/**
* HTML methods
*/
static VALUE rb_eu_escape_html(int argc, VALUE *argv, VALUE self)
{
VALUE rb_out_buf, str, rb_secure;
struct buf *out_buf;
int secure = g_html_secure;
if (rb_scan_args(argc, argv, "11", &str, &rb_secure) == 2) {
if (rb_secure == Qfalse) {
secure = 0;
}
}
Check_Type(str, T_STRING);
out_buf = bufnew(128);
houdini_escape_html0(out_buf, (uint8_t *)RSTRING_PTR(str), RSTRING_LEN(str), secure);
rb_out_buf = rb_str_new((char *)out_buf->data, out_buf->size);
bufrelease(out_buf);
#ifdef HAVE_RUBY_ENCODING_H
rb_enc_copy(rb_out_buf, str);
#endif
return rb_out_buf;
}
/**
* Generic template
*/
static VALUE
rb_eu__generic(
VALUE self, VALUE str,
houdini_cb callback,
size_t chunk_size)
{
VALUE result;
struct buf *out_buf;
if (NIL_P(str))
return rb_str_new2("");
Check_Type(str, T_STRING);
out_buf = bufnew(chunk_size);
callback(out_buf, (uint8_t *)RSTRING_PTR(str), RSTRING_LEN(str));
result = rb_str_new((char *)out_buf->data, out_buf->size);
bufrelease(out_buf);
#ifdef HAVE_RUBY_ENCODING_H
rb_enc_copy(result, str);
#endif
return result;
}
void
rb_attr(VALUE klass, ID id, int read, int write, int ex)
{
const char *name;
ID attriv;
VALUE aname;
int noex;
if (!ex) {
noex = NOEX_PUBLIC;
}
else {
if (SCOPE_TEST(NOEX_PRIVATE)) {
noex = NOEX_PRIVATE;
rb_warning((SCOPE_CHECK(NOEX_MODFUNC)) ?
"attribute accessor as module_function" :
"private attribute?");
}
else if (SCOPE_TEST(NOEX_PROTECTED)) {
noex = NOEX_PROTECTED;
}
else {
noex = NOEX_PUBLIC;
}
}
if (!rb_is_local_id(id) && !rb_is_const_id(id)) {
rb_name_error(id, "invalid attribute name `%s'", rb_id2name(id));
}
name = rb_id2name(id);
if (!name) {
rb_raise(rb_eArgError, "argument needs to be symbol or string");
}
aname = rb_sprintf("@%s", name);
rb_enc_copy(aname, rb_id2str(id));
attriv = rb_intern_str(aname);
if (read) {
rb_add_method(klass, id, NEW_IVAR(attriv), noex);
}
if (write) {
rb_add_method(klass, rb_id_attrset(id), NEW_ATTRSET(attriv), noex);
}
}
static VALUE rb_mysql_client_escape(RB_MYSQL_UNUSED VALUE klass, VALUE str) {
unsigned char *newStr;
VALUE rb_str;
unsigned long newLen, oldLen;
Check_Type(str, T_STRING);
oldLen = RSTRING_LEN(str);
newStr = malloc(oldLen*2+1);
newLen = mysql_escape_string((char *)newStr, StringValuePtr(str), oldLen);
if (newLen == oldLen) {
// no need to return a new ruby string if nothing changed
free(newStr);
return str;
} else {
rb_str = rb_str_new((const char*)newStr, newLen);
#ifdef HAVE_RUBY_ENCODING_H
rb_enc_copy(rb_str, str);
#endif
free(newStr);
return rb_str;
}
}
VALUE
rb_str_format(int argc, const VALUE *argv, VALUE fmt)
{
enum {default_float_precision = 6};
rb_encoding *enc;
const char *p, *end;
char *buf;
long blen, bsiz;
VALUE result;
long scanned = 0;
int coderange = ENC_CODERANGE_7BIT;
int width, prec, flags = FNONE;
int nextarg = 1;
int posarg = 0;
int tainted = 0;
VALUE nextvalue;
VALUE tmp;
VALUE str;
volatile VALUE hash = Qundef;
#define CHECK_FOR_WIDTH(f) \
if ((f) & FWIDTH) { \
rb_raise(rb_eArgError, "width given twice"); \
} \
if ((f) & FPREC0) { \
rb_raise(rb_eArgError, "width after precision"); \
}
#define CHECK_FOR_FLAGS(f) \
if ((f) & FWIDTH) { \
rb_raise(rb_eArgError, "flag after width"); \
} \
if ((f) & FPREC0) { \
rb_raise(rb_eArgError, "flag after precision"); \
}
++argc;
--argv;
if (OBJ_TAINTED(fmt)) tainted = 1;
StringValue(fmt);
enc = rb_enc_get(fmt);
fmt = rb_str_new4(fmt);
p = RSTRING_PTR(fmt);
end = p + RSTRING_LEN(fmt);
blen = 0;
bsiz = 120;
result = rb_str_buf_new(bsiz);
rb_enc_copy(result, fmt);
buf = RSTRING_PTR(result);
memset(buf, 0, bsiz);
ENC_CODERANGE_SET(result, coderange);
for (; p < end; p++) {
const char *t;
int n;
VALUE sym = Qnil;
for (t = p; t < end && *t != '%'; t++) ;
PUSH(p, t - p);
if (coderange != ENC_CODERANGE_BROKEN && scanned < blen) {
scanned += rb_str_coderange_scan_restartable(buf+scanned, buf+blen, enc, &coderange);
ENC_CODERANGE_SET(result, coderange);
}
if (t >= end) {
/* end of fmt string */
goto sprint_exit;
}
p = t + 1; /* skip `%' */
width = prec = -1;
nextvalue = Qundef;
retry:
switch (*p) {
default:
if (rb_enc_isprint(*p, enc))
rb_raise(rb_eArgError, "malformed format string - %%%c", *p);
else
rb_raise(rb_eArgError, "malformed format string");
break;
case ' ':
CHECK_FOR_FLAGS(flags);
flags |= FSPACE;
p++;
goto retry;
case '#':
CHECK_FOR_FLAGS(flags);
flags |= FSHARP;
p++;
goto retry;
case '+':
CHECK_FOR_FLAGS(flags);
flags |= FPLUS;
p++;
goto retry;
case '-':
CHECK_FOR_FLAGS(flags);
flags |= FMINUS;
p++;
goto retry;
case '0':
CHECK_FOR_FLAGS(flags);
flags |= FZERO;
p++;
goto retry;
case '1': case '2': case '3': case '4':
case '5': case '6': case '7': case '8': case '9':
n = 0;
GETNUM(n, width);
if (*p == '$') {
if (nextvalue != Qundef) {
rb_raise(rb_eArgError, "value given twice - %d$", n);
}
nextvalue = GETPOSARG(n);
p++;
goto retry;
}
CHECK_FOR_WIDTH(flags);
width = n;
flags |= FWIDTH;
goto retry;
case '<':
case '{':
{
const char *start = p;
char term = (*p == '<') ? '>' : '}';
int len;
for (; p < end && *p != term; ) {
p += rb_enc_mbclen(p, end, enc);
}
if (p >= end) {
rb_raise(rb_eArgError, "malformed name - unmatched parenthesis");
}
#if SIZEOF_INT < SIZEOF_SIZE_T
if ((size_t)(p - start) >= INT_MAX) {
const int message_limit = 20;
len = (int)(rb_enc_right_char_head(start, start + message_limit, p, enc) - start);
rb_enc_raise(enc, rb_eArgError,
"too long name (%"PRIdSIZE" bytes) - %.*s...%c",
(size_t)(p - start - 2), len, start, term);
}
#endif
len = (int)(p - start + 1); /* including parenthesis */
if (sym != Qnil) {
rb_enc_raise(enc, rb_eArgError, "named%.*s after <%"PRIsVALUE">",
len, start, rb_sym2str(sym));
}
CHECKNAMEARG(start, len, enc);
get_hash(&hash, argc, argv);
sym = rb_check_symbol_cstr(start + 1,
len - 2 /* without parenthesis */,
enc);
if (sym != Qnil) nextvalue = rb_hash_lookup2(hash, sym, Qundef);
if (nextvalue == Qundef) {
rb_enc_raise(enc, rb_eKeyError, "key%.*s not found", len, start);
}
if (term == '}') goto format_s;
p++;
goto retry;
}
case '*':
CHECK_FOR_WIDTH(flags);
flags |= FWIDTH;
GETASTER(width);
if (width < 0) {
flags |= FMINUS;
width = -width;
}
p++;
goto retry;
case '.':
if (flags & FPREC0) {
rb_raise(rb_eArgError, "precision given twice");
}
flags |= FPREC|FPREC0;
prec = 0;
p++;
if (*p == '*') {
GETASTER(prec);
if (prec < 0) { /* ignore negative precision */
flags &= ~FPREC;
}
p++;
goto retry;
}
GETNUM(prec, precision);
goto retry;
case '\n':
case '\0':
p--;
case '%':
if (flags != FNONE) {
rb_raise(rb_eArgError, "invalid format character - %%");
}
PUSH("%", 1);
break;
case 'c':
{
VALUE val = GETARG();
VALUE tmp;
unsigned int c;
int n;
tmp = rb_check_string_type(val);
if (!NIL_P(tmp)) {
if (rb_enc_strlen(RSTRING_PTR(tmp),RSTRING_END(tmp),enc) != 1) {
rb_raise(rb_eArgError, "%%c requires a character");
}
c = rb_enc_codepoint_len(RSTRING_PTR(tmp), RSTRING_END(tmp), &n, enc);
RB_GC_GUARD(tmp);
}
else {
c = NUM2INT(val);
n = rb_enc_codelen(c, enc);
}
if (n <= 0) {
rb_raise(rb_eArgError, "invalid character");
}
if (!(flags & FWIDTH)) {
CHECK(n);
rb_enc_mbcput(c, &buf[blen], enc);
blen += n;
}
else if ((flags & FMINUS)) {
CHECK(n);
rb_enc_mbcput(c, &buf[blen], enc);
blen += n;
FILL(' ', width-1);
}
else {
FILL(' ', width-1);
CHECK(n);
rb_enc_mbcput(c, &buf[blen], enc);
blen += n;
}
}
break;
case 's':
case 'p':
format_s:
{
VALUE arg = GETARG();
long len, slen;
if (*p == 'p') arg = rb_inspect(arg);
str = rb_obj_as_string(arg);
if (OBJ_TAINTED(str)) tainted = 1;
len = RSTRING_LEN(str);
rb_str_set_len(result, blen);
if (coderange != ENC_CODERANGE_BROKEN && scanned < blen) {
int cr = coderange;
scanned += rb_str_coderange_scan_restartable(buf+scanned, buf+blen, enc, &cr);
ENC_CODERANGE_SET(result,
(cr == ENC_CODERANGE_UNKNOWN ?
ENC_CODERANGE_BROKEN : (coderange = cr)));
}
enc = rb_enc_check(result, str);
if (flags&(FPREC|FWIDTH)) {
slen = rb_enc_strlen(RSTRING_PTR(str),RSTRING_END(str),enc);
if (slen < 0) {
rb_raise(rb_eArgError, "invalid mbstring sequence");
}
if ((flags&FPREC) && (prec < slen)) {
char *p = rb_enc_nth(RSTRING_PTR(str), RSTRING_END(str),
prec, enc);
slen = prec;
len = p - RSTRING_PTR(str);
}
/* need to adjust multi-byte string pos */
if ((flags&FWIDTH) && (width > slen)) {
width -= (int)slen;
if (!(flags&FMINUS)) {
CHECK(width);
while (width--) {
buf[blen++] = ' ';
}
}
CHECK(len);
memcpy(&buf[blen], RSTRING_PTR(str), len);
RB_GC_GUARD(str);
blen += len;
if (flags&FMINUS) {
CHECK(width);
while (width--) {
buf[blen++] = ' ';
}
}
rb_enc_associate(result, enc);
break;
}
}
PUSH(RSTRING_PTR(str), len);
RB_GC_GUARD(str);
rb_enc_associate(result, enc);
}
break;
case 'd':
case 'i':
case 'o':
case 'x':
case 'X':
case 'b':
case 'B':
case 'u':
{
volatile VALUE val = GETARG();
int valsign;
char nbuf[64], *s;
const char *prefix = 0;
int sign = 0, dots = 0;
char sc = 0;
long v = 0;
int base, bignum = 0;
int len;
switch (*p) {
case 'd':
case 'i':
case 'u':
sign = 1; break;
case 'o':
case 'x':
case 'X':
case 'b':
case 'B':
if (flags&(FPLUS|FSPACE)) sign = 1;
break;
}
if (flags & FSHARP) {
switch (*p) {
case 'o':
prefix = "0"; break;
case 'x':
prefix = "0x"; break;
case 'X':
prefix = "0X"; break;
case 'b':
prefix = "0b"; break;
case 'B':
prefix = "0B"; break;
}
}
bin_retry:
switch (TYPE(val)) {
case T_FLOAT:
if (FIXABLE(RFLOAT_VALUE(val))) {
val = LONG2FIX((long)RFLOAT_VALUE(val));
goto bin_retry;
}
val = rb_dbl2big(RFLOAT_VALUE(val));
if (FIXNUM_P(val)) goto bin_retry;
bignum = 1;
break;
case T_STRING:
val = rb_str_to_inum(val, 0, TRUE);
goto bin_retry;
case T_BIGNUM:
bignum = 1;
break;
case T_FIXNUM:
v = FIX2LONG(val);
break;
default:
val = rb_Integer(val);
goto bin_retry;
}
switch (*p) {
case 'o':
base = 8; break;
case 'x':
case 'X':
base = 16; break;
case 'b':
case 'B':
base = 2; break;
case 'u':
case 'd':
case 'i':
default:
base = 10; break;
}
if (base != 10) {
int numbits = ffs(base)-1;
size_t abs_nlz_bits;
size_t numdigits = rb_absint_numwords(val, numbits, &abs_nlz_bits);
long i;
if (INT_MAX-1 < numdigits) /* INT_MAX is used because rb_long2int is used later. */
rb_raise(rb_eArgError, "size too big");
if (sign) {
if (numdigits == 0)
numdigits = 1;
tmp = rb_str_new(NULL, numdigits);
valsign = rb_integer_pack(val, RSTRING_PTR(tmp), RSTRING_LEN(tmp),
1, CHAR_BIT-numbits, INTEGER_PACK_BIG_ENDIAN);
for (i = 0; i < RSTRING_LEN(tmp); i++)
RSTRING_PTR(tmp)[i] = ruby_digitmap[((unsigned char *)RSTRING_PTR(tmp))[i]];
s = RSTRING_PTR(tmp);
if (valsign < 0) {
sc = '-';
width--;
}
else if (flags & FPLUS) {
sc = '+';
width--;
}
else if (flags & FSPACE) {
sc = ' ';
width--;
}
}
else {
/* Following conditional "numdigits++" guarantees the
* most significant digit as
* - '1'(bin), '7'(oct) or 'f'(hex) for negative numbers
* - '0' for zero
* - not '0' for positive numbers.
*
* It also guarantees the most significant two
* digits will not be '11'(bin), '77'(oct), 'ff'(hex)
* or '00'. */
if (numdigits == 0 ||
((abs_nlz_bits != (size_t)(numbits-1) ||
!rb_absint_singlebit_p(val)) &&
(!bignum ? v < 0 : BIGNUM_NEGATIVE_P(val))))
numdigits++;
tmp = rb_str_new(NULL, numdigits);
valsign = rb_integer_pack(val, RSTRING_PTR(tmp), RSTRING_LEN(tmp),
1, CHAR_BIT-numbits, INTEGER_PACK_2COMP | INTEGER_PACK_BIG_ENDIAN);
for (i = 0; i < RSTRING_LEN(tmp); i++)
RSTRING_PTR(tmp)[i] = ruby_digitmap[((unsigned char *)RSTRING_PTR(tmp))[i]];
s = RSTRING_PTR(tmp);
dots = valsign < 0;
}
len = rb_long2int(RSTRING_END(tmp) - s);
}
else if (!bignum) {
valsign = 1;
if (v < 0) {
v = -v;
sc = '-';
width--;
valsign = -1;
}
else if (flags & FPLUS) {
sc = '+';
width--;
}
else if (flags & FSPACE) {
sc = ' ';
width--;
}
snprintf(nbuf, sizeof(nbuf), "%ld", v);
s = nbuf;
len = (int)strlen(s);
}
else {
tmp = rb_big2str(val, 10);
s = RSTRING_PTR(tmp);
valsign = 1;
if (s[0] == '-') {
s++;
sc = '-';
width--;
valsign = -1;
}
else if (flags & FPLUS) {
sc = '+';
width--;
}
else if (flags & FSPACE) {
sc = ' ';
width--;
}
len = rb_long2int(RSTRING_END(tmp) - s);
}
if (dots) {
prec -= 2;
width -= 2;
}
if (*p == 'X') {
char *pp = s;
int c;
while ((c = (int)(unsigned char)*pp) != 0) {
*pp = rb_enc_toupper(c, enc);
pp++;
}
}
if (prefix && !prefix[1]) { /* octal */
if (dots) {
prefix = 0;
}
else if (len == 1 && *s == '0') {
len = 0;
if (flags & FPREC) prec--;
}
else if ((flags & FPREC) && (prec > len)) {
prefix = 0;
}
}
else if (len == 1 && *s == '0') {
prefix = 0;
}
if (prefix) {
width -= (int)strlen(prefix);
}
if ((flags & (FZERO|FMINUS|FPREC)) == FZERO) {
prec = width;
width = 0;
}
else {
if (prec < len) {
if (!prefix && prec == 0 && len == 1 && *s == '0') len = 0;
prec = len;
}
width -= prec;
}
if (!(flags&FMINUS)) {
CHECK(width);
while (width-- > 0) {
buf[blen++] = ' ';
}
}
if (sc) PUSH(&sc, 1);
if (prefix) {
int plen = (int)strlen(prefix);
PUSH(prefix, plen);
}
CHECK(prec - len);
if (dots) PUSH("..", 2);
if (!sign && valsign < 0) {
char c = sign_bits(base, p);
while (len < prec--) {
buf[blen++] = c;
}
}
else if ((flags & (FMINUS|FPREC)) != FMINUS) {
while (len < prec--) {
buf[blen++] = '0';
}
}
PUSH(s, len);
RB_GC_GUARD(tmp);
CHECK(width);
while (width-- > 0) {
buf[blen++] = ' ';
}
}
break;
case 'f':
{
VALUE val = GETARG(), num, den;
int sign = (flags&FPLUS) ? 1 : 0, zero = 0;
long len, done = 0;
int prefix = 0;
if (!RB_TYPE_P(val, T_RATIONAL)) {
nextvalue = val;
goto float_value;
}
if (!(flags&FPREC)) prec = default_float_precision;
den = rb_rational_den(val);
num = rb_rational_num(val);
if (FIXNUM_P(num)) {
if ((SIGNED_VALUE)num < 0) {
long n = -FIX2LONG(num);
num = LONG2FIX(n);
sign = -1;
}
}
else if (rb_num_negative_p(num)) {
sign = -1;
num = rb_funcallv(num, idUMinus, 0, 0);
}
if (den != INT2FIX(1) || prec > 1) {
const ID idDiv = rb_intern("div");
VALUE p10 = rb_int_positive_pow(10, prec);
VALUE den_2 = rb_funcall(den, idDiv, 1, INT2FIX(2));
num = rb_funcallv(num, '*', 1, &p10);
num = rb_funcallv(num, '+', 1, &den_2);
num = rb_funcallv(num, idDiv, 1, &den);
}
else if (prec >= 0) {
zero = prec;
}
val = rb_obj_as_string(num);
len = RSTRING_LEN(val) + zero;
if (prec >= len) ++len; /* integer part 0 */
if (sign || (flags&FSPACE)) ++len;
if (prec > 0) ++len; /* period */
CHECK(len > width ? len : width);
if (sign || (flags&FSPACE)) {
buf[blen++] = sign > 0 ? '+' : sign < 0 ? '-' : ' ';
prefix++;
done++;
}
len = RSTRING_LEN(val) + zero;
t = RSTRING_PTR(val);
if (len > prec) {
memcpy(&buf[blen], t, len - prec);
blen += len - prec;
done += len - prec;
}
else {
buf[blen++] = '0';
done++;
}
if (prec > 0) {
buf[blen++] = '.';
done++;
}
if (zero) {
FILL('0', zero);
done += zero;
}
else if (prec > len) {
FILL('0', prec - len);
memcpy(&buf[blen], t, len);
blen += len;
done += prec;
}
else if (prec > 0) {
memcpy(&buf[blen], t + len - prec, prec);
blen += prec;
done += prec;
}
if ((flags & FWIDTH) && width > done) {
int fill = ' ';
long shifting = 0;
if (!(flags&FMINUS)) {
shifting = done;
if (flags&FZERO) {
shifting -= prefix;
fill = '0';
}
blen -= shifting;
memmove(&buf[blen + width - done], &buf[blen], shifting);
}
FILL(fill, width - done);
blen += shifting;
}
RB_GC_GUARD(val);
break;
}
case 'g':
case 'G':
case 'e':
case 'E':
/* TODO: rational support */
case 'a':
case 'A':
float_value:
{
VALUE val = GETARG();
double fval;
int i, need;
char fbuf[32];
fval = RFLOAT_VALUE(rb_Float(val));
if (isnan(fval) || isinf(fval)) {
const char *expr;
if (isnan(fval)) {
expr = "NaN";
}
else {
expr = "Inf";
}
need = (int)strlen(expr);
if ((!isnan(fval) && fval < 0.0) || (flags & FPLUS))
need++;
if ((flags & FWIDTH) && need < width)
need = width;
CHECK(need + 1);
snprintf(&buf[blen], need + 1, "%*s", need, "");
if (flags & FMINUS) {
if (!isnan(fval) && fval < 0.0)
buf[blen++] = '-';
else if (flags & FPLUS)
buf[blen++] = '+';
else if (flags & FSPACE)
blen++;
memcpy(&buf[blen], expr, strlen(expr));
}
else {
if (!isnan(fval) && fval < 0.0)
buf[blen + need - strlen(expr) - 1] = '-';
else if (flags & FPLUS)
buf[blen + need - strlen(expr) - 1] = '+';
else if ((flags & FSPACE) && need > width)
blen++;
memcpy(&buf[blen + need - strlen(expr)], expr,
strlen(expr));
}
blen += strlen(&buf[blen]);
break;
}
fmt_setup(fbuf, sizeof(fbuf), *p, flags, width, prec);
need = 0;
if (*p != 'e' && *p != 'E') {
i = INT_MIN;
frexp(fval, &i);
if (i > 0)
need = BIT_DIGITS(i);
}
need += (flags&FPREC) ? prec : default_float_precision;
if ((flags&FWIDTH) && need < width)
need = width;
need += 20;
CHECK(need);
snprintf(&buf[blen], need, fbuf, fval);
blen += strlen(&buf[blen]);
}
break;
}
flags = FNONE;
}
sprint_exit:
RB_GC_GUARD(fmt);
/* XXX - We cannot validate the number of arguments if (digit)$ style used.
*/
if (posarg >= 0 && nextarg < argc) {
const char *mesg = "too many arguments for format string";
if (RTEST(ruby_debug)) rb_raise(rb_eArgError, "%s", mesg);
if (RTEST(ruby_verbose)) rb_warn("%s", mesg);
}
rb_str_resize(result, blen);
if (tainted) OBJ_TAINT(result);
return result;
}
static VALUE encoding_spec_rb_enc_copy(VALUE self, VALUE dest, VALUE src) {
rb_enc_copy(dest, src);
return dest;
}
VALUE
rb_str_format(int argc, const VALUE *argv, VALUE fmt)
{
rb_encoding *enc;
const char *p, *end;
char *buf;
long blen, bsiz;
VALUE result;
long scanned = 0;
int coderange = ENC_CODERANGE_7BIT;
int width, prec, flags = FNONE;
int nextarg = 1;
int posarg = 0;
int tainted = 0;
VALUE nextvalue;
VALUE tmp;
VALUE str;
volatile VALUE hash = Qundef;
#define CHECK_FOR_WIDTH(f) \
if ((f) & FWIDTH) { \
rb_raise(rb_eArgError, "width given twice"); \
} \
if ((f) & FPREC0) { \
rb_raise(rb_eArgError, "width after precision"); \
}
#define CHECK_FOR_FLAGS(f) \
if ((f) & FWIDTH) { \
rb_raise(rb_eArgError, "flag after width"); \
} \
if ((f) & FPREC0) { \
rb_raise(rb_eArgError, "flag after precision"); \
}
++argc;
--argv;
if (OBJ_TAINTED(fmt)) tainted = 1;
StringValue(fmt);
enc = rb_enc_get(fmt);
fmt = rb_str_new4(fmt);
p = RSTRING_PTR(fmt);
end = p + RSTRING_LEN(fmt);
blen = 0;
bsiz = 120;
result = rb_str_buf_new(bsiz);
rb_enc_copy(result, fmt);
buf = RSTRING_PTR(result);
memset(buf, 0, bsiz);
ENC_CODERANGE_SET(result, coderange);
for (; p < end; p++) {
const char *t;
int n;
ID id = 0;
for (t = p; t < end && *t != '%'; t++) ;
PUSH(p, t - p);
if (coderange != ENC_CODERANGE_BROKEN && scanned < blen) {
scanned += rb_str_coderange_scan_restartable(buf+scanned, buf+blen, enc, &coderange);
ENC_CODERANGE_SET(result, coderange);
}
if (t >= end) {
/* end of fmt string */
goto sprint_exit;
}
p = t + 1; /* skip `%' */
width = prec = -1;
nextvalue = Qundef;
retry:
switch (*p) {
default:
if (rb_enc_isprint(*p, enc))
rb_raise(rb_eArgError, "malformed format string - %%%c", *p);
else
rb_raise(rb_eArgError, "malformed format string");
break;
case ' ':
CHECK_FOR_FLAGS(flags);
flags |= FSPACE;
p++;
goto retry;
case '#':
CHECK_FOR_FLAGS(flags);
flags |= FSHARP;
p++;
goto retry;
case '+':
CHECK_FOR_FLAGS(flags);
flags |= FPLUS;
p++;
goto retry;
case '-':
CHECK_FOR_FLAGS(flags);
flags |= FMINUS;
p++;
goto retry;
case '0':
CHECK_FOR_FLAGS(flags);
flags |= FZERO;
p++;
goto retry;
case '1': case '2': case '3': case '4':
case '5': case '6': case '7': case '8': case '9':
n = 0;
GETNUM(n, width);
if (*p == '$') {
if (nextvalue != Qundef) {
rb_raise(rb_eArgError, "value given twice - %d$", n);
}
nextvalue = GETPOSARG(n);
p++;
goto retry;
}
CHECK_FOR_WIDTH(flags);
width = n;
flags |= FWIDTH;
goto retry;
case '<':
case '{':
{
const char *start = p;
char term = (*p == '<') ? '>' : '}';
int len;
for (; p < end && *p != term; ) {
p += rb_enc_mbclen(p, end, enc);
}
if (p >= end) {
rb_raise(rb_eArgError, "malformed name - unmatched parenthesis");
}
#if SIZEOF_INT < SIZEOF_SIZE_T
if ((size_t)(p - start) >= INT_MAX) {
const int message_limit = 20;
len = (int)(rb_enc_right_char_head(start, start + message_limit, p, enc) - start);
rb_enc_raise(enc, rb_eArgError,
"too long name (%"PRIdSIZE" bytes) - %.*s...%c",
(size_t)(p - start - 2), len, start, term);
}
#endif
len = (int)(p - start + 1); /* including parenthesis */
if (id) {
rb_enc_raise(enc, rb_eArgError, "named%.*s after <%s>",
len, start, rb_id2name(id));
}
nextvalue = GETNAMEARG((id = rb_check_id_cstr(start + 1,
len - 2 /* without parenthesis */,
enc),
ID2SYM(id)),
start, len, enc);
if (nextvalue == Qundef) {
rb_enc_raise(enc, rb_eKeyError, "key%.*s not found", len, start);
}
if (term == '}') goto format_s;
p++;
goto retry;
}
case '*':
CHECK_FOR_WIDTH(flags);
flags |= FWIDTH;
GETASTER(width);
if (width < 0) {
flags |= FMINUS;
width = -width;
}
p++;
goto retry;
case '.':
if (flags & FPREC0) {
rb_raise(rb_eArgError, "precision given twice");
}
flags |= FPREC|FPREC0;
prec = 0;
p++;
if (*p == '*') {
GETASTER(prec);
if (prec < 0) { /* ignore negative precision */
flags &= ~FPREC;
}
p++;
goto retry;
}
GETNUM(prec, precision);
goto retry;
case '\n':
case '\0':
p--;
case '%':
if (flags != FNONE) {
rb_raise(rb_eArgError, "invalid format character - %%");
}
PUSH("%", 1);
break;
case 'c':
{
VALUE val = GETARG();
VALUE tmp;
unsigned int c;
int n;
tmp = rb_check_string_type(val);
if (!NIL_P(tmp)) {
if (rb_enc_strlen(RSTRING_PTR(tmp),RSTRING_END(tmp),enc) != 1) {
rb_raise(rb_eArgError, "%%c requires a character");
}
c = rb_enc_codepoint_len(RSTRING_PTR(tmp), RSTRING_END(tmp), &n, enc);
RB_GC_GUARD(tmp);
}
else {
c = NUM2INT(val);
n = rb_enc_codelen(c, enc);
}
if (n <= 0) {
rb_raise(rb_eArgError, "invalid character");
}
if (!(flags & FWIDTH)) {
CHECK(n);
rb_enc_mbcput(c, &buf[blen], enc);
blen += n;
}
else if ((flags & FMINUS)) {
CHECK(n);
rb_enc_mbcput(c, &buf[blen], enc);
blen += n;
FILL(' ', width-1);
}
else {
FILL(' ', width-1);
CHECK(n);
rb_enc_mbcput(c, &buf[blen], enc);
blen += n;
}
}
break;
case 's':
case 'p':
format_s:
{
VALUE arg = GETARG();
long len, slen;
if (*p == 'p') arg = rb_inspect(arg);
str = rb_obj_as_string(arg);
if (OBJ_TAINTED(str)) tainted = 1;
len = RSTRING_LEN(str);
rb_str_set_len(result, blen);
if (coderange != ENC_CODERANGE_BROKEN && scanned < blen) {
int cr = coderange;
scanned += rb_str_coderange_scan_restartable(buf+scanned, buf+blen, enc, &cr);
ENC_CODERANGE_SET(result,
(cr == ENC_CODERANGE_UNKNOWN ?
ENC_CODERANGE_BROKEN : (coderange = cr)));
}
enc = rb_enc_check(result, str);
if (flags&(FPREC|FWIDTH)) {
slen = rb_enc_strlen(RSTRING_PTR(str),RSTRING_END(str),enc);
if (slen < 0) {
rb_raise(rb_eArgError, "invalid mbstring sequence");
}
if ((flags&FPREC) && (prec < slen)) {
char *p = rb_enc_nth(RSTRING_PTR(str), RSTRING_END(str),
prec, enc);
slen = prec;
len = p - RSTRING_PTR(str);
}
/* need to adjust multi-byte string pos */
if ((flags&FWIDTH) && (width > slen)) {
width -= (int)slen;
if (!(flags&FMINUS)) {
CHECK(width);
while (width--) {
buf[blen++] = ' ';
}
}
CHECK(len);
memcpy(&buf[blen], RSTRING_PTR(str), len);
RB_GC_GUARD(str);
blen += len;
if (flags&FMINUS) {
CHECK(width);
while (width--) {
buf[blen++] = ' ';
}
}
rb_enc_associate(result, enc);
break;
}
}
PUSH(RSTRING_PTR(str), len);
RB_GC_GUARD(str);
rb_enc_associate(result, enc);
}
break;
case 'd':
case 'i':
case 'o':
case 'x':
case 'X':
case 'b':
case 'B':
case 'u':
{
volatile VALUE val = GETARG();
char fbuf[32], nbuf[64], *s;
const char *prefix = 0;
int sign = 0, dots = 0;
char sc = 0;
long v = 0;
int base, bignum = 0;
int len;
switch (*p) {
case 'd':
case 'i':
case 'u':
sign = 1; break;
case 'o':
case 'x':
case 'X':
case 'b':
case 'B':
if (flags&(FPLUS|FSPACE)) sign = 1;
break;
}
if (flags & FSHARP) {
switch (*p) {
case 'o':
prefix = "0"; break;
case 'x':
prefix = "0x"; break;
case 'X':
prefix = "0X"; break;
case 'b':
prefix = "0b"; break;
case 'B':
prefix = "0B"; break;
}
}
bin_retry:
switch (TYPE(val)) {
case T_FLOAT:
if (FIXABLE(RFLOAT_VALUE(val))) {
val = LONG2FIX((long)RFLOAT_VALUE(val));
goto bin_retry;
}
val = rb_dbl2big(RFLOAT_VALUE(val));
if (FIXNUM_P(val)) goto bin_retry;
bignum = 1;
break;
case T_STRING:
val = rb_str_to_inum(val, 0, TRUE);
goto bin_retry;
case T_BIGNUM:
bignum = 1;
break;
case T_FIXNUM:
v = FIX2LONG(val);
break;
default:
val = rb_Integer(val);
goto bin_retry;
}
switch (*p) {
case 'o':
base = 8; break;
case 'x':
case 'X':
base = 16; break;
case 'b':
case 'B':
base = 2; break;
case 'u':
case 'd':
case 'i':
default:
base = 10; break;
}
if (!bignum) {
if (base == 2) {
val = rb_int2big(v);
goto bin_retry;
}
if (sign) {
char c = *p;
if (c == 'i') c = 'd'; /* %d and %i are identical */
if (v < 0) {
v = -v;
sc = '-';
width--;
}
else if (flags & FPLUS) {
sc = '+';
width--;
}
else if (flags & FSPACE) {
sc = ' ';
width--;
}
snprintf(fbuf, sizeof(fbuf), "%%l%c", c);
snprintf(nbuf, sizeof(nbuf), fbuf, v);
s = nbuf;
}
else {
s = nbuf;
if (v < 0) {
dots = 1;
}
snprintf(fbuf, sizeof(fbuf), "%%l%c", *p == 'X' ? 'x' : *p);
snprintf(++s, sizeof(nbuf) - 1, fbuf, v);
if (v < 0) {
char d = 0;
s = remove_sign_bits(s, base);
switch (base) {
case 16:
d = 'f'; break;
case 8:
d = '7'; break;
}
if (d && *s != d) {
*--s = d;
}
}
}
len = (int)strlen(s);
}
else {
if (sign) {
tmp = rb_big2str(val, base);
s = RSTRING_PTR(tmp);
if (s[0] == '-') {
s++;
sc = '-';
width--;
}
else if (flags & FPLUS) {
sc = '+';
width--;
}
else if (flags & FSPACE) {
sc = ' ';
width--;
}
}
else {
if (!RBIGNUM_SIGN(val)) {
val = rb_big_clone(val);
rb_big_2comp(val);
}
tmp = rb_big2str0(val, base, RBIGNUM_SIGN(val));
s = RSTRING_PTR(tmp);
if (*s == '-') {
dots = 1;
if (base == 10) {
rb_warning("negative number for %%u specifier");
}
s = remove_sign_bits(++s, base);
switch (base) {
case 16:
if (s[0] != 'f') *--s = 'f'; break;
case 8:
if (s[0] != '7') *--s = '7'; break;
case 2:
if (s[0] != '1') *--s = '1'; break;
}
}
}
len = rb_long2int(RSTRING_END(tmp) - s);
}
if (dots) {
prec -= 2;
width -= 2;
}
if (*p == 'X') {
char *pp = s;
int c;
while ((c = (int)(unsigned char)*pp) != 0) {
*pp = rb_enc_toupper(c, enc);
pp++;
}
}
if (prefix && !prefix[1]) { /* octal */
if (dots) {
prefix = 0;
}
else if (len == 1 && *s == '0') {
len = 0;
if (flags & FPREC) prec--;
}
else if ((flags & FPREC) && (prec > len)) {
prefix = 0;
}
}
else if (len == 1 && *s == '0') {
prefix = 0;
}
if (prefix) {
width -= (int)strlen(prefix);
}
if ((flags & (FZERO|FMINUS|FPREC)) == FZERO) {
prec = width;
width = 0;
}
else {
if (prec < len) {
if (!prefix && prec == 0 && len == 1 && *s == '0') len = 0;
prec = len;
}
width -= prec;
}
if (!(flags&FMINUS)) {
CHECK(width);
while (width-- > 0) {
buf[blen++] = ' ';
}
}
if (sc) PUSH(&sc, 1);
if (prefix) {
int plen = (int)strlen(prefix);
PUSH(prefix, plen);
}
CHECK(prec - len);
if (dots) PUSH("..", 2);
if (!bignum && v < 0) {
char c = sign_bits(base, p);
while (len < prec--) {
buf[blen++] = c;
}
}
else if ((flags & (FMINUS|FPREC)) != FMINUS) {
char c;
if (!sign && bignum && !RBIGNUM_SIGN(val))
c = sign_bits(base, p);
else
c = '0';
while (len < prec--) {
buf[blen++] = c;
}
}
PUSH(s, len);
RB_GC_GUARD(tmp);
CHECK(width);
while (width-- > 0) {
buf[blen++] = ' ';
}
}
break;
case 'f':
case 'g':
case 'G':
case 'e':
case 'E':
case 'a':
case 'A':
{
VALUE val = GETARG();
double fval;
int i, need = 6;
char fbuf[32];
fval = RFLOAT_VALUE(rb_Float(val));
if (isnan(fval) || isinf(fval)) {
const char *expr;
if (isnan(fval)) {
expr = "NaN";
}
else {
expr = "Inf";
}
need = (int)strlen(expr);
if ((!isnan(fval) && fval < 0.0) || (flags & FPLUS))
need++;
if ((flags & FWIDTH) && need < width)
need = width;
CHECK(need + 1);
snprintf(&buf[blen], need + 1, "%*s", need, "");
if (flags & FMINUS) {
if (!isnan(fval) && fval < 0.0)
buf[blen++] = '-';
else if (flags & FPLUS)
buf[blen++] = '+';
else if (flags & FSPACE)
blen++;
memcpy(&buf[blen], expr, strlen(expr));
}
else {
if (!isnan(fval) && fval < 0.0)
buf[blen + need - strlen(expr) - 1] = '-';
else if (flags & FPLUS)
buf[blen + need - strlen(expr) - 1] = '+';
else if ((flags & FSPACE) && need > width)
blen++;
memcpy(&buf[blen + need - strlen(expr)], expr,
strlen(expr));
}
blen += strlen(&buf[blen]);
break;
}
fmt_setup(fbuf, sizeof(fbuf), *p, flags, width, prec);
need = 0;
if (*p != 'e' && *p != 'E') {
i = INT_MIN;
frexp(fval, &i);
if (i > 0)
need = BIT_DIGITS(i);
}
need += (flags&FPREC) ? prec : 6;
if ((flags&FWIDTH) && need < width)
need = width;
need += 20;
CHECK(need);
snprintf(&buf[blen], need, fbuf, fval);
blen += strlen(&buf[blen]);
}
break;
}
flags = FNONE;
}
sprint_exit:
RB_GC_GUARD(fmt);
/* XXX - We cannot validate the number of arguments if (digit)$ style used.
*/
if (posarg >= 0 && nextarg < argc) {
const char *mesg = "too many arguments for format string";
if (RTEST(ruby_debug)) rb_raise(rb_eArgError, "%s", mesg);
if (RTEST(ruby_verbose)) rb_warn("%s", mesg);
}
rb_str_resize(result, blen);
if (tainted) OBJ_TAINT(result);
return result;
}
/*
* Document-method: auto_link
*
* call-seq:
* auto_link(text, mode=:all, link_attr=nil, skip_tags=nil, flags=0)
* auto_link(text, mode=:all, link_attr=nil, skip_tags=nil, flags=0) { |link_text| ... }
*
* Parses a block of text looking for "safe" urls or email addresses,
* and turns them into HTML links with the given attributes.
*
* NOTE: The block of text may or may not be HTML; if the text is HTML,
* Rinku will skip the relevant tags to prevent double-linking and linking
* inside `pre` blocks by default.
*
* NOTE: If the input text is HTML, it's expected to be already escaped.
* Rinku will perform no escaping.
*
* NOTE: Currently the follow protocols are considered safe and are the
* only ones that will be autolinked.
*
* http:// https:// ftp:// mailto://
*
* Email addresses are also autolinked by default. URLs without a protocol
* specifier but starting with 'www.' will also be autolinked, defaulting to
* the 'http://' protocol.
*
* - `text` is a string in plain text or HTML markup. If the string is formatted in
* HTML, Rinku is smart enough to skip the links that are already enclosed in `<a>`
* tags.`
*
* - `mode` is a symbol, either `:all`, `:urls` or `:email_addresses`,
* which specifies which kind of links will be auto-linked.
*
* - `link_attr` is a string containing the link attributes for each link that
* will be generated. These attributes are not sanitized and will be include as-is
* in each generated link, e.g.
*
* ~~~~~ruby
* auto_link('http://www.pokemon.com', :all, 'target="_blank"')
* # => '<a href="http://www.pokemon.com" target="_blank">http://www.pokemon.com</a>'
* ~~~~~
*
* This string can be autogenerated from a hash using the Rails `tag_options` helper.
*
* - `skip_tags` is a list of strings with the names of HTML tags that will be skipped
* when autolinking. If `nil`, this defaults to the value of the global `Rinku.skip_tags`,
* which is initially `["a", "pre", "code", "kbd", "script"]`.
*
* - `flag` is an optional boolean value specifying whether to recognize
* 'http://foo' as a valid domain, or require at least one '.'. It defaults to false.
*
* - `&block` is an optional block argument. If a block is passed, it will
* be yielded for each found link in the text, and its return value will be used instead
* of the name of the link. E.g.
*
* ~~~~~ruby
* auto_link('Check it out at http://www.pokemon.com') do |url|
* "THE POKEMAN WEBSITEZ"
* end
* # => 'Check it out at <a href="http://www.pokemon.com">THE POKEMAN WEBSITEZ</a>'
* ~~~~~~
*/
static VALUE
rb_rinku_autolink(int argc, VALUE *argv, VALUE self)
{
static const char *SKIP_TAGS[] = {"a", "pre", "code", "kbd", "script", NULL};
VALUE result, rb_text, rb_mode, rb_html, rb_skip, rb_flags, rb_block;
struct buf *output_buf;
int link_mode, count;
unsigned int link_flags = 0;
const char *link_attr = NULL;
const char **skip_tags = NULL;
ID mode_sym;
rb_scan_args(argc, argv, "14&", &rb_text, &rb_mode,
&rb_html, &rb_skip, &rb_flags, &rb_block);
Check_Type(rb_text, T_STRING);
if (!NIL_P(rb_mode)) {
Check_Type(rb_mode, T_SYMBOL);
mode_sym = SYM2ID(rb_mode);
} else {
mode_sym = rb_intern("all");
}
if (!NIL_P(rb_html)) {
Check_Type(rb_html, T_STRING);
link_attr = RSTRING_PTR(rb_html);
}
if (NIL_P(rb_skip))
rb_skip = rb_iv_get(self, "@skip_tags");
if (NIL_P(rb_skip)) {
skip_tags = SKIP_TAGS;
} else {
skip_tags = rinku_load_tags(rb_skip);
}
if (!NIL_P(rb_flags)) {
Check_Type(rb_flags, T_FIXNUM);
link_flags = FIX2INT(rb_flags);
}
output_buf = bufnew(32);
if (mode_sym == rb_intern("all"))
link_mode = AUTOLINK_ALL;
else if (mode_sym == rb_intern("email_addresses"))
link_mode = AUTOLINK_EMAILS;
else if (mode_sym == rb_intern("urls"))
link_mode = AUTOLINK_URLS;
else
rb_raise(rb_eTypeError,
"Invalid linking mode (possible values are :all, :urls, :email_addresses)");
count = rinku_autolink(
output_buf,
RSTRING_PTR(rb_text),
RSTRING_LEN(rb_text),
link_mode,
link_flags,
link_attr,
skip_tags,
RTEST(rb_block) ? &autolink_callback : NULL,
(void*)rb_block);
if (count == 0)
result = rb_text;
else {
result = rb_str_new(output_buf->data, output_buf->size);
rb_enc_copy(result, rb_text);
}
if (skip_tags != SKIP_TAGS)
xfree(skip_tags);
bufrelease(output_buf);
return result;
}
