbminlen) {
ret = rb_enc_precise_mbclen(q, q + clen, enc);
if (MBCLEN_NEEDMORE_P(ret)) break;
if (MBCLEN_INVALID_P(ret)) continue;
UNREACHABLE;
}
}
if (rep) {
rb_str_buf_cat(buf, rep, replen);
}
else {
repl = rb_yield(rb_enc_str_new(p, e-p, enc));
repl = str_compat_and_valid(repl, enc);
rb_str_buf_cat(buf, RSTRING_PTR(repl), RSTRING_LEN(repl));
}
p += clen;
p1 = p;
}
else {
UNREACHABLE;
}
}
if (NIL_P(buf)) {
if (p == e) {
ENC_CODERANGE_SET(str, ENC_CODERANGE_VALID);
return Qnil;
}
buf = rb_str_buf_new(RSTRING_LEN(str));
}
if (p1 < p) {
rb_str_buf_cat(buf, p1, p - p1);
}
if (p < e) {
if (rep) {
rb_str_buf_cat(buf, rep, replen);
}
else {
repl = rb_yield(rb_enc_str_new(p, e-p, enc));
repl = str_compat_and_valid(repl, enc);
rb_str_buf_cat(buf, RSTRING_PTR(repl), RSTRING_LEN(repl));
}
}
ENCODING_CODERANGE_SET(buf, rb_enc_to_index(enc), ENC_CODERANGE_VALID);
return buf;
}
}
static VALUE
optimized_unescape(VALUE str, VALUE encoding)
{
long i, len, beg = 0;
VALUE dest = 0;
const char *cstr;
int cr, origenc, encidx = rb_to_encoding_index(encoding);
len = RSTRING_LEN(str);
cstr = RSTRING_PTR(str);
for (i = 0; i < len; ++i) {
char buf[1];
const char c = cstr[i];
int clen = 0;
if (c == '%') {
if (i + 3 > len) break;
if (!ISXDIGIT(cstr[i+1])) continue;
if (!ISXDIGIT(cstr[i+2])) continue;
buf[0] = ((char_to_number(cstr[i+1]) << 4)
| char_to_number(cstr[i+2]));
clen = 2;
}
else if (c == '+') {
buf[0] = ' ';
}
else {
continue;
}
if (!dest) {
dest = rb_str_buf_new(len);
}
rb_str_cat(dest, cstr + beg, i - beg);
i += clen;
beg = i + 1;
rb_str_cat(dest, buf, 1);
}
if (dest) {
rb_str_cat(dest, cstr + beg, len - beg);
preserve_original_state(str, dest);
cr = ENC_CODERANGE_UNKNOWN;
}
else {
dest = rb_str_dup(str);
cr = ENC_CODERANGE(str);
}
origenc = rb_enc_get_index(str);
if (origenc != encidx) {
rb_enc_associate_index(dest, encidx);
if (!ENC_CODERANGE_CLEAN_P(rb_enc_str_coderange(dest))) {
rb_enc_associate_index(dest, origenc);
if (cr != ENC_CODERANGE_UNKNOWN)
ENC_CODERANGE_SET(dest, cr);
}
}
return dest;
}
static int
str_transcode(int argc, VALUE *argv, VALUE *self)
{
VALUE dest;
VALUE str = *self;
long blen, slen;
unsigned char *buf, *bp, *sp, *fromp;
rb_encoding *from_enc, *to_enc;
const char *from_e, *to_e;
int from_encidx, to_encidx;
VALUE from_encval, to_encval;
const rb_transcoder *my_transcoder;
rb_transcoding my_transcoding;
int final_encoding = 0;
VALUE opt;
int options = 0;
opt = rb_check_convert_type(argv[argc-1], T_HASH, "Hash", "to_hash");
if (!NIL_P(opt)) {
VALUE v;
argc--;
v = rb_hash_aref(opt, sym_invalid);
if (NIL_P(v)) {
rb_raise(rb_eArgError, "unknown value for invalid: setting");
}
else if (v==sym_ignore) {
options |= INVALID_IGNORE;
}
}
if (argc < 1 || argc > 2) {
rb_raise(rb_eArgError, "wrong number of arguments (%d for 1..2)", argc);
}
if ((to_encidx = rb_to_encoding_index(to_encval = argv[0])) < 0) {
to_enc = 0;
to_encidx = 0;
to_e = StringValueCStr(to_encval);
}
else {
to_enc = rb_enc_from_index(to_encidx);
to_e = rb_enc_name(to_enc);
}
if (argc==1) {
from_encidx = rb_enc_get_index(str);
from_enc = rb_enc_from_index(from_encidx);
from_e = rb_enc_name(from_enc);
}
else if ((from_encidx = rb_to_encoding_index(from_encval = argv[1])) < 0) {
from_enc = 0;
from_e = StringValueCStr(from_encval);
}
else {
from_enc = rb_enc_from_index(from_encidx);
from_e = rb_enc_name(from_enc);
}
if (from_enc && from_enc == to_enc) {
return -1;
}
if (from_enc && to_enc && rb_enc_asciicompat(from_enc) && rb_enc_asciicompat(to_enc)) {
if (ENC_CODERANGE(str) == ENC_CODERANGE_7BIT) {
return to_encidx;
}
}
if (encoding_equal(from_e, to_e)) {
return -1;
}
do { /* loop for multistep transcoding */
/* later, maybe use smaller intermediate strings for very long strings */
if (!(my_transcoder = transcode_dispatch(from_e, to_e))) {
rb_raise(rb_eArgError, "transcoding not supported (from %s to %s)", from_e, to_e);
}
my_transcoding.transcoder = my_transcoder;
if (my_transcoder->preprocessor) {
fromp = sp = (unsigned char *)RSTRING_PTR(str);
slen = RSTRING_LEN(str);
blen = slen + 30; /* len + margin */
dest = rb_str_tmp_new(blen);
bp = (unsigned char *)RSTRING_PTR(dest);
my_transcoding.ruby_string_dest = dest;
(*my_transcoder->preprocessor)(&fromp, &bp, (sp+slen), (bp+blen), &my_transcoding);
if (fromp != sp+slen) {
rb_raise(rb_eArgError, "not fully converted, %td bytes left", sp+slen-fromp);
}
buf = (unsigned char *)RSTRING_PTR(dest);
*bp = '\0';
rb_str_set_len(dest, bp - buf);
str = dest;
}
fromp = sp = (unsigned char *)RSTRING_PTR(str);
slen = RSTRING_LEN(str);
blen = slen + 30; /* len + margin */
dest = rb_str_tmp_new(blen);
bp = (unsigned char *)RSTRING_PTR(dest);
my_transcoding.ruby_string_dest = dest;
my_transcoding.flush_func = str_transcoding_resize;
transcode_loop(&fromp, &bp, (sp+slen), (bp+blen), my_transcoder, &my_transcoding, options);
if (fromp != sp+slen) {
rb_raise(rb_eArgError, "not fully converted, %td bytes left", sp+slen-fromp);
}
buf = (unsigned char *)RSTRING_PTR(dest);
*bp = '\0';
rb_str_set_len(dest, bp - buf);
if (my_transcoder->postprocessor) {
str = dest;
fromp = sp = (unsigned char *)RSTRING_PTR(str);
slen = RSTRING_LEN(str);
blen = slen + 30; /* len + margin */
dest = rb_str_tmp_new(blen);
bp = (unsigned char *)RSTRING_PTR(dest);
my_transcoding.ruby_string_dest = dest;
(*my_transcoder->postprocessor)(&fromp, &bp, (sp+slen), (bp+blen), &my_transcoding);
if (fromp != sp+slen) {
rb_raise(rb_eArgError, "not fully converted, %td bytes left", sp+slen-fromp);
}
buf = (unsigned char *)RSTRING_PTR(dest);
*bp = '\0';
rb_str_set_len(dest, bp - buf);
}
if (encoding_equal(my_transcoder->to_encoding, to_e)) {
final_encoding = 1;
}
else {
from_e = my_transcoder->to_encoding;
str = dest;
}
} while (!final_encoding);
/* set encoding */
if (!to_enc) {
to_encidx = rb_define_dummy_encoding(to_e);
}
*self = dest;
return to_encidx;
}
