static VALUE
str_compat_and_valid(VALUE str, rb_encoding *enc)
{
int cr;
str = StringValue(str);
cr = rb_enc_str_coderange(str);
if (cr == ENC_CODERANGE_BROKEN) {
#ifdef PRIsVALUE
rb_raise(rb_eArgError, "replacement must be valid byte sequence '%+"PRIsVALUE"'", str);
#else
str = rb_inspect(str);
rb_raise(rb_eArgError, "replacement must be valid byte sequence '%s'", RSTRING_PTR(str));
RB_GC_GUARD(str);
#endif
}
else if (cr == ENC_CODERANGE_7BIT) {
rb_encoding *e = STR_ENC_GET(str);
if (!rb_enc_asciicompat(enc)) {
rb_raise(rb_eEncCompatError, "incompatible character encodings: %s and %s",
rb_enc_name(enc), rb_enc_name(e));
}
}
else { /* ENC_CODERANGE_VALID */
rb_encoding *e = STR_ENC_GET(str);
if (enc != e) {
rb_raise(rb_eEncCompatError, "incompatible character encodings: %s and %s",
rb_enc_name(enc), rb_enc_name(e));
}
}
return str;
}
static mrb_value
sym_inspect(mrb_state *mrb, mrb_value sym)
{
#ifdef INCLUDE_ENCODING
#define STR_ENC_GET(mrb, str) mrb_enc_from_index(mrb, ENCODING_GET(mrb, str))
mrb_value str;
mrb_sym id = SYM2ID(sym);
mrb_encoding *enc;
const char *ptr;
long len;
char *dest;
mrb_encoding *resenc = mrb_default_internal_encoding(mrb);
if (resenc == NULL) resenc = mrb_default_external_encoding(mrb);
sym = mrb_str_new_cstr(mrb, mrb_sym2name(mrb, id));//mrb_id2str(id);
enc = STR_ENC_GET(mrb, sym);
ptr = RSTRING_PTR(sym);
len = RSTRING_LEN(sym);
if ((resenc != enc && !mrb_str_is_ascii_only_p(mrb, sym)) || len != (long)strlen(ptr) ||
!mrb_enc_symname_p(ptr, enc) || !sym_printable(mrb, ptr, ptr + len, enc)) {
str = mrb_str_inspect(mrb, sym);
len = RSTRING_LEN(str);
mrb_str_resize(mrb, str, len + 1);
dest = RSTRING_PTR(str);
memmove(dest + 1, dest, len);
dest[0] = ':';
}
else {
char *dest;
str = mrb_enc_str_new(mrb, 0, len + 1, enc);
dest = RSTRING_PTR(str);
dest[0] = ':';
memcpy(dest + 1, ptr, len);
}
return str;
#else
mrb_value str;
const char *name;
mrb_sym id = SYM2ID(sym);
name = mrb_sym2name(mrb, id); //mrb_id2name(id);
str = mrb_str_new(mrb, 0, strlen(name)+1);
RSTRING(str)->buf[0] = ':';
strcpy(RSTRING(str)->buf+1, name);
if (!mrb_symname_p(name)) {
str = mrb_str_dump(mrb, str);
strncpy(RSTRING(str)->buf, ":\"", 2);
}
return str;
#endif
}
static VALUE
rb_str_blank_as(VALUE str)
{
rb_encoding *enc;
char *s, *e;
enc = STR_ENC_GET(str);
s = RSTRING_PTR(str);
if (!s || RSTRING_LEN(str) == 0) return Qtrue;
e = RSTRING_END(str);
while (s < e) {
int n;
unsigned int cc = rb_enc_codepoint_len(s, e, &n, enc);
switch (cc) {
case 9:
case 0xa:
case 0xb:
case 0xc:
case 0xd:
case 0x20:
case 0x85:
case 0xa0:
case 0x1680:
case 0x180e:
case 0x2000:
case 0x2001:
case 0x2002:
case 0x2003:
case 0x2004:
case 0x2005:
case 0x2006:
case 0x2007:
case 0x2008:
case 0x2009:
case 0x200a:
case 0x2028:
case 0x2029:
case 0x202f:
case 0x205f:
case 0x3000:
/* found */
break;
default:
return Qfalse;
}
s += n;
}
return Qtrue;
}
static VALUE
rb_str_blank(VALUE str)
{
rb_encoding *enc;
char *s, *e;
enc = STR_ENC_GET(str);
s = RSTRING_PTR(str);
if (!s || RSTRING_LEN(str) == 0) return Qtrue;
e = RSTRING_END(str);
while (s < e) {
int n;
unsigned int cc = rb_enc_codepoint_len(s, e, &n, enc);
if (!rb_isspace(cc) && cc != 0) return Qfalse;
s += n;
}
return Qtrue;
}
/**
* @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;
}
}
