VALUE
rb_grn_context_rb_string_encode (grn_ctx *context, VALUE rb_string)
{
#ifdef HAVE_RUBY_ENCODING_H
rb_encoding *encoding, *to_encode;
encoding = rb_enc_get(rb_string);
to_encode = rb_grn_encoding_to_ruby_encoding(context->encoding);
if (rb_enc_to_index(encoding) != rb_enc_to_index(to_encode))
rb_string = rb_str_encode(rb_string, rb_enc_from_encoding(to_encode),
0, Qnil);
#endif
return rb_string;
}
static int
enc_set_default_encoding(struct default_encoding *def, VALUE encoding, const char *name)
{
int overridden = FALSE;
if (def->index != -2)
/* Already set */
overridden = TRUE;
if (NIL_P(encoding)) {
def->index = -1;
def->enc = 0;
st_insert(enc_table.names, (st_data_t)strdup(name),
(st_data_t)UNSPECIFIED_ENCODING);
}
else {
def->index = rb_enc_to_index(rb_to_encoding(encoding));
def->enc = 0;
enc_alias_internal(name, def->index);
}
if (def == &default_external)
enc_set_filesystem_encoding();
return overridden;
}
/*
* call-seq:
* res.check -> nil
*
* Raises appropriate exception if PG::Result is in a bad state.
*/
VALUE
pg_result_check( VALUE self )
{
VALUE error, exception, klass;
VALUE rb_pgconn = rb_iv_get( self, "@connection" );
PGconn *conn = pg_get_pgconn(rb_pgconn);
PGresult *result;
#ifdef M17N_SUPPORTED
rb_encoding *enc = pg_conn_enc_get( conn );
#endif
char * sqlstate;
Data_Get_Struct(self, PGresult, result);
if(result == NULL)
{
error = rb_str_new2( PQerrorMessage(conn) );
}
else
{
switch (PQresultStatus(result))
{
case PGRES_TUPLES_OK:
case PGRES_COPY_OUT:
case PGRES_COPY_IN:
#ifdef HAVE_CONST_PGRES_COPY_BOTH
case PGRES_COPY_BOTH:
#endif
#ifdef HAVE_CONST_PGRES_SINGLE_TUPLE
case PGRES_SINGLE_TUPLE:
#endif
case PGRES_EMPTY_QUERY:
case PGRES_COMMAND_OK:
return Qnil;
case PGRES_BAD_RESPONSE:
case PGRES_FATAL_ERROR:
case PGRES_NONFATAL_ERROR:
error = rb_str_new2( PQresultErrorMessage(result) );
break;
default:
error = rb_str_new2( "internal error : unknown result status." );
}
}
#ifdef M17N_SUPPORTED
rb_enc_set_index( error, rb_enc_to_index(enc) );
#endif
sqlstate = PQresultErrorField( result, PG_DIAG_SQLSTATE );
klass = lookup_error_class( sqlstate );
exception = rb_exc_new3( klass, error );
rb_iv_set( exception, "@connection", rb_pgconn );
rb_iv_set( exception, "@result", self );
rb_exc_raise( exception );
/* Not reached */
return Qnil;
}
static int
enc_set_filesystem_encoding(void)
{
int idx;
#if defined NO_LOCALE_CHARMAP
idx = rb_enc_to_index(rb_default_external_encoding());
#elif defined _WIN32 || defined __CYGWIN__
char cp[sizeof(int) * 8 / 3 + 4];
snprintf(cp, sizeof cp, "CP%d", AreFileApisANSI() ? GetACP() : GetOEMCP());
idx = rb_enc_find_index(cp);
if (idx < 0) idx = rb_ascii8bit_encindex();
#else
idx = rb_enc_to_index(rb_default_external_encoding());
#endif
enc_alias_internal("filesystem", idx);
return idx;
}
static int
check_encoding(rb_encoding *enc)
{
int index = rb_enc_to_index(enc);
if (rb_enc_from_index(index) != enc)
return -1;
if (enc_autoload_p(enc)) {
index = enc_autoload(enc);
}
return index;
}
/*
* Result constructor
*/
VALUE
pg_new_result(PGresult *result, VALUE rb_pgconn)
{
PGconn *conn = pg_get_pgconn( rb_pgconn );
VALUE val = Data_Wrap_Struct(rb_cPGresult, NULL, pgresult_gc_free, result);
#ifdef M17N_SUPPORTED
rb_encoding *enc = pg_conn_enc_get( conn );
ENCODING_SET( val, rb_enc_to_index(enc) );
#endif
rb_iv_set( val, "@connection", rb_pgconn );
return val;
}
int
Init_enc_set_filesystem_encoding(void)
{
int idx;
#if defined NO_LOCALE_CHARMAP
# error NO_LOCALE_CHARMAP defined
#elif defined _WIN32 || defined __CYGWIN__
char cp[SIZEOF_CP_NAME];
CP_FORMAT(cp, AreFileApisANSI() ? GetACP() : GetOEMCP());
idx = rb_enc_find_index(cp);
if (idx < 0) idx = ENCINDEX_ASCII;
#else
idx = rb_enc_to_index(rb_default_external_encoding());
#endif
return idx;
}
/*
* call-seq:
* enc.inspect -> string
*
* Returns a string which represents the encoding for programmers.
*
* Encoding::UTF_8.inspect #=> "#<Encoding:UTF-8>"
* Encoding::ISO_2022_JP.inspect #=> "#<Encoding:ISO-2022-JP (dummy)>"
*/
static VALUE
enc_inspect(VALUE self)
{
rb_encoding *enc;
if (!is_data_encoding(self)) {
not_encoding(self);
}
if (!(enc = DATA_PTR(self)) || rb_enc_from_index(rb_enc_to_index(enc)) != enc) {
rb_raise(rb_eTypeError, "broken Encoding");
}
return rb_enc_sprintf(rb_usascii_encoding(),
"#<%"PRIsVALUE":%s%s%s>", rb_obj_class(self),
rb_enc_name(enc),
(ENC_DUMMY_P(enc) ? " (dummy)" : ""),
enc_autoload_p(enc) ? " (autoload)" : "");
}
/*
Return code page number of the encoding.
Cache code page into a hash for performance since finding the code page in
Encoding#names is slow.
*/
static UINT
code_page(rb_encoding *enc)
{
int enc_idx;
if (!enc)
return system_code_page();
enc_idx = rb_enc_to_index(enc);
/* map US-ASCII and ASCII-8bit as code page 1252 (us-ascii) */
if (enc_idx == rb_usascii_encindex() || enc_idx == rb_ascii8bit_encindex()) {
return 1252;
}
if (enc_idx == rb_utf8_encindex()) {
return CP_UTF8;
}
if (0 <= enc_idx && (unsigned int)enc_idx < rb_code_page.count)
return rb_code_page.table[enc_idx];
return INVALID_CODE_PAGE;
}
VALUE
rb_enc_associate(VALUE obj, rb_encoding *enc)
{
return rb_enc_associate_index(obj, rb_enc_to_index(enc));
}
static VALUE encoding_spec_rb_enc_to_index(VALUE self, VALUE name) {
return INT2NUM(rb_enc_to_index(rb_enc_find(RSTRING_PTR(name))));
}
int
Init_enc_set_filesystem_encoding(void)
{
return rb_enc_to_index(rb_default_external_encoding());
}
/**
* @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;
rb_encoding *enc = rb_to_encoding(encoding);
int cr, origenc, encidx = rb_enc_to_index(enc);
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;
}
