1
0
Fork 0
mirror of https://github.com/ruby/ruby.git synced 2022-11-09 12:17:21 -05:00
ruby--ruby/marshal.c

2377 lines
56 KiB
C
Raw Normal View History

/**********************************************************************
marshal.c -
$Author$
created at: Thu Apr 27 16:30:01 JST 1995
* encoding.c: provide basic features for M17N. * parse.y: encoding aware parsing. * parse.y (pragma_encoding): encoding specification pragma. * parse.y (rb_intern3): encoding specified symbols. * string.c (rb_str_length): length based on characters. for older behavior, bytesize method added. * string.c (rb_str_index_m): index based on characters. rindex as well. * string.c (succ_char): encoding aware succeeding string. * string.c (rb_str_reverse): reverse based on characters. * string.c (rb_str_inspect): encoding aware string description. * string.c (rb_str_upcase_bang): encoding aware case conversion. downcase, capitalize, swapcase as well. * string.c (rb_str_tr_bang): tr based on characters. delete, squeeze, tr_s, count as well. * string.c (rb_str_split_m): split based on characters. * string.c (rb_str_each_line): encoding aware each_line. * string.c (rb_str_each_char): added. iteration based on characters. * string.c (rb_str_strip_bang): encoding aware whitespace stripping. lstrip, rstrip as well. * string.c (rb_str_justify): encoding aware justifying (ljust, rjust, center). * string.c (str_encoding): get encoding attribute from a string. * re.c (rb_reg_initialize): encoding aware regular expression * sprintf.c (rb_str_format): formatting (i.e. length count) based on characters. * io.c (rb_io_getc): getc to return one-character string. for older behavior, getbyte method added. * ext/stringio/stringio.c (strio_getc): ditto. * io.c (rb_io_ungetc): allow pushing arbitrary string at the current reading point. * ext/stringio/stringio.c (strio_ungetc): ditto. * ext/strscan/strscan.c: encoding support. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@13261 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2007-08-24 23:29:39 -04:00
Copyright (C) 1993-2007 Yukihiro Matsumoto
**********************************************************************/
#include "ruby/internal/config.h"
#include <math.h>
#ifdef HAVE_FLOAT_H
#include <float.h>
#endif
#ifdef HAVE_IEEEFP_H
#include <ieeefp.h>
#endif
#include "encindex.h"
#include "id_table.h"
#include "internal.h"
#include "internal/array.h"
#include "internal/bignum.h"
#include "internal/class.h"
#include "internal/encoding.h"
#include "internal/error.h"
#include "internal/hash.h"
#include "internal/object.h"
#include "internal/struct.h"
#include "internal/util.h"
#include "internal/vm.h"
#include "ruby/io.h"
#include "ruby/ruby.h"
#include "ruby/st.h"
#include "ruby/util.h"
#define BITSPERSHORT (2*CHAR_BIT)
#define SHORTMASK ((1<<BITSPERSHORT)-1)
#define SHORTDN(x) RSHIFT((x),BITSPERSHORT)
#if SIZEOF_SHORT == SIZEOF_BDIGIT
#define SHORTLEN(x) (x)
#else
static size_t
shortlen(size_t len, BDIGIT *ds)
{
BDIGIT num;
int offset = 0;
num = ds[len-1];
while (num) {
num = SHORTDN(num);
offset++;
}
return (len - 1)*SIZEOF_BDIGIT/2 + offset;
}
#define SHORTLEN(x) shortlen((x),d)
#endif
#define MARSHAL_MAJOR 4
#define MARSHAL_MINOR 8
#define TYPE_NIL '0'
#define TYPE_TRUE 'T'
#define TYPE_FALSE 'F'
#define TYPE_FIXNUM 'i'
#define TYPE_EXTENDED 'e'
#define TYPE_UCLASS 'C'
#define TYPE_OBJECT 'o'
#define TYPE_DATA 'd'
#define TYPE_USERDEF 'u'
#define TYPE_USRMARSHAL 'U'
#define TYPE_FLOAT 'f'
#define TYPE_BIGNUM 'l'
#define TYPE_STRING '"'
#define TYPE_REGEXP '/'
#define TYPE_ARRAY '['
#define TYPE_HASH '{'
#define TYPE_HASH_DEF '}'
#define TYPE_STRUCT 'S'
#define TYPE_MODULE_OLD 'M'
#define TYPE_CLASS 'c'
#define TYPE_MODULE 'm'
#define TYPE_SYMBOL ':'
#define TYPE_SYMLINK ';'
#define TYPE_IVAR 'I'
#define TYPE_LINK '@'
static ID s_dump, s_load, s_mdump, s_mload;
static ID s_dump_data, s_load_data, s_alloc, s_call;
static ID s_getbyte, s_read, s_write, s_binmode;
static ID s_encoding_short, s_ruby2_keywords_flag;
#define name_s_dump "_dump"
#define name_s_load "_load"
#define name_s_mdump "marshal_dump"
#define name_s_mload "marshal_load"
#define name_s_dump_data "_dump_data"
#define name_s_load_data "_load_data"
#define name_s_alloc "_alloc"
#define name_s_call "call"
#define name_s_getbyte "getbyte"
#define name_s_read "read"
#define name_s_write "write"
#define name_s_binmode "binmode"
#define name_s_encoding_short "E"
#define name_s_ruby2_keywords_flag "K"
typedef struct {
VALUE newclass;
VALUE oldclass;
VALUE (*dumper)(VALUE);
VALUE (*loader)(VALUE, VALUE);
} marshal_compat_t;
static st_table *compat_allocator_tbl;
static VALUE compat_allocator_tbl_wrapper;
static VALUE rb_marshal_dump_limited(VALUE obj, VALUE port, int limit);
static VALUE rb_marshal_load_with_proc(VALUE port, VALUE proc);
static int
mark_marshal_compat_i(st_data_t key, st_data_t value, st_data_t _)
{
marshal_compat_t *p = (marshal_compat_t *)value;
rb_gc_mark(p->newclass);
rb_gc_mark(p->oldclass);
return ST_CONTINUE;
}
static void
mark_marshal_compat_t(void *tbl)
{
if (!tbl) return;
st_foreach(tbl, mark_marshal_compat_i, 0);
}
static st_table *compat_allocator_table(void);
void
rb_marshal_define_compat(VALUE newclass, VALUE oldclass, VALUE (*dumper)(VALUE), VALUE (*loader)(VALUE, VALUE))
{
marshal_compat_t *compat;
rb_alloc_func_t allocator = rb_get_alloc_func(newclass);
if (!allocator) {
rb_raise(rb_eTypeError, "no allocator");
}
compat = ALLOC(marshal_compat_t);
compat->newclass = Qnil;
compat->oldclass = Qnil;
compat->newclass = newclass;
compat->oldclass = oldclass;
compat->dumper = dumper;
compat->loader = loader;
st_insert(compat_allocator_table(), (st_data_t)allocator, (st_data_t)compat);
}
struct dump_arg {
VALUE str, dest;
st_table *symbols;
st_table *data;
st_table *compat_tbl;
st_table *encodings;
};
struct dump_call_arg {
VALUE obj;
struct dump_arg *arg;
int limit;
};
static VALUE
check_dump_arg(VALUE ret, struct dump_arg *arg, const char *name)
{
if (!arg->symbols) {
rb_raise(rb_eRuntimeError, "Marshal.dump reentered at %s",
name);
}
return ret;
}
static VALUE
check_userdump_arg(VALUE obj, ID sym, int argc, const VALUE *argv,
struct dump_arg *arg, const char *name)
{
VALUE ret = rb_funcallv(obj, sym, argc, argv);
VALUE klass = CLASS_OF(obj);
if (CLASS_OF(ret) == klass) {
rb_raise(rb_eRuntimeError, "%"PRIsVALUE"#%s returned same class instance",
klass, name);
}
return check_dump_arg(ret, arg, name);
}
#define dump_funcall(arg, obj, sym, argc, argv) \
check_userdump_arg(obj, sym, argc, argv, arg, name_##sym)
#define dump_check_funcall(arg, obj, sym, argc, argv) \
check_dump_arg(rb_check_funcall(obj, sym, argc, argv), arg, name_##sym)
static void clear_dump_arg(struct dump_arg *arg);
static void
mark_dump_arg(void *ptr)
{
struct dump_arg *p = ptr;
if (!p->symbols)
return;
rb_mark_set(p->symbols);
rb_mark_set(p->data);
rb_mark_hash(p->compat_tbl);
rb_gc_mark(p->str);
}
static void
free_dump_arg(void *ptr)
{
clear_dump_arg(ptr);
xfree(ptr);
}
static size_t
memsize_dump_arg(const void *ptr)
{
return sizeof(struct dump_arg);
}
static const rb_data_type_t dump_arg_data = {
"dump_arg",
{mark_dump_arg, free_dump_arg, memsize_dump_arg,},
0, 0, RUBY_TYPED_FREE_IMMEDIATELY
};
static VALUE
must_not_be_anonymous(const char *type, VALUE path)
{
char *n = RSTRING_PTR(path);
if (!rb_enc_asciicompat(rb_enc_get(path))) {
/* cannot occur? */
rb_raise(rb_eTypeError, "can't dump non-ascii %s name % "PRIsVALUE,
type, path);
}
if (n[0] == '#') {
rb_raise(rb_eTypeError, "can't dump anonymous %s % "PRIsVALUE,
type, path);
}
return path;
}
static VALUE
class2path(VALUE klass)
{
VALUE path = rb_class_path(klass);
must_not_be_anonymous((RB_TYPE_P(klass, T_CLASS) ? "class" : "module"), path);
if (rb_path_to_class(path) != rb_class_real(klass)) {
rb_raise(rb_eTypeError, "% "PRIsVALUE" can't be referred to", path);
}
return path;
}
int ruby_marshal_write_long(long x, char *buf);
static void w_long(long, struct dump_arg*);
static int w_encoding(VALUE encname, struct dump_call_arg *arg);
static VALUE encoding_name(VALUE obj, struct dump_arg *arg);
static void
w_nbyte(const char *s, long n, struct dump_arg *arg)
{
VALUE buf = arg->str;
rb_str_buf_cat(buf, s, n);
if (arg->dest && RSTRING_LEN(buf) >= BUFSIZ) {
rb_io_write(arg->dest, buf);
rb_str_resize(buf, 0);
}
}
static void
w_byte(char c, struct dump_arg *arg)
{
w_nbyte(&c, 1, arg);
}
static void
w_bytes(const char *s, long n, struct dump_arg *arg)
{
w_long(n, arg);
w_nbyte(s, n, arg);
}
#define w_cstr(s, arg) w_bytes((s), strlen(s), (arg))
static void
w_short(int x, struct dump_arg *arg)
{
w_byte((char)((x >> 0) & 0xff), arg);
w_byte((char)((x >> 8) & 0xff), arg);
}
static void
w_long(long x, struct dump_arg *arg)
{
char buf[sizeof(long)+1];
int i = ruby_marshal_write_long(x, buf);
if (i < 0) {
rb_raise(rb_eTypeError, "long too big to dump");
}
w_nbyte(buf, i, arg);
}
int
ruby_marshal_write_long(long x, char *buf)
{
int i;
#if SIZEOF_LONG > 4
if (!(RSHIFT(x, 31) == 0 || RSHIFT(x, 31) == -1)) {
/* big long does not fit in 4 bytes */
return -1;
}
#endif
if (x == 0) {
buf[0] = 0;
return 1;
}
if (0 < x && x < 123) {
buf[0] = (char)(x + 5);
return 1;
}
if (-124 < x && x < 0) {
buf[0] = (char)((x - 5)&0xff);
return 1;
}
for (i=1;i<(int)sizeof(long)+1;i++) {
buf[i] = (char)(x & 0xff);
x = RSHIFT(x,8);
if (x == 0) {
buf[0] = i;
break;
}
if (x == -1) {
buf[0] = -i;
break;
}
}
return i+1;
}
#ifdef DBL_MANT_DIG
#define DECIMAL_MANT (53-16) /* from IEEE754 double precision */
#if DBL_MANT_DIG > 32
#define MANT_BITS 32
#elif DBL_MANT_DIG > 24
#define MANT_BITS 24
#elif DBL_MANT_DIG > 16
#define MANT_BITS 16
#else
#define MANT_BITS 8
#endif
static double
load_mantissa(double d, const char *buf, long len)
{
if (!len) return d;
if (--len > 0 && !*buf++) { /* binary mantissa mark */
int e, s = d < 0, dig = 0;
unsigned long m;
modf(ldexp(frexp(fabs(d), &e), DECIMAL_MANT), &d);
do {
m = 0;
switch (len) {
2019-07-14 01:04:34 -04:00
default: m = *buf++ & 0xff; /* fall through */
#if MANT_BITS > 24
2019-07-14 01:04:34 -04:00
case 3: m = (m << 8) | (*buf++ & 0xff); /* fall through */
#endif
#if MANT_BITS > 16
2019-07-14 01:04:34 -04:00
case 2: m = (m << 8) | (*buf++ & 0xff); /* fall through */
#endif
#if MANT_BITS > 8
case 1: m = (m << 8) | (*buf++ & 0xff);
#endif
}
dig -= len < MANT_BITS / 8 ? 8 * (unsigned)len : MANT_BITS;
d += ldexp((double)m, dig);
} while ((len -= MANT_BITS / 8) > 0);
d = ldexp(d, e - DECIMAL_MANT);
if (s) d = -d;
}
return d;
}
#else
#define load_mantissa(d, buf, len) (d)
#endif
#ifdef DBL_DIG
#define FLOAT_DIG (DBL_DIG+2)
#else
#define FLOAT_DIG 17
#endif
static void
w_float(double d, struct dump_arg *arg)
{
char buf[FLOAT_DIG + (DECIMAL_MANT + 7) / 8 + 10];
if (isinf(d)) {
if (d < 0) w_cstr("-inf", arg);
else w_cstr("inf", arg);
}
else if (isnan(d)) {
w_cstr("nan", arg);
}
else if (d == 0.0) {
if (signbit(d)) w_cstr("-0", arg);
else w_cstr("0", arg);
}
else {
int decpt, sign, digs, len = 0;
char *e, *p = ruby_dtoa(d, 0, 0, &decpt, &sign, &e);
if (sign) buf[len++] = '-';
digs = (int)(e - p);
if (decpt < -3 || decpt > digs) {
buf[len++] = p[0];
if (--digs > 0) buf[len++] = '.';
memcpy(buf + len, p + 1, digs);
len += digs;
len += snprintf(buf + len, sizeof(buf) - len, "e%d", decpt - 1);
}
else if (decpt > 0) {
memcpy(buf + len, p, decpt);
len += decpt;
if ((digs -= decpt) > 0) {
buf[len++] = '.';
memcpy(buf + len, p + decpt, digs);
len += digs;
}
}
else {
buf[len++] = '0';
buf[len++] = '.';
if (decpt) {
memset(buf + len, '0', -decpt);
len -= decpt;
}
memcpy(buf + len, p, digs);
len += digs;
}
xfree(p);
w_bytes(buf, len, arg);
}
}
static void
w_symbol(VALUE sym, struct dump_arg *arg)
{
st_data_t num;
VALUE encname;
if (st_lookup(arg->symbols, sym, &num)) {
w_byte(TYPE_SYMLINK, arg);
w_long((long)num, arg);
}
else {
const VALUE orig_sym = sym;
sym = rb_sym2str(sym);
if (!sym) {
rb_raise(rb_eTypeError, "can't dump anonymous ID %"PRIdVALUE, sym);
}
encname = encoding_name(sym, arg);
if (NIL_P(encname) ||
rb_enc_str_coderange(sym) == ENC_CODERANGE_7BIT) {
encname = Qnil;
}
else {
w_byte(TYPE_IVAR, arg);
}
w_byte(TYPE_SYMBOL, arg);
w_bytes(RSTRING_PTR(sym), RSTRING_LEN(sym), arg);
st_add_direct(arg->symbols, orig_sym, arg->symbols->num_entries);
if (!NIL_P(encname)) {
struct dump_call_arg c_arg;
c_arg.limit = 1;
c_arg.arg = arg;
w_long(1L, arg);
w_encoding(encname, &c_arg);
}
}
}
static void
w_unique(VALUE s, struct dump_arg *arg)
{
must_not_be_anonymous("class", s);
w_symbol(rb_str_intern(s), arg);
}
static void w_object(VALUE,struct dump_arg*,int);
static int
hash_each(VALUE key, VALUE value, VALUE v)
{
struct dump_call_arg *arg = (void *)v;
w_object(key, arg->arg, arg->limit);
w_object(value, arg->arg, arg->limit);
return ST_CONTINUE;
}
#define SINGLETON_DUMP_UNABLE_P(klass) \
(rb_id_table_size(RCLASS_M_TBL(klass)) > 0 || \
(RCLASS_IV_TBL(klass) && RCLASS_IV_TBL(klass)->num_entries > 1))
static void
w_extended(VALUE klass, struct dump_arg *arg, int check)
{
if (check && FL_TEST(klass, FL_SINGLETON)) {
VALUE origin = RCLASS_ORIGIN(klass);
if (SINGLETON_DUMP_UNABLE_P(klass) ||
(origin != klass && SINGLETON_DUMP_UNABLE_P(origin))) {
rb_raise(rb_eTypeError, "singleton can't be dumped");
}
klass = RCLASS_SUPER(klass);
}
while (BUILTIN_TYPE(klass) == T_ICLASS) {
Ensure origins for all included, prepended, and refined modules This fixes various issues when a module is included in or prepended to a module or class, and then refined, or refined and then included or prepended to a module or class. Implement by renaming ensure_origin to rb_ensure_origin, making it non-static, and calling it when refining a module. Fix Module#initialize_copy to handle origins correctly. Previously, Module#initialize_copy did not handle origins correctly. For example, this code: ```ruby module B; end class A def b; 2 end prepend B end a = A.dup.new class A def b; 1 end end p a.b ``` Printed 1 instead of 2. This is because the super chain for a.singleton_class was: ``` a.singleton_class A.dup B(iclass) B(iclass origin) A(origin) # not A.dup(origin) ``` The B iclasses would not be modified, so the includer entry would be still be set to A and not A.dup. This modifies things so that if the class/module has an origin, all iclasses between the class/module and the origin are duplicated and have the correct includer entry set, and the correct origin is created. This requires other changes to make sure all tests still pass: * rb_undef_methods_from doesn't automatically handle classes with origins, so pass it the origin for Comparable when undefing methods in Complex. This fixed a failure in the Complex tests. * When adding a method, the method cache was not cleared correctly if klass has an origin. Clear the method cache for the klass before switching to the origin of klass. This fixed failures in the autoload tests related to overridding require, without breaking the optimization tests. Also clear the method cache for both the module and origin when removing a method. * Module#include? is fixed to skip origin iclasses. * Refinements are fixed to use the origin class of the module that has an origin. * RCLASS_REFINED_BY_ANY is removed as it was only used in a single place and is no longer needed. * Marshal#dump is fixed to skip iclass origins. * rb_method_entry_make is fixed to handled overridden optimized methods for modules that have origins. Fixes [Bug #16852]
2020-05-23 23:16:27 -04:00
if (!FL_TEST(klass, RICLASS_IS_ORIGIN) ||
BUILTIN_TYPE(RBASIC(klass)->klass) != T_MODULE) {
VALUE path = rb_class_name(RBASIC(klass)->klass);
w_byte(TYPE_EXTENDED, arg);
w_unique(path, arg);
}
klass = RCLASS_SUPER(klass);
}
}
static void
w_class(char type, VALUE obj, struct dump_arg *arg, int check)
{
VALUE path;
st_data_t real_obj;
VALUE klass;
if (arg->compat_tbl &&
st_lookup(arg->compat_tbl, (st_data_t)obj, &real_obj)) {
obj = (VALUE)real_obj;
}
klass = CLASS_OF(obj);
* gc.c (Init_stack): stack region is far smaller than usual if pthread is used. * marshal.c (w_extended): singleton methods should not be checked when dumping via marshal_dump() or _dump(). [ruby-talk:85909] * file.c (getcwdofdrv): avoid using getcwd() directly, use my_getcwd() instead. * merged NeXT, OpenStep, Rhapsody ports patch from Eric Sunshine <sunshine@sunshineco.com>. [ruby-core:01596] * marshal.c (w_object): LINK check earlier than anything else, i.e. do not dump TYPE_IVAR for already dumped objects. (ruby-bugs PR#1220) * eval.c (rb_eval): call "inherited" only when a new class is generated; not on reopening. * eval.c (eval): prepend error position in evaluating string to * configure.in: revived NextStep, OpenStep, and Rhapsody ports which had become unbuildable; enhanced --enable-fat-binary option so that it accepts a list of desired architectures (rather than assuming a fixed list), or defaults to a platform-appropriate list if user does not provide an explicit list; made the default list of architectures for MAB (fat binary) more comprehensive; now uses -fno-common even when building the interpreter (in addition to using it for extensions), thus allowing the interpreter to be embedded into a plugin module of an external project (in addition to allowing embedding directly into an application); added checks for <netinet/in_systm.h> (needed by `socket' extension) and getcwd(); now ensures that -I/usr/local/include is employed when extensions' extconf.rb scripts invoke have_header() since extension checks on NextStep and OpenStep will fail without it if the desired resource resides in the /usr/local tree; fixed formatting of --help message. * Makefile.in: $(LIBRUBY_A) rule now deletes the archive before invoking $(AR) since `ar' on Apple/NeXT can not "update" MAB archives (see configure's --enable-fat-binary option); added rule for new missing/getcwd.c. * defines.h: fixed endian handling during MAB build (see configure's --enable-fat-binary option) to ensure that all portions of the project see the correct WORDS_BIGENDIAN value (some extension modules were getting the wrong endian setting); added missing constants GETPGRP_VOID, WNOHANG, WUNTRACED, X_OK, and type pid_t for NextStep and OpenStep; removed unnecessary and problematic HAVE_SYS_WAIT_H define in NeXT section. * dir.c: do not allow NAMLEN() macro to trust dirent::d_namlen on NextStep since, on some installations, this value always resolves uselessly to zero. * dln.c: added error reporting to NextStep extension loader since the previous behavior of failing silently was not useful; now ensures that NSLINKMODULE_OPTION_BINDNOW compatibility constant is defined for OpenStep and Rhapsody; no longer includes <mach-o/dyld.h> twice on Rhapsody since this header lacks multiple-include protection, which resulted in "redefinition" compilation errors. * main.c: also create hard reference to objc_msgSend() on NeXT platforms (in addition to Apple platforms). * lib/mkmf.rb: now exports XCFLAGS from configure script to extension makefiles so that extensions can be built MAB (see configure's --enable-fat-binary option); also utilize XCFLAGS in cc_command() (but not cpp_command() because MAB flags are incompatible with direct invocation of `cpp'). * ext/curses/extconf.rb: now additionally checks for presence of these curses functions which are not present on NextStep or Openstep: bkgd(), bkgdset(), color(), curs(), getbkgd(), init(), scrl(), set(), setscrreg(), wattroff(), wattron(), wattrset(), wbkgd(), wbkgdset(), wscrl(), wsetscrreg() * ext/curses/curses.c: added appropriate #ifdef's for additional set of curses functions now checked by extconf.rb; fixed curses_bkgd() and window_bkgd() to correctly return boolean result rather than numeric result; fixed window_getbkgd() to correctly signal an error by returning nil rather than -1. * ext/etc/etc.c: setup_passwd() and setup_group() now check for null pointers before invoking rb_tainted_str_new2() upon fields extracted from `struct passwd' and `struct group' since null pointers in some fields are common on NextStep/OpenStep (especially so for the `pw_comment' field) and rb_tainted_str_new2() throws an exception when it receives a null pointer. * ext/pty/pty.c: include "util.h" for strdup()/ruby_strdup() for platforms such as NextStep and OpenStep which lack strdup(). * ext/socket/getaddrinfo.c: cast first argument of getservbyname(), gethostbyaddr(), and gethostbyname() from (const char*) to non-const (char*) for older platforms such as NextStep and OpenStep. * ext/socket/socket.c: include "util.h" for strdup()/ruby_strdup() for platforms such as NextStep and OpenStep which lack strdup(); include <netinet/in_systm.h> if present for NextStep and OpenStep; cast first argument of gethostbyaddr() and getservbyname() from (const char*) to non-const (char*) for older platforms. * ext/syslog/syslog.c: include "util.h" for strdup()/ruby_strdup() for platforms such as NextStep and OpenStep which lack strdup(). git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@5002 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2003-11-21 23:00:03 -05:00
w_extended(klass, arg, check);
w_byte(type, arg);
path = class2path(rb_class_real(klass));
w_unique(path, arg);
}
static void
w_uclass(VALUE obj, VALUE super, struct dump_arg *arg)
{
VALUE klass = CLASS_OF(obj);
w_extended(klass, arg, TRUE);
klass = rb_class_real(klass);
if (klass != super) {
w_byte(TYPE_UCLASS, arg);
w_unique(class2path(klass), arg);
}
}
#define to_be_skipped_id(id) (id == rb_id_encoding() || id == s_encoding_short || id == s_ruby2_keywords_flag || !rb_id2str(id))
struct w_ivar_arg {
struct dump_call_arg *dump;
st_data_t num_ivar;
};
static int
w_obj_each(st_data_t key, st_data_t val, st_data_t a)
{
ID id = (ID)key;
VALUE value = (VALUE)val;
struct w_ivar_arg *ivarg = (struct w_ivar_arg *)a;
struct dump_call_arg *arg = ivarg->dump;
if (to_be_skipped_id(id)) {
if (id == s_encoding_short) {
2020-01-11 02:32:56 -05:00
rb_warn("instance variable `"name_s_encoding_short"' on class %"PRIsVALUE" is not dumped",
CLASS_OF(arg->obj));
}
if (id == s_ruby2_keywords_flag) {
rb_warn("instance variable `"name_s_ruby2_keywords_flag"' on class %"PRIsVALUE" is not dumped",
CLASS_OF(arg->obj));
}
return ST_CONTINUE;
}
if (!ivarg->num_ivar) {
rb_raise(rb_eRuntimeError, "instance variable added to %"PRIsVALUE" instance",
CLASS_OF(arg->obj));
}
--ivarg->num_ivar;
w_symbol(ID2SYM(id), arg->arg);
w_object(value, arg->arg, arg->limit);
return ST_CONTINUE;
}
static int
obj_count_ivars(st_data_t key, st_data_t val, st_data_t a)
{
ID id = (ID)key;
if (!to_be_skipped_id(id)) ++*(st_index_t *)a;
return ST_CONTINUE;
}
static VALUE
encoding_name(VALUE obj, struct dump_arg *arg)
{
if (rb_enc_capable(obj)) {
int encidx = rb_enc_get_index(obj);
rb_encoding *enc = 0;
st_data_t name;
if (encidx <= 0 || !(enc = rb_enc_from_index(encidx))) {
return Qnil;
}
/* special treatment for US-ASCII and UTF-8 */
if (encidx == rb_usascii_encindex()) {
return Qfalse;
}
else if (encidx == rb_utf8_encindex()) {
return Qtrue;
}
if (arg->encodings ?
!st_lookup(arg->encodings, (st_data_t)rb_enc_name(enc), &name) :
(arg->encodings = st_init_strcasetable(), 1)) {
name = (st_data_t)rb_str_new_cstr(rb_enc_name(enc));
st_insert(arg->encodings, (st_data_t)rb_enc_name(enc), name);
}
return (VALUE)name;
}
else {
return Qnil;
}
}
static int
w_encoding(VALUE encname, struct dump_call_arg *arg)
{
int limit = arg->limit;
if (limit >= 0) ++limit;
switch (encname) {
case Qfalse:
case Qtrue:
2019-08-09 22:43:07 -04:00
w_symbol(ID2SYM(s_encoding_short), arg->arg);
w_object(encname, arg->arg, limit);
return 1;
case Qnil:
2019-07-01 03:25:56 -04:00
return 0;
}
w_symbol(ID2SYM(rb_id_encoding()), arg->arg);
w_object(encname, arg->arg, limit);
return 1;
}
static st_index_t
has_ivars(VALUE obj, VALUE encname, VALUE *ivobj)
{
st_index_t enc = !NIL_P(encname);
st_index_t num = 0;
st_index_t ruby2_keywords_flag = 0;
if (SPECIAL_CONST_P(obj)) goto generic;
switch (BUILTIN_TYPE(obj)) {
case T_OBJECT:
case T_CLASS:
case T_MODULE:
break; /* counted elsewhere */
case T_HASH:
ruby2_keywords_flag = RHASH(obj)->basic.flags & RHASH_PASS_AS_KEYWORDS ? 1 : 0;
/* fall through */
default:
generic:
rb_ivar_foreach(obj, obj_count_ivars, (st_data_t)&num);
if (ruby2_keywords_flag || num) *ivobj = obj;
}
return num + enc + ruby2_keywords_flag;
}
2019-07-01 03:18:51 -04:00
static void
w_ivar_each(VALUE obj, st_index_t num, struct dump_call_arg *arg)
{
struct w_ivar_arg ivarg = {arg, num};
if (!num) return;
rb_ivar_foreach(obj, w_obj_each, (st_data_t)&ivarg);
if (ivarg.num_ivar) {
rb_raise(rb_eRuntimeError, "instance variable removed from %"PRIsVALUE" instance",
CLASS_OF(arg->obj));
}
2019-07-01 03:18:51 -04:00
}
static void
w_ivar(st_index_t num, VALUE ivobj, VALUE encname, struct dump_call_arg *arg)
{
w_long(num, arg->arg);
num -= w_encoding(encname, arg);
if (RB_TYPE_P(ivobj, T_HASH) && (RHASH(ivobj)->basic.flags & RHASH_PASS_AS_KEYWORDS)) {
int limit = arg->limit;
if (limit >= 0) ++limit;
w_symbol(ID2SYM(s_ruby2_keywords_flag), arg->arg);
w_object(Qtrue, arg->arg, limit);
num--;
}
if (ivobj != Qundef && num) {
2019-07-01 03:18:51 -04:00
w_ivar_each(ivobj, num, arg);
}
}
static void
w_objivar(VALUE obj, struct dump_call_arg *arg)
{
st_data_t num = 0;
rb_ivar_foreach(obj, obj_count_ivars, (st_data_t)&num);
w_long(num, arg->arg);
2019-07-01 03:18:51 -04:00
w_ivar_each(obj, num, arg);
}
static void
w_object(VALUE obj, struct dump_arg *arg, int limit)
{
struct dump_call_arg c_arg;
VALUE ivobj = Qundef;
* gc.c (Init_stack): stack region is far smaller than usual if pthread is used. * marshal.c (w_extended): singleton methods should not be checked when dumping via marshal_dump() or _dump(). [ruby-talk:85909] * file.c (getcwdofdrv): avoid using getcwd() directly, use my_getcwd() instead. * merged NeXT, OpenStep, Rhapsody ports patch from Eric Sunshine <sunshine@sunshineco.com>. [ruby-core:01596] * marshal.c (w_object): LINK check earlier than anything else, i.e. do not dump TYPE_IVAR for already dumped objects. (ruby-bugs PR#1220) * eval.c (rb_eval): call "inherited" only when a new class is generated; not on reopening. * eval.c (eval): prepend error position in evaluating string to * configure.in: revived NextStep, OpenStep, and Rhapsody ports which had become unbuildable; enhanced --enable-fat-binary option so that it accepts a list of desired architectures (rather than assuming a fixed list), or defaults to a platform-appropriate list if user does not provide an explicit list; made the default list of architectures for MAB (fat binary) more comprehensive; now uses -fno-common even when building the interpreter (in addition to using it for extensions), thus allowing the interpreter to be embedded into a plugin module of an external project (in addition to allowing embedding directly into an application); added checks for <netinet/in_systm.h> (needed by `socket' extension) and getcwd(); now ensures that -I/usr/local/include is employed when extensions' extconf.rb scripts invoke have_header() since extension checks on NextStep and OpenStep will fail without it if the desired resource resides in the /usr/local tree; fixed formatting of --help message. * Makefile.in: $(LIBRUBY_A) rule now deletes the archive before invoking $(AR) since `ar' on Apple/NeXT can not "update" MAB archives (see configure's --enable-fat-binary option); added rule for new missing/getcwd.c. * defines.h: fixed endian handling during MAB build (see configure's --enable-fat-binary option) to ensure that all portions of the project see the correct WORDS_BIGENDIAN value (some extension modules were getting the wrong endian setting); added missing constants GETPGRP_VOID, WNOHANG, WUNTRACED, X_OK, and type pid_t for NextStep and OpenStep; removed unnecessary and problematic HAVE_SYS_WAIT_H define in NeXT section. * dir.c: do not allow NAMLEN() macro to trust dirent::d_namlen on NextStep since, on some installations, this value always resolves uselessly to zero. * dln.c: added error reporting to NextStep extension loader since the previous behavior of failing silently was not useful; now ensures that NSLINKMODULE_OPTION_BINDNOW compatibility constant is defined for OpenStep and Rhapsody; no longer includes <mach-o/dyld.h> twice on Rhapsody since this header lacks multiple-include protection, which resulted in "redefinition" compilation errors. * main.c: also create hard reference to objc_msgSend() on NeXT platforms (in addition to Apple platforms). * lib/mkmf.rb: now exports XCFLAGS from configure script to extension makefiles so that extensions can be built MAB (see configure's --enable-fat-binary option); also utilize XCFLAGS in cc_command() (but not cpp_command() because MAB flags are incompatible with direct invocation of `cpp'). * ext/curses/extconf.rb: now additionally checks for presence of these curses functions which are not present on NextStep or Openstep: bkgd(), bkgdset(), color(), curs(), getbkgd(), init(), scrl(), set(), setscrreg(), wattroff(), wattron(), wattrset(), wbkgd(), wbkgdset(), wscrl(), wsetscrreg() * ext/curses/curses.c: added appropriate #ifdef's for additional set of curses functions now checked by extconf.rb; fixed curses_bkgd() and window_bkgd() to correctly return boolean result rather than numeric result; fixed window_getbkgd() to correctly signal an error by returning nil rather than -1. * ext/etc/etc.c: setup_passwd() and setup_group() now check for null pointers before invoking rb_tainted_str_new2() upon fields extracted from `struct passwd' and `struct group' since null pointers in some fields are common on NextStep/OpenStep (especially so for the `pw_comment' field) and rb_tainted_str_new2() throws an exception when it receives a null pointer. * ext/pty/pty.c: include "util.h" for strdup()/ruby_strdup() for platforms such as NextStep and OpenStep which lack strdup(). * ext/socket/getaddrinfo.c: cast first argument of getservbyname(), gethostbyaddr(), and gethostbyname() from (const char*) to non-const (char*) for older platforms such as NextStep and OpenStep. * ext/socket/socket.c: include "util.h" for strdup()/ruby_strdup() for platforms such as NextStep and OpenStep which lack strdup(); include <netinet/in_systm.h> if present for NextStep and OpenStep; cast first argument of gethostbyaddr() and getservbyname() from (const char*) to non-const (char*) for older platforms. * ext/syslog/syslog.c: include "util.h" for strdup()/ruby_strdup() for platforms such as NextStep and OpenStep which lack strdup(). git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@5002 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2003-11-21 23:00:03 -05:00
st_data_t num;
st_index_t hasiv = 0;
VALUE encname = Qnil;
if (limit == 0) {
rb_raise(rb_eArgError, "exceed depth limit");
}
if (limit > 0) limit--;
c_arg.limit = limit;
c_arg.arg = arg;
c_arg.obj = obj;
* gc.c (Init_stack): stack region is far smaller than usual if pthread is used. * marshal.c (w_extended): singleton methods should not be checked when dumping via marshal_dump() or _dump(). [ruby-talk:85909] * file.c (getcwdofdrv): avoid using getcwd() directly, use my_getcwd() instead. * merged NeXT, OpenStep, Rhapsody ports patch from Eric Sunshine <sunshine@sunshineco.com>. [ruby-core:01596] * marshal.c (w_object): LINK check earlier than anything else, i.e. do not dump TYPE_IVAR for already dumped objects. (ruby-bugs PR#1220) * eval.c (rb_eval): call "inherited" only when a new class is generated; not on reopening. * eval.c (eval): prepend error position in evaluating string to * configure.in: revived NextStep, OpenStep, and Rhapsody ports which had become unbuildable; enhanced --enable-fat-binary option so that it accepts a list of desired architectures (rather than assuming a fixed list), or defaults to a platform-appropriate list if user does not provide an explicit list; made the default list of architectures for MAB (fat binary) more comprehensive; now uses -fno-common even when building the interpreter (in addition to using it for extensions), thus allowing the interpreter to be embedded into a plugin module of an external project (in addition to allowing embedding directly into an application); added checks for <netinet/in_systm.h> (needed by `socket' extension) and getcwd(); now ensures that -I/usr/local/include is employed when extensions' extconf.rb scripts invoke have_header() since extension checks on NextStep and OpenStep will fail without it if the desired resource resides in the /usr/local tree; fixed formatting of --help message. * Makefile.in: $(LIBRUBY_A) rule now deletes the archive before invoking $(AR) since `ar' on Apple/NeXT can not "update" MAB archives (see configure's --enable-fat-binary option); added rule for new missing/getcwd.c. * defines.h: fixed endian handling during MAB build (see configure's --enable-fat-binary option) to ensure that all portions of the project see the correct WORDS_BIGENDIAN value (some extension modules were getting the wrong endian setting); added missing constants GETPGRP_VOID, WNOHANG, WUNTRACED, X_OK, and type pid_t for NextStep and OpenStep; removed unnecessary and problematic HAVE_SYS_WAIT_H define in NeXT section. * dir.c: do not allow NAMLEN() macro to trust dirent::d_namlen on NextStep since, on some installations, this value always resolves uselessly to zero. * dln.c: added error reporting to NextStep extension loader since the previous behavior of failing silently was not useful; now ensures that NSLINKMODULE_OPTION_BINDNOW compatibility constant is defined for OpenStep and Rhapsody; no longer includes <mach-o/dyld.h> twice on Rhapsody since this header lacks multiple-include protection, which resulted in "redefinition" compilation errors. * main.c: also create hard reference to objc_msgSend() on NeXT platforms (in addition to Apple platforms). * lib/mkmf.rb: now exports XCFLAGS from configure script to extension makefiles so that extensions can be built MAB (see configure's --enable-fat-binary option); also utilize XCFLAGS in cc_command() (but not cpp_command() because MAB flags are incompatible with direct invocation of `cpp'). * ext/curses/extconf.rb: now additionally checks for presence of these curses functions which are not present on NextStep or Openstep: bkgd(), bkgdset(), color(), curs(), getbkgd(), init(), scrl(), set(), setscrreg(), wattroff(), wattron(), wattrset(), wbkgd(), wbkgdset(), wscrl(), wsetscrreg() * ext/curses/curses.c: added appropriate #ifdef's for additional set of curses functions now checked by extconf.rb; fixed curses_bkgd() and window_bkgd() to correctly return boolean result rather than numeric result; fixed window_getbkgd() to correctly signal an error by returning nil rather than -1. * ext/etc/etc.c: setup_passwd() and setup_group() now check for null pointers before invoking rb_tainted_str_new2() upon fields extracted from `struct passwd' and `struct group' since null pointers in some fields are common on NextStep/OpenStep (especially so for the `pw_comment' field) and rb_tainted_str_new2() throws an exception when it receives a null pointer. * ext/pty/pty.c: include "util.h" for strdup()/ruby_strdup() for platforms such as NextStep and OpenStep which lack strdup(). * ext/socket/getaddrinfo.c: cast first argument of getservbyname(), gethostbyaddr(), and gethostbyname() from (const char*) to non-const (char*) for older platforms such as NextStep and OpenStep. * ext/socket/socket.c: include "util.h" for strdup()/ruby_strdup() for platforms such as NextStep and OpenStep which lack strdup(); include <netinet/in_systm.h> if present for NextStep and OpenStep; cast first argument of gethostbyaddr() and getservbyname() from (const char*) to non-const (char*) for older platforms. * ext/syslog/syslog.c: include "util.h" for strdup()/ruby_strdup() for platforms such as NextStep and OpenStep which lack strdup(). git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@5002 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2003-11-21 23:00:03 -05:00
if (st_lookup(arg->data, obj, &num)) {
w_byte(TYPE_LINK, arg);
w_long((long)num, arg);
return;
}
if (obj == Qnil) {
w_byte(TYPE_NIL, arg);
}
else if (obj == Qtrue) {
w_byte(TYPE_TRUE, arg);
}
else if (obj == Qfalse) {
w_byte(TYPE_FALSE, arg);
}
else if (FIXNUM_P(obj)) {
#if SIZEOF_LONG <= 4
w_byte(TYPE_FIXNUM, arg);
w_long(FIX2INT(obj), arg);
#else
if (RSHIFT((long)obj, 31) == 0 || RSHIFT((long)obj, 31) == -1) {
w_byte(TYPE_FIXNUM, arg);
w_long(FIX2LONG(obj), arg);
}
else {
w_object(rb_int2big(FIX2LONG(obj)), arg, limit);
}
#endif
}
else if (SYMBOL_P(obj)) {
w_symbol(obj, arg);
}
else if (FLONUM_P(obj)) {
st_add_direct(arg->data, obj, arg->data->num_entries);
w_byte(TYPE_FLOAT, arg);
w_float(RFLOAT_VALUE(obj), arg);
}
else {
VALUE v;
if (!RBASIC_CLASS(obj)) {
rb_raise(rb_eTypeError, "can't dump internal %s",
rb_builtin_type_name(BUILTIN_TYPE(obj)));
}
if (rb_obj_respond_to(obj, s_mdump, TRUE)) {
st_add_direct(arg->data, obj, arg->data->num_entries);
v = dump_funcall(arg, obj, s_mdump, 0, 0);
w_class(TYPE_USRMARSHAL, obj, arg, FALSE);
w_object(v, arg, limit);
return;
}
if (rb_obj_respond_to(obj, s_dump, TRUE)) {
VALUE ivobj2 = Qundef;
st_index_t hasiv2;
VALUE encname2;
v = INT2NUM(limit);
v = dump_funcall(arg, obj, s_dump, 1, &v);
if (!RB_TYPE_P(v, T_STRING)) {
rb_raise(rb_eTypeError, "_dump() must return string");
}
hasiv = has_ivars(obj, (encname = encoding_name(obj, arg)), &ivobj);
hasiv2 = has_ivars(v, (encname2 = encoding_name(v, arg)), &ivobj2);
if (hasiv2) {
hasiv = hasiv2;
ivobj = ivobj2;
encname = encname2;
}
if (hasiv) w_byte(TYPE_IVAR, arg);
w_class(TYPE_USERDEF, obj, arg, FALSE);
w_bytes(RSTRING_PTR(v), RSTRING_LEN(v), arg);
if (hasiv) {
w_ivar(hasiv, ivobj, encname, &c_arg);
}
st_add_direct(arg->data, obj, arg->data->num_entries);
return;
}
st_add_direct(arg->data, obj, arg->data->num_entries);
hasiv = has_ivars(obj, (encname = encoding_name(obj, arg)), &ivobj);
{
st_data_t compat_data;
rb_alloc_func_t allocator = rb_get_alloc_func(RBASIC(obj)->klass);
if (st_lookup(compat_allocator_tbl,
(st_data_t)allocator,
&compat_data)) {
marshal_compat_t *compat = (marshal_compat_t*)compat_data;
VALUE real_obj = obj;
obj = compat->dumper(real_obj);
if (!arg->compat_tbl) {
arg->compat_tbl = rb_init_identtable();
}
st_insert(arg->compat_tbl, (st_data_t)obj, (st_data_t)real_obj);
if (obj != real_obj && ivobj == Qundef) hasiv = 0;
}
}
if (hasiv) w_byte(TYPE_IVAR, arg);
switch (BUILTIN_TYPE(obj)) {
case T_CLASS:
if (FL_TEST(obj, FL_SINGLETON)) {
rb_raise(rb_eTypeError, "singleton class can't be dumped");
}
w_byte(TYPE_CLASS, arg);
{
VALUE path = class2path(obj);
w_bytes(RSTRING_PTR(path), RSTRING_LEN(path), arg);
RB_GC_GUARD(path);
}
break;
case T_MODULE:
w_byte(TYPE_MODULE, arg);
{
VALUE path = class2path(obj);
w_bytes(RSTRING_PTR(path), RSTRING_LEN(path), arg);
RB_GC_GUARD(path);
}
break;
case T_FLOAT:
w_byte(TYPE_FLOAT, arg);
w_float(RFLOAT_VALUE(obj), arg);
break;
case T_BIGNUM:
w_byte(TYPE_BIGNUM, arg);
{
char sign = BIGNUM_SIGN(obj) ? '+' : '-';
size_t len = BIGNUM_LEN(obj);
size_t slen;
size_t j;
BDIGIT *d = BIGNUM_DIGITS(obj);
slen = SHORTLEN(len);
if (LONG_MAX < slen) {
rb_raise(rb_eTypeError, "too big Bignum can't be dumped");
}
w_byte(sign, arg);
w_long((long)slen, arg);
for (j = 0; j < len; j++) {
#if SIZEOF_BDIGIT > SIZEOF_SHORT
BDIGIT num = *d;
int i;
for (i=0; i<SIZEOF_BDIGIT; i+=SIZEOF_SHORT) {
w_short(num & SHORTMASK, arg);
num = SHORTDN(num);
if (j == len - 1 && num == 0) break;
}
#else
w_short(*d, arg);
#endif
d++;
}
}
break;
case T_STRING:
w_uclass(obj, rb_cString, arg);
w_byte(TYPE_STRING, arg);
w_bytes(RSTRING_PTR(obj), RSTRING_LEN(obj), arg);
break;
case T_REGEXP:
w_uclass(obj, rb_cRegexp, arg);
w_byte(TYPE_REGEXP, arg);
{
int opts = rb_reg_options(obj);
w_bytes(RREGEXP_SRC_PTR(obj), RREGEXP_SRC_LEN(obj), arg);
w_byte((char)opts, arg);
}
break;
case T_ARRAY:
w_uclass(obj, rb_cArray, arg);
w_byte(TYPE_ARRAY, arg);
{
long i, len = RARRAY_LEN(obj);
w_long(len, arg);
for (i=0; i<RARRAY_LEN(obj); i++) {
w_object(RARRAY_AREF(obj, i), arg, limit);
if (len != RARRAY_LEN(obj)) {
rb_raise(rb_eRuntimeError, "array modified during dump");
}
}
}
break;
case T_HASH:
w_uclass(obj, rb_cHash, arg);
if (NIL_P(RHASH_IFNONE(obj))) {
w_byte(TYPE_HASH, arg);
}
2019-07-30 21:22:47 -04:00
else if (FL_TEST(obj, RHASH_PROC_DEFAULT)) {
* array.c: replace rb_protect_inspect() and rb_inspecting_p() by rb_exec_recursive() in eval.c. * eval.c (rb_exec_recursive): new function. * array.c (rb_ary_join): use rb_exec_recursive(). * array.c (rb_ary_inspect, rb_ary_hash): ditto. * file.c (rb_file_join): ditto. * hash.c (rb_hash_inspect, rb_hash_to_s, rb_hash_hash): ditto. * io.c (rb_io_puts): ditto. * object.c (rb_obj_inspect): ditto * struct.c (rb_struct_inspect): ditto. * lib/set.rb (SortedSet::setup): a hack to shut up warning. [ruby-talk:132866] * lib/time.rb (Time::strptime): add new function. inspired by [ruby-talk:132815]. * lib/parsedate.rb (ParseDate::strptime): ditto. * regparse.c: move st_*_strend() functions from st.c. fixed some potential memory leaks. * exception error messages updated. [ruby-core:04497] * ext/socket/socket.c (Init_socket): add bunch of Socket constants. Patch from Sam Roberts <sroberts@uniserve.com>. [ruby-core:04409] * array.c (rb_ary_s_create): no need for negative argc check. [ruby-core:04463] * array.c (rb_ary_unshift_m): ditto. * lib/xmlrpc/parser.rb (XMLRPC::FaultException): make it subclass of StandardError class, not Exception class. [ruby-core:04429] * parse.y (fcall_gen): lvar(arg) will be evaluated as lvar.call(arg) when lvar is a defined local variable. [new] * object.c (rb_class_initialize): call inherited method before calling initializing block. * eval.c (rb_thread_start_1): initialize newly pushed frame. * lib/open3.rb (Open3::popen3): $? should not be EXIT_FAILURE. fixed: [ruby-core:04444] * eval.c (is_defined): NODE_IASGN is an assignment. * ext/readline/readline.c (Readline.readline): use rl_outstream and rl_instream. [ruby-dev:25699] * ext/etc/etc.c (Init_etc): sGroup needs HAVE_ST_GR_PASSWD check [ruby-dev:25675] * misc/ruby-mode.el: [ruby-core:04415] * lib/rdoc/generators/html_generator.rb: [ruby-core:04412] * lib/rdoc/generators/ri_generator.rb: ditto. * struct.c (make_struct): fixed: [ruby-core:04402] * ext/curses/curses.c (window_color_set): [ruby-core:04393] * ext/socket/socket.c (Init_socket): SO_REUSEPORT added. [ruby-talk:130092] * object.c: [ruby-doc:818] * parse.y (open_args): fix too verbose warnings for the space before argument parentheses. [ruby-dev:25492] * parse.y (parser_yylex): ditto. * parse.y (parser_yylex): the first expression in the parentheses should not be a command. [ruby-dev:25492] * lib/irb/context.rb (IRB::Context::initialize): [ruby-core:04330] * object.c (Init_Object): remove Object#type. [ruby-core:04335] * st.c (st_foreach): report success/failure by return value. [ruby-Bugs-1396] * parse.y: forgot to initialize parser struct. [ruby-dev:25492] * parse.y (parser_yylex): no tLABEL on EXPR_BEG. [ruby-talk:127711] * document updates - [ruby-core:04296], [ruby-core:04301], [ruby-core:04302], [ruby-core:04307] * dir.c (rb_push_glob): should work for NUL delimited patterns. * dir.c (rb_glob2): should aware of offset in the pattern. * string.c (rb_str_new4): should propagate taintedness. * env.h: rename member names in struct FRAME; last_func -> callee, orig_func -> this_func, last_class -> this_class. * struct.c (rb_struct_set): use original method name, not callee name, to retrieve member slot. [ruby-core:04268] * time.c (time_strftime): protect from format modification from GC finalizers. * object.c (Init_Object): remove rb_obj_id_obsolete() * eval.c (rb_mod_define_method): incomplete subclass check. [ruby-dev:25464] * gc.c (rb_data_object_alloc): klass may be NULL. [ruby-list:40498] * bignum.c (rb_big_rand): should return positive random number. [ruby-dev:25401] * bignum.c (rb_big_rand): do not use rb_big_modulo to generate random bignums. [ruby-dev:25396] * variable.c (rb_autoload): [ruby-dev:25373] * eval.c (svalue_to_avalue): [ruby-dev:25366] * string.c (rb_str_justify): [ruby-dev:25367] * io.c (rb_f_select): [ruby-dev:25312] * ext/socket/socket.c (sock_s_getservbyport): [ruby-talk:124072] * struct.c (make_struct): [ruby-dev:25249] * dir.c (dir_open_dir): new function. [ruby-dev:25242] * io.c (rb_f_open): add type check for return value from to_open. * lib/pstore.rb (PStore#transaction): Use the empty content when a file is not found. [ruby-dev:24561] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@8068 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2005-03-04 01:47:45 -05:00
rb_raise(rb_eTypeError, "can't dump hash with default proc");
}
else {
w_byte(TYPE_HASH_DEF, arg);
}
w_long(rb_hash_size_num(obj), arg);
rb_hash_foreach(obj, hash_each, (st_data_t)&c_arg);
if (!NIL_P(RHASH_IFNONE(obj))) {
w_object(RHASH_IFNONE(obj), arg, limit);
}
break;
case T_STRUCT:
w_class(TYPE_STRUCT, obj, arg, TRUE);
{
long len = RSTRUCT_LEN(obj);
VALUE mem;
long i;
w_long(len, arg);
mem = rb_struct_members(obj);
for (i=0; i<len; i++) {
w_symbol(RARRAY_AREF(mem, i), arg);
w_object(RSTRUCT_GET(obj, i), arg, limit);
}
}
break;
case T_OBJECT:
w_class(TYPE_OBJECT, obj, arg, TRUE);
w_objivar(obj, &c_arg);
break;
case T_DATA:
{
VALUE v;
if (!rb_obj_respond_to(obj, s_dump_data, TRUE)) {
rb_raise(rb_eTypeError,
"no _dump_data is defined for class %"PRIsVALUE,
rb_obj_class(obj));
}
v = dump_funcall(arg, obj, s_dump_data, 0, 0);
w_class(TYPE_DATA, obj, arg, TRUE);
w_object(v, arg, limit);
}
break;
default:
rb_raise(rb_eTypeError, "can't dump %"PRIsVALUE,
rb_obj_class(obj));
break;
}
RB_GC_GUARD(obj);
}
if (hasiv) {
w_ivar(hasiv, ivobj, encname, &c_arg);
}
}
static void
clear_dump_arg(struct dump_arg *arg)
{
if (!arg->symbols) return;
st_free_table(arg->symbols);
arg->symbols = 0;
st_free_table(arg->data);
arg->data = 0;
if (arg->compat_tbl) {
st_free_table(arg->compat_tbl);
arg->compat_tbl = 0;
}
if (arg->encodings) {
st_free_table(arg->encodings);
arg->encodings = 0;
}
}
NORETURN(static inline void io_needed(void));
static inline void
io_needed(void)
{
rb_raise(rb_eTypeError, "instance of IO needed");
}
/*
* call-seq:
* dump( obj [, anIO] , limit=-1 ) -> anIO
*
* Serializes obj and all descendant objects. If anIO is
* specified, the serialized data will be written to it, otherwise the
* data will be returned as a String. If limit is specified, the
* traversal of subobjects will be limited to that depth. If limit is
* negative, no checking of depth will be performed.
*
* class Klass
* def initialize(str)
* @str = str
* end
* def say_hello
* @str
* end
* end
*
* (produces no output)
*
* o = Klass.new("hello\n")
* data = Marshal.dump(o)
* obj = Marshal.load(data)
* obj.say_hello #=> "hello\n"
*
* Marshal can't dump following objects:
* * anonymous Class/Module.
* * objects which are related to system (ex: Dir, File::Stat, IO, File, Socket
* and so on)
* * an instance of MatchData, Data, Method, UnboundMethod, Proc, Thread,
* ThreadGroup, Continuation
* * objects which define singleton methods
*/
static VALUE
marshal_dump(int argc, VALUE *argv, VALUE _)
{
VALUE obj, port, a1, a2;
int limit = -1;
port = Qnil;
rb_scan_args(argc, argv, "12", &obj, &a1, &a2);
if (argc == 3) {
if (!NIL_P(a2)) limit = NUM2INT(a2);
if (NIL_P(a1)) io_needed();
port = a1;
}
else if (argc == 2) {
if (FIXNUM_P(a1)) limit = FIX2INT(a1);
else if (NIL_P(a1)) io_needed();
else port = a1;
}
return rb_marshal_dump_limited(obj, port, limit);
}
VALUE
rb_marshal_dump_limited(VALUE obj, VALUE port, int limit)
{
struct dump_arg *arg;
VALUE wrapper; /* used to avoid memory leak in case of exception */
wrapper = TypedData_Make_Struct(0, struct dump_arg, &dump_arg_data, arg);
arg->dest = 0;
arg->symbols = st_init_numtable();
arg->data = rb_init_identtable();
arg->compat_tbl = 0;
arg->encodings = 0;
arg->str = rb_str_buf_new(0);
if (!NIL_P(port)) {
if (!rb_respond_to(port, s_write)) {
io_needed();
}
arg->dest = port;
dump_check_funcall(arg, port, s_binmode, 0, 0);
}
else {
port = arg->str;
}
w_byte(MARSHAL_MAJOR, arg);
w_byte(MARSHAL_MINOR, arg);
w_object(obj, arg, limit);
if (arg->dest) {
rb_io_write(arg->dest, arg->str);
rb_str_resize(arg->str, 0);
}
clear_dump_arg(arg);
RB_GC_GUARD(wrapper);
return port;
}
struct load_arg {
VALUE src;
char *buf;
long buflen;
long readable;
long offset;
st_table *symbols;
st_table *data;
VALUE proc;
st_table *compat_tbl;
};
static VALUE
check_load_arg(VALUE ret, struct load_arg *arg, const char *name)
{
if (!arg->symbols) {
rb_raise(rb_eRuntimeError, "Marshal.load reentered at %s",
name);
}
return ret;
}
#define load_funcall(arg, obj, sym, argc, argv) \
check_load_arg(rb_funcallv(obj, sym, argc, argv), arg, name_##sym)
static void clear_load_arg(struct load_arg *arg);
static void
mark_load_arg(void *ptr)
{
struct load_arg *p = ptr;
if (!p->symbols)
return;
rb_mark_tbl(p->symbols);
rb_mark_tbl(p->data);
rb_mark_hash(p->compat_tbl);
}
static void
free_load_arg(void *ptr)
{
clear_load_arg(ptr);
xfree(ptr);
}
static size_t
memsize_load_arg(const void *ptr)
{
return sizeof(struct load_arg);
}
static const rb_data_type_t load_arg_data = {
"load_arg",
{mark_load_arg, free_load_arg, memsize_load_arg,},
0, 0, RUBY_TYPED_FREE_IMMEDIATELY
};
#define r_entry(v, arg) r_entry0((v), (arg)->data->num_entries, (arg))
static VALUE r_entry0(VALUE v, st_index_t num, struct load_arg *arg);
static VALUE r_object(struct load_arg *arg);
static VALUE r_symbol(struct load_arg *arg);
static VALUE path2class(VALUE path);
NORETURN(static void too_short(void));
static void
too_short(void)
{
rb_raise(rb_eArgError, "marshal data too short");
}
static st_index_t
r_prepare(struct load_arg *arg)
{
st_index_t idx = arg->data->num_entries;
st_insert(arg->data, (st_data_t)idx, (st_data_t)Qundef);
return idx;
}
static unsigned char
r_byte1_buffered(struct load_arg *arg)
{
if (arg->buflen == 0) {
long readable = arg->readable < BUFSIZ ? arg->readable : BUFSIZ;
VALUE str, n = LONG2NUM(readable);
str = load_funcall(arg, arg->src, s_read, 1, &n);
if (NIL_P(str)) too_short();
StringValue(str);
memcpy(arg->buf, RSTRING_PTR(str), RSTRING_LEN(str));
arg->offset = 0;
arg->buflen = RSTRING_LEN(str);
}
arg->buflen--;
return arg->buf[arg->offset++];
}
static int
r_byte(struct load_arg *arg)
{
int c;
if (RB_TYPE_P(arg->src, T_STRING)) {
if (RSTRING_LEN(arg->src) > arg->offset) {
c = (unsigned char)RSTRING_PTR(arg->src)[arg->offset++];
}
else {
too_short();
}
}
else {
if (arg->readable >0 || arg->buflen > 0) {
c = r_byte1_buffered(arg);
}
else {
VALUE v = load_funcall(arg, arg->src, s_getbyte, 0, 0);
if (NIL_P(v)) rb_eof_error();
c = (unsigned char)NUM2CHR(v);
}
}
return c;
}
NORETURN(static void long_toobig(int size));
static void
long_toobig(int size)
{
rb_raise(rb_eTypeError, "long too big for this architecture (size "
STRINGIZE(SIZEOF_LONG)", given %d)", size);
}
static long
r_long(struct load_arg *arg)
{
register long x;
int c = (signed char)r_byte(arg);
long i;
if (c == 0) return 0;
if (c > 0) {
if (4 < c && c < 128) {
return c - 5;
}
if (c > (int)sizeof(long)) long_toobig(c);
x = 0;
for (i=0;i<c;i++) {
x |= (long)r_byte(arg) << (8*i);
}
}
else {
if (-129 < c && c < -4) {
return c + 5;
}
c = -c;
if (c > (int)sizeof(long)) long_toobig(c);
x = -1;
for (i=0;i<c;i++) {
x &= ~((long)0xff << (8*i));
x |= (long)r_byte(arg) << (8*i);
}
}
return x;
}
long
ruby_marshal_read_long(const char **buf, long len)
{
long x;
struct RString src;
struct load_arg arg;
memset(&arg, 0, sizeof(arg));
arg.src = rb_setup_fake_str(&src, *buf, len, 0);
x = r_long(&arg);
*buf += arg.offset;
return x;
}
static VALUE
r_bytes1(long len, struct load_arg *arg)
{
VALUE str, n = LONG2NUM(len);
str = load_funcall(arg, arg->src, s_read, 1, &n);
if (NIL_P(str)) too_short();
StringValue(str);
if (RSTRING_LEN(str) != len) too_short();
return str;
}
static VALUE
r_bytes1_buffered(long len, struct load_arg *arg)
{
VALUE str;
if (len <= arg->buflen) {
str = rb_str_new(arg->buf+arg->offset, len);
arg->offset += len;
arg->buflen -= len;
}
else {
long buflen = arg->buflen;
long readable = arg->readable + 1;
long tmp_len, read_len, need_len = len - buflen;
VALUE tmp, n;
readable = readable < BUFSIZ ? readable : BUFSIZ;
read_len = need_len > readable ? need_len : readable;
n = LONG2NUM(read_len);
tmp = load_funcall(arg, arg->src, s_read, 1, &n);
if (NIL_P(tmp)) too_short();
StringValue(tmp);
tmp_len = RSTRING_LEN(tmp);
if (tmp_len < need_len) too_short();
str = rb_str_new(arg->buf+arg->offset, buflen);
rb_str_cat(str, RSTRING_PTR(tmp), need_len);
if (tmp_len > need_len) {
buflen = tmp_len - need_len;
memcpy(arg->buf, RSTRING_PTR(tmp)+need_len, buflen);
arg->buflen = buflen;
}
else {
arg->buflen = 0;
}
arg->offset = 0;
}
return str;
}
#define r_bytes(arg) r_bytes0(r_long(arg), (arg))
static VALUE
r_bytes0(long len, struct load_arg *arg)
{
VALUE str;
if (len == 0) return rb_str_new(0, 0);
if (RB_TYPE_P(arg->src, T_STRING)) {
if (RSTRING_LEN(arg->src) - arg->offset >= len) {
str = rb_str_new(RSTRING_PTR(arg->src)+arg->offset, len);
arg->offset += len;
}
else {
too_short();
}
}
else {
if (arg->readable > 0 || arg->buflen > 0) {
str = r_bytes1_buffered(len, arg);
}
else {
str = r_bytes1(len, arg);
}
}
return str;
}
2020-01-11 02:32:56 -05:00
static inline int
name_equal(const char *name, size_t nlen, const char *p, long l)
{
if ((size_t)l != nlen || *p != *name) return 0;
return nlen == 1 || memcmp(p+1, name+1, nlen-1) == 0;
}
static int
sym2encidx(VALUE sym, VALUE val)
{
static const char name_encoding[8] = "encoding";
const char *p;
long l;
if (rb_enc_get_index(sym) != ENCINDEX_US_ASCII) return -1;
RSTRING_GETMEM(sym, p, l);
if (l <= 0) return -1;
2020-01-11 02:32:56 -05:00
if (name_equal(name_encoding, sizeof(name_encoding), p, l)) {
int idx = rb_enc_find_index(StringValueCStr(val));
return idx;
}
2020-01-11 02:32:56 -05:00
if (name_equal(name_s_encoding_short, rb_strlen_lit(name_s_encoding_short), p, l)) {
if (val == Qfalse) return rb_usascii_encindex();
else if (val == Qtrue) return rb_utf8_encindex();
/* bogus ignore */
}
return -1;
}
static int
ruby2_keywords_flag_check(VALUE sym)
{
const char *p;
long l;
RSTRING_GETMEM(sym, p, l);
if (l <= 0) return 0;
if (name_equal(name_s_ruby2_keywords_flag, rb_strlen_lit(name_s_ruby2_keywords_flag), p, 1)) {
return 1;
}
return 0;
}
static VALUE
r_symlink(struct load_arg *arg)
{
st_data_t sym;
long num = r_long(arg);
if (!st_lookup(arg->symbols, num, &sym)) {
rb_raise(rb_eArgError, "bad symbol");
}
return (VALUE)sym;
}
static VALUE
r_symreal(struct load_arg *arg, int ivar)
{
VALUE s = r_bytes(arg);
VALUE sym;
int idx = -1;
st_index_t n = arg->symbols->num_entries;
if (rb_enc_str_asciionly_p(s)) rb_enc_associate_index(s, ENCINDEX_US_ASCII);
st_insert(arg->symbols, (st_data_t)n, (st_data_t)s);
if (ivar) {
long num = r_long(arg);
while (num-- > 0) {
sym = r_symbol(arg);
idx = sym2encidx(sym, r_object(arg));
}
}
if (idx > 0) rb_enc_associate_index(s, idx);
return s;
}
static VALUE
r_symbol(struct load_arg *arg)
{
int type, ivar = 0;
again:
switch ((type = r_byte(arg))) {
default:
rb_raise(rb_eArgError, "dump format error for symbol(0x%x)", type);
case TYPE_IVAR:
ivar = 1;
goto again;
case TYPE_SYMBOL:
return r_symreal(arg, ivar);
case TYPE_SYMLINK:
if (ivar) {
rb_raise(rb_eArgError, "dump format error (symlink with encoding)");
}
return r_symlink(arg);
}
}
static VALUE
r_unique(struct load_arg *arg)
{
return r_symbol(arg);
}
static VALUE
r_string(struct load_arg *arg)
{
return r_bytes(arg);
}
static VALUE
r_entry0(VALUE v, st_index_t num, struct load_arg *arg)
{
2020-09-05 04:28:06 -04:00
st_data_t real_obj = (st_data_t)v;
if (arg->compat_tbl) {
/* real_obj is kept if not found */
st_lookup(arg->compat_tbl, v, &real_obj);
}
2020-09-05 04:28:06 -04:00
st_insert(arg->data, num, real_obj);
return v;
}
static VALUE
r_fixup_compat(VALUE v, struct load_arg *arg)
{
st_data_t data;
st_data_t key = (st_data_t)v;
if (arg->compat_tbl && st_delete(arg->compat_tbl, &key, &data)) {
VALUE real_obj = (VALUE)data;
rb_alloc_func_t allocator = rb_get_alloc_func(CLASS_OF(real_obj));
if (st_lookup(compat_allocator_tbl, (st_data_t)allocator, &data)) {
marshal_compat_t *compat = (marshal_compat_t*)data;
compat->loader(real_obj, v);
}
v = real_obj;
}
return v;
}
static VALUE
r_post_proc(VALUE v, struct load_arg *arg)
{
if (arg->proc) {
v = load_funcall(arg, arg->proc, s_call, 1, &v);
}
return v;
}
static VALUE
r_leave(VALUE v, struct load_arg *arg)
{
v = r_fixup_compat(v, arg);
v = r_post_proc(v, arg);
return v;
}
static int
copy_ivar_i(st_data_t key, st_data_t val, st_data_t arg)
{
VALUE obj = (VALUE)arg, value = (VALUE)val;
ID vid = (ID)key;
if (!rb_ivar_defined(obj, vid))
rb_ivar_set(obj, vid, value);
return ST_CONTINUE;
}
static VALUE
r_copy_ivar(VALUE v, VALUE data)
{
rb_ivar_foreach(data, copy_ivar_i, (st_data_t)v);
return v;
}
static void
r_ivar(VALUE obj, int *has_encoding, struct load_arg *arg)
{
long len;
len = r_long(arg);
if (len > 0) {
do {
VALUE sym = r_symbol(arg);
VALUE val = r_object(arg);
int idx = sym2encidx(sym, val);
if (idx >= 0) {
if (rb_enc_capable(obj)) {
rb_enc_associate_index(obj, idx);
}
else {
rb_raise(rb_eArgError, "%"PRIsVALUE" is not enc_capable", obj);
}
if (has_encoding) *has_encoding = TRUE;
}
else if (ruby2_keywords_flag_check(sym)) {
if (RB_TYPE_P(obj, T_HASH)) {
RHASH(obj)->basic.flags |= RHASH_PASS_AS_KEYWORDS;
}
else {
rb_raise(rb_eArgError, "ruby2_keywords flag is given but %"PRIsVALUE" is not a Hash", obj);
}
}
else {
rb_ivar_set(obj, rb_intern_str(sym), val);
}
} while (--len > 0);
}
}
static VALUE
path2class(VALUE path)
{
VALUE v = rb_path_to_class(path);
if (!RB_TYPE_P(v, T_CLASS)) {
rb_raise(rb_eArgError, "%"PRIsVALUE" does not refer to class", path);
}
return v;
}
#define path2module(path) must_be_module(rb_path_to_class(path), path)
static VALUE
must_be_module(VALUE v, VALUE path)
{
if (!RB_TYPE_P(v, T_MODULE)) {
rb_raise(rb_eArgError, "%"PRIsVALUE" does not refer to module", path);
}
return v;
}
static VALUE
obj_alloc_by_klass(VALUE klass, struct load_arg *arg, VALUE *oldclass)
{
st_data_t data;
rb_alloc_func_t allocator;
allocator = rb_get_alloc_func(klass);
if (st_lookup(compat_allocator_tbl, (st_data_t)allocator, &data)) {
marshal_compat_t *compat = (marshal_compat_t*)data;
VALUE real_obj = rb_obj_alloc(klass);
VALUE obj = rb_obj_alloc(compat->oldclass);
if (oldclass) *oldclass = compat->oldclass;
if (!arg->compat_tbl) {
arg->compat_tbl = rb_init_identtable();
}
st_insert(arg->compat_tbl, (st_data_t)obj, (st_data_t)real_obj);
return obj;
}
return rb_obj_alloc(klass);
}
static VALUE
obj_alloc_by_path(VALUE path, struct load_arg *arg)
{
return obj_alloc_by_klass(path2class(path), arg, 0);
}
static VALUE
append_extmod(VALUE obj, VALUE extmod)
{
long i = RARRAY_LEN(extmod);
while (i > 0) {
VALUE m = RARRAY_AREF(extmod, --i);
rb_extend_object(obj, m);
}
return obj;
}
#define prohibit_ivar(type, str) do { \
if (!ivp || !*ivp) break; \
rb_raise(rb_eTypeError, \
"can't override instance variable of "type" `%"PRIsVALUE"'", \
(str)); \
} while (0)
static VALUE
r_object0(struct load_arg *arg, int *ivp, VALUE extmod)
{
VALUE v = Qnil;
int type = r_byte(arg);
long id;
st_data_t link;
switch (type) {
case TYPE_LINK:
id = r_long(arg);
if (!st_lookup(arg->data, (st_data_t)id, &link)) {
rb_raise(rb_eArgError, "dump format error (unlinked)");
}
v = (VALUE)link;
v = r_post_proc(v, arg);
break;
case TYPE_IVAR:
{
int ivar = TRUE;
v = r_object0(arg, &ivar, extmod);
if (ivar) r_ivar(v, NULL, arg);
}
break;
case TYPE_EXTENDED:
{
VALUE path = r_unique(arg);
VALUE m = rb_path_to_class(path);
if (NIL_P(extmod)) extmod = rb_ary_tmp_new(0);
if (RB_TYPE_P(m, T_CLASS)) { /* prepended */
VALUE c;
v = r_object0(arg, 0, Qnil);
c = CLASS_OF(v);
if (c != m || FL_TEST(c, FL_SINGLETON)) {
rb_raise(rb_eArgError,
"prepended class %"PRIsVALUE" differs from class %"PRIsVALUE,
path, rb_class_name(c));
}
c = rb_singleton_class(v);
while (RARRAY_LEN(extmod) > 0) {
m = rb_ary_pop(extmod);
rb_prepend_module(c, m);
}
}
else {
must_be_module(m, path);
rb_ary_push(extmod, m);
v = r_object0(arg, 0, extmod);
while (RARRAY_LEN(extmod) > 0) {
m = rb_ary_pop(extmod);
rb_extend_object(v, m);
}
}
}
break;
case TYPE_UCLASS:
{
VALUE c = path2class(r_unique(arg));
if (FL_TEST(c, FL_SINGLETON)) {
rb_raise(rb_eTypeError, "singleton can't be loaded");
}
v = r_object0(arg, 0, extmod);
if (rb_special_const_p(v) || RB_TYPE_P(v, T_OBJECT) || RB_TYPE_P(v, T_CLASS)) {
goto format_error;
}
if (RB_TYPE_P(v, T_MODULE) || !RTEST(rb_class_inherited_p(c, RBASIC(v)->klass))) {
VALUE tmp = rb_obj_alloc(c);
if (TYPE(v) != TYPE(tmp)) goto format_error;
}
* include/ruby/ruby.h: constify RBasic::klass and add RBASIC_CLASS(obj) macro which returns a class of `obj'. This change is a part of RGENGC branch [ruby-trunk - Feature #8339]. * object.c: add new function rb_obj_reveal(). This function reveal interal (hidden) object by rb_obj_hide(). Note that do not change class before and after hiding. Only permitted example is: klass = RBASIC_CLASS(obj); rb_obj_hide(obj); .... rb_obj_reveal(obj, klass); TODO: API design. rb_obj_reveal() should be replaced with others. TODO: modify constified variables using cast may be harmful for compiler's analysis and optimizaton. Any idea to prohibt inserting RBasic::klass directly? If rename RBasic::klass and force to use RBASIC_CLASS(obj), then all codes such as `RBASIC(obj)->klass' will be compilation error. Is it acceptable? (We have similar experience at Ruby 1.9, for example "RARRAY(ary)->ptr" to "RARRAY_PTR(ary)". * internal.h: add some macros. * RBASIC_CLEAR_CLASS(obj) clear RBasic::klass to make it internal object. * RBASIC_SET_CLASS(obj, cls) set RBasic::klass. * RBASIC_SET_CLASS_RAW(obj, cls) same as RBASIC_SET_CLASS without write barrier (planned). * RCLASS_SET_SUPER(a, b) set super class of a. * array.c, class.c, compile.c, encoding.c, enum.c, error.c, eval.c, file.c, gc.c, hash.c, io.c, iseq.c, marshal.c, object.c, parse.y, proc.c, process.c, random.c, ruby.c, sprintf.c, string.c, thread.c, transcode.c, vm.c, vm_eval.c, win32/file.c: Use above macros and functions to access RBasic::klass. * ext/coverage/coverage.c, ext/readline/readline.c, ext/socket/ancdata.c, ext/socket/init.c, * ext/zlib/zlib.c: ditto. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40691 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2013-05-13 06:49:11 -04:00
RBASIC_SET_CLASS(v, c);
}
break;
format_error:
rb_raise(rb_eArgError, "dump format error (user class)");
case TYPE_NIL:
v = Qnil;
v = r_leave(v, arg);
break;
case TYPE_TRUE:
v = Qtrue;
v = r_leave(v, arg);
break;
case TYPE_FALSE:
v = Qfalse;
v = r_leave(v, arg);
break;
case TYPE_FIXNUM:
{
long i = r_long(arg);
v = LONG2FIX(i);
}
v = r_leave(v, arg);
break;
case TYPE_FLOAT:
{
double d;
VALUE str = r_bytes(arg);
const char *ptr = RSTRING_PTR(str);
if (strcmp(ptr, "nan") == 0) {
d = nan("");
}
else if (strcmp(ptr, "inf") == 0) {
d = HUGE_VAL;
}
else if (strcmp(ptr, "-inf") == 0) {
d = -HUGE_VAL;
}
else {
char *e;
d = strtod(ptr, &e);
d = load_mantissa(d, e, RSTRING_LEN(str) - (e - ptr));
}
v = DBL2NUM(d);
v = r_entry(v, arg);
v = r_leave(v, arg);
}
break;
case TYPE_BIGNUM:
{
long len;
VALUE data;
int sign;
sign = r_byte(arg);
len = r_long(arg);
data = r_bytes0(len * 2, arg);
v = rb_integer_unpack(RSTRING_PTR(data), len, 2, 0,
INTEGER_PACK_LITTLE_ENDIAN | (sign == '-' ? INTEGER_PACK_NEGATIVE : 0));
rb_str_resize(data, 0L);
v = r_entry(v, arg);
v = r_leave(v, arg);
}
break;
case TYPE_STRING:
v = r_entry(r_string(arg), arg);
v = r_leave(v, arg);
break;
case TYPE_REGEXP:
{
VALUE str = r_bytes(arg);
int options = r_byte(arg);
int has_encoding = FALSE;
st_index_t idx = r_prepare(arg);
if (ivp) {
r_ivar(str, &has_encoding, arg);
*ivp = FALSE;
}
if (!has_encoding) {
/* 1.8 compatibility; remove escapes undefined in 1.8 */
char *ptr = RSTRING_PTR(str), *dst = ptr, *src = ptr;
long len = RSTRING_LEN(str);
long bs = 0;
for (; len-- > 0; *dst++ = *src++) {
switch (*src) {
case '\\': bs++; break;
case 'g': case 'h': case 'i': case 'j': case 'k': case 'l':
case 'm': case 'o': case 'p': case 'q': case 'u': case 'y':
case 'E': case 'F': case 'H': case 'I': case 'J': case 'K':
case 'L': case 'N': case 'O': case 'P': case 'Q': case 'R':
case 'S': case 'T': case 'U': case 'V': case 'X': case 'Y':
if (bs & 1) --dst;
/* fall through */
default: bs = 0; break;
}
}
rb_str_set_len(str, dst - ptr);
}
v = r_entry0(rb_reg_new_str(str, options), idx, arg);
v = r_leave(v, arg);
}
break;
case TYPE_ARRAY:
{
long len = r_long(arg);
v = rb_ary_new2(len);
v = r_entry(v, arg);
arg->readable += len - 1;
while (len--) {
rb_ary_push(v, r_object(arg));
arg->readable--;
}
v = r_leave(v, arg);
arg->readable++;
}
break;
case TYPE_HASH:
case TYPE_HASH_DEF:
{
long len = r_long(arg);
v = rb_hash_new_with_size(len);
v = r_entry(v, arg);
arg->readable += (len - 1) * 2;
while (len--) {
VALUE key = r_object(arg);
VALUE value = r_object(arg);
rb_hash_aset(v, key, value);
arg->readable -= 2;
}
arg->readable += 2;
if (type == TYPE_HASH_DEF) {
RHASH_SET_IFNONE(v, r_object(arg));
}
v = r_leave(v, arg);
}
break;
case TYPE_STRUCT:
{
VALUE mem, values;
long i;
VALUE slot;
st_index_t idx = r_prepare(arg);
VALUE klass = path2class(r_unique(arg));
long len = r_long(arg);
v = rb_obj_alloc(klass);
if (!RB_TYPE_P(v, T_STRUCT)) {
rb_raise(rb_eTypeError, "class %"PRIsVALUE" not a struct", rb_class_name(klass));
}
mem = rb_struct_s_members(klass);
if (RARRAY_LEN(mem) != len) {
rb_raise(rb_eTypeError, "struct %"PRIsVALUE" not compatible (struct size differs)",
rb_class_name(klass));
}
arg->readable += (len - 1) * 2;
v = r_entry0(v, idx, arg);
values = rb_ary_new2(len);
{
VALUE keywords = Qfalse;
if (RTEST(rb_struct_s_keyword_init(klass))) {
keywords = rb_hash_new();
rb_ary_push(values, keywords);
}
for (i=0; i<len; i++) {
VALUE n = rb_sym2str(RARRAY_AREF(mem, i));
slot = r_symbol(arg);
if (!rb_str_equal(n, slot)) {
rb_raise(rb_eTypeError, "struct %"PRIsVALUE" not compatible (:%"PRIsVALUE" for :%"PRIsVALUE")",
rb_class_name(klass),
slot, n);
}
if (keywords) {
rb_hash_aset(keywords, RARRAY_AREF(mem, i), r_object(arg));
}
else {
rb_ary_push(values, r_object(arg));
}
arg->readable -= 2;
}
}
rb_struct_initialize(v, values);
v = r_leave(v, arg);
arg->readable += 2;
}
break;
case TYPE_USERDEF:
{
VALUE name = r_unique(arg);
VALUE klass = path2class(name);
VALUE data;
st_data_t d;
if (!rb_obj_respond_to(klass, s_load, TRUE)) {
rb_raise(rb_eTypeError, "class %"PRIsVALUE" needs to have method `_load'",
name);
}
data = r_string(arg);
if (ivp) {
r_ivar(data, NULL, arg);
*ivp = FALSE;
}
v = load_funcall(arg, klass, s_load, 1, &data);
v = r_entry(v, arg);
if (st_lookup(compat_allocator_tbl, (st_data_t)rb_get_alloc_func(klass), &d)) {
marshal_compat_t *compat = (marshal_compat_t*)d;
v = compat->loader(klass, v);
}
v = r_post_proc(v, arg);
}
break;
case TYPE_USRMARSHAL:
{
VALUE name = r_unique(arg);
VALUE klass = path2class(name);
VALUE oldclass = 0;
VALUE data;
v = obj_alloc_by_klass(klass, arg, &oldclass);
if (!NIL_P(extmod)) {
/* for the case marshal_load is overridden */
append_extmod(v, extmod);
}
if (!rb_obj_respond_to(v, s_mload, TRUE)) {
rb_raise(rb_eTypeError, "instance of %"PRIsVALUE" needs to have method `marshal_load'",
name);
}
v = r_entry(v, arg);
data = r_object(arg);
load_funcall(arg, v, s_mload, 1, &data);
v = r_fixup_compat(v, arg);
v = r_copy_ivar(v, data);
v = r_post_proc(v, arg);
if (!NIL_P(extmod)) {
if (oldclass) append_extmod(v, extmod);
rb_ary_clear(extmod);
}
}
break;
case TYPE_OBJECT:
{
st_index_t idx = r_prepare(arg);
v = obj_alloc_by_path(r_unique(arg), arg);
if (!RB_TYPE_P(v, T_OBJECT)) {
rb_raise(rb_eArgError, "dump format error");
}
v = r_entry0(v, idx, arg);
r_ivar(v, NULL, arg);
v = r_leave(v, arg);
}
break;
case TYPE_DATA:
{
VALUE name = r_unique(arg);
VALUE klass = path2class(name);
VALUE oldclass = 0;
VALUE r;
v = obj_alloc_by_klass(klass, arg, &oldclass);
if (!RB_TYPE_P(v, T_DATA)) {
rb_raise(rb_eArgError, "dump format error");
}
v = r_entry(v, arg);
if (!rb_obj_respond_to(v, s_load_data, TRUE)) {
rb_raise(rb_eTypeError,
"class %"PRIsVALUE" needs to have instance method `_load_data'",
name);
}
r = r_object0(arg, 0, extmod);
load_funcall(arg, v, s_load_data, 1, &r);
v = r_leave(v, arg);
}
break;
case TYPE_MODULE_OLD:
{
VALUE str = r_bytes(arg);
v = rb_path_to_class(str);
prohibit_ivar("class/module", str);
v = r_entry(v, arg);
v = r_leave(v, arg);
}
break;
case TYPE_CLASS:
{
VALUE str = r_bytes(arg);
v = path2class(str);
prohibit_ivar("class", str);
v = r_entry(v, arg);
v = r_leave(v, arg);
}
break;
case TYPE_MODULE:
{
VALUE str = r_bytes(arg);
v = path2module(str);
prohibit_ivar("module", str);
v = r_entry(v, arg);
v = r_leave(v, arg);
}
break;
case TYPE_SYMBOL:
if (ivp) {
v = r_symreal(arg, *ivp);
*ivp = FALSE;
}
else {
v = r_symreal(arg, 0);
}
v = rb_str_intern(v);
v = r_leave(v, arg);
break;
case TYPE_SYMLINK:
v = rb_str_intern(r_symlink(arg));
break;
default:
rb_raise(rb_eArgError, "dump format error(0x%x)", type);
break;
}
if (v == Qundef) {
rb_raise(rb_eArgError, "dump format error (bad link)");
}
return v;
}
static VALUE
r_object(struct load_arg *arg)
{
return r_object0(arg, 0, Qnil);
}
static void
clear_load_arg(struct load_arg *arg)
{
if (arg->buf) {
xfree(arg->buf);
arg->buf = 0;
}
arg->buflen = 0;
arg->offset = 0;
arg->readable = 0;
if (!arg->symbols) return;
st_free_table(arg->symbols);
arg->symbols = 0;
st_free_table(arg->data);
arg->data = 0;
if (arg->compat_tbl) {
st_free_table(arg->compat_tbl);
arg->compat_tbl = 0;
}
}
/*
* call-seq:
* load( source [, proc] ) -> obj
* restore( source [, proc] ) -> obj
*
* Returns the result of converting the serialized data in source into a
* Ruby object (possibly with associated subordinate objects). source
* may be either an instance of IO or an object that responds to
* to_str. If proc is specified, each object will be passed to the proc, as the object
* is being deserialized.
*
* Never pass untrusted data (including user supplied input) to this method.
* Please see the overview for further details.
*/
static VALUE
marshal_load(int argc, VALUE *argv, VALUE _)
{
VALUE port, proc;
rb_check_arity(argc, 1, 2);
port = argv[0];
proc = argc > 1 ? argv[1] : Qnil;
return rb_marshal_load_with_proc(port, proc);
}
VALUE
rb_marshal_load_with_proc(VALUE port, VALUE proc)
{
int major, minor;
VALUE v;
VALUE wrapper; /* used to avoid memory leak in case of exception */
struct load_arg *arg;
v = rb_check_string_type(port);
if (!NIL_P(v)) {
port = v;
}
else if (rb_respond_to(port, s_getbyte) && rb_respond_to(port, s_read)) {
rb_check_funcall(port, s_binmode, 0, 0);
}
else {
io_needed();
}
wrapper = TypedData_Make_Struct(0, struct load_arg, &load_arg_data, arg);
arg->src = port;
arg->offset = 0;
arg->symbols = st_init_numtable();
arg->data = rb_init_identtable();
arg->compat_tbl = 0;
arg->proc = 0;
arg->readable = 0;
if (NIL_P(v))
arg->buf = xmalloc(BUFSIZ);
else
arg->buf = 0;
major = r_byte(arg);
minor = r_byte(arg);
if (major != MARSHAL_MAJOR || minor > MARSHAL_MINOR) {
clear_load_arg(arg);
rb_raise(rb_eTypeError, "incompatible marshal file format (can't be read)\n\
\tformat version %d.%d required; %d.%d given",
MARSHAL_MAJOR, MARSHAL_MINOR, major, minor);
}
if (RTEST(ruby_verbose) && minor != MARSHAL_MINOR) {
rb_warn("incompatible marshal file format (can be read)\n\
\tformat version %d.%d required; %d.%d given",
MARSHAL_MAJOR, MARSHAL_MINOR, major, minor);
}
if (!NIL_P(proc)) arg->proc = proc;
v = r_object(arg);
clear_load_arg(arg);
RB_GC_GUARD(wrapper);
return v;
}
/*
* The marshaling library converts collections of Ruby objects into a
* byte stream, allowing them to be stored outside the currently
* active script. This data may subsequently be read and the original
* objects reconstituted.
*
* Marshaled data has major and minor version numbers stored along
* with the object information. In normal use, marshaling can only
* load data written with the same major version number and an equal
* or lower minor version number. If Ruby's ``verbose'' flag is set
* (normally using -d, -v, -w, or --verbose) the major and minor
* numbers must match exactly. Marshal versioning is independent of
* Ruby's version numbers. You can extract the version by reading the
* first two bytes of marshaled data.
*
* str = Marshal.dump("thing")
* RUBY_VERSION #=> "1.9.0"
* str[0].ord #=> 4
* str[1].ord #=> 8
*
* Some objects cannot be dumped: if the objects to be dumped include
* bindings, procedure or method objects, instances of class IO, or
* singleton objects, a TypeError will be raised.
*
* If your class has special serialization needs (for example, if you
* want to serialize in some specific format), or if it contains
* objects that would otherwise not be serializable, you can implement
* your own serialization strategy.
*
* There are two methods of doing this, your object can define either
* marshal_dump and marshal_load or _dump and _load. marshal_dump will take
* precedence over _dump if both are defined. marshal_dump may result in
* smaller Marshal strings.
*
* == Security considerations
*
* By design, Marshal.load can deserialize almost any class loaded into the
* Ruby process. In many cases this can lead to remote code execution if the
* Marshal data is loaded from an untrusted source.
*
* As a result, Marshal.load is not suitable as a general purpose serialization
* format and you should never unmarshal user supplied input or other untrusted
* data.
*
* If you need to deserialize untrusted data, use JSON or another serialization
* format that is only able to load simple, 'primitive' types such as String,
* Array, Hash, etc. Never allow user input to specify arbitrary types to
* deserialize into.
*
* == marshal_dump and marshal_load
*
* When dumping an object the method marshal_dump will be called.
* marshal_dump must return a result containing the information necessary for
* marshal_load to reconstitute the object. The result can be any object.
*
* When loading an object dumped using marshal_dump the object is first
* allocated then marshal_load is called with the result from marshal_dump.
* marshal_load must recreate the object from the information in the result.
*
* Example:
*
* class MyObj
* def initialize name, version, data
* @name = name
* @version = version
* @data = data
* end
*
* def marshal_dump
* [@name, @version]
* end
*
* def marshal_load array
* @name, @version = array
* end
* end
*
* == _dump and _load
*
* Use _dump and _load when you need to allocate the object you're restoring
* yourself.
*
* When dumping an object the instance method _dump is called with an Integer
* which indicates the maximum depth of objects to dump (a value of -1 implies
* that you should disable depth checking). _dump must return a String
* containing the information necessary to reconstitute the object.
*
* The class method _load should take a String and use it to return an object
* of the same class.
*
* Example:
*
* class MyObj
* def initialize name, version, data
* @name = name
* @version = version
* @data = data
* end
*
* def _dump level
* [@name, @version].join ':'
* end
*
* def self._load args
* new(*args.split(':'))
* end
* end
*
* Since Marshal.dump outputs a string you can have _dump return a Marshal
* string which is Marshal.loaded in _load for complex objects.
*/
void
Init_marshal(void)
{
VALUE rb_mMarshal = rb_define_module("Marshal");
#define set_id(sym) sym = rb_intern_const(name_##sym)
set_id(s_dump);
set_id(s_load);
set_id(s_mdump);
set_id(s_mload);
set_id(s_dump_data);
set_id(s_load_data);
set_id(s_alloc);
set_id(s_call);
set_id(s_getbyte);
set_id(s_read);
set_id(s_write);
set_id(s_binmode);
set_id(s_encoding_short);
set_id(s_ruby2_keywords_flag);
rb_define_module_function(rb_mMarshal, "dump", marshal_dump, -1);
rb_define_module_function(rb_mMarshal, "load", marshal_load, -1);
rb_define_module_function(rb_mMarshal, "restore", marshal_load, -1);
/* major version */
rb_define_const(rb_mMarshal, "MAJOR_VERSION", INT2FIX(MARSHAL_MAJOR));
/* minor version */
rb_define_const(rb_mMarshal, "MINOR_VERSION", INT2FIX(MARSHAL_MINOR));
}
static st_table *
compat_allocator_table(void)
{
if (compat_allocator_tbl) return compat_allocator_tbl;
compat_allocator_tbl = st_init_numtable();
#undef RUBY_UNTYPED_DATA_WARNING
#define RUBY_UNTYPED_DATA_WARNING 0
compat_allocator_tbl_wrapper =
Data_Wrap_Struct(0, mark_marshal_compat_t, 0, compat_allocator_tbl);
rb_gc_register_mark_object(compat_allocator_tbl_wrapper);
return compat_allocator_tbl;
}
VALUE
rb_marshal_dump(VALUE obj, VALUE port)
{
return rb_marshal_dump_limited(obj, port, -1);
}
VALUE
rb_marshal_load(VALUE port)
{
return rb_marshal_load_with_proc(port, Qnil);
}