ruby/math.c

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

1147 строки
27 KiB
C
Исходник Обычный вид История

/**********************************************************************
math.c -
$Author$
created at: Tue Jan 25 14:12:56 JST 1994
* 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
2007-08-25 07:29:39 +04:00
Copyright (C) 1993-2007 Yukihiro Matsumoto
**********************************************************************/
#include "ruby/internal/config.h"
#ifdef _MSC_VER
# define _USE_MATH_DEFINES 1
#endif
#include <errno.h>
#include <float.h>
#include <math.h>
#include "internal.h"
#include "internal/bignum.h"
#include "internal/complex.h"
#include "internal/math.h"
#include "internal/object.h"
#include "internal/vm.h"
VALUE rb_mMath;
VALUE rb_eMathDomainError;
#define Get_Double(x) rb_num_to_dbl(x)
#define domain_error(msg) \
rb_raise(rb_eMathDomainError, "Numerical argument is out of domain - " msg)
2021-07-04 13:22:11 +03:00
#define domain_check_min(val, min, msg) \
((val) < (min) ? domain_error(msg) : (void)0)
#define domain_check_range(val, min, max, msg) \
((val) < (min) || (max) < (val) ? domain_error(msg) : (void)0)
/*
* call-seq:
* Math.atan2(y, x) -> float
*
* Returns the {arc tangent}[https://en.wikipedia.org/wiki/Atan2] of +y+ and +x+
* in {radians}[https://en.wikipedia.org/wiki/Trigonometric_functions#Radians_versus_degrees].
*
* - Domain of +y+: <tt>[-INFINITY, INFINITY]</tt>.
* - Domain of +x+: <tt>[-INFINITY, INFINITY]</tt>.
* - Range: <tt>[-PI, PI]</tt>.
*
* Examples:
*
* atan2(-1.0, -1.0) # => -2.356194490192345 # -3*PI/4
* atan2(-1.0, 0.0) # => -1.5707963267948966 # -PI/2
* atan2(-1.0, 1.0) # => -0.7853981633974483 # -PI/4
* atan2(0.0, -1.0) # => 3.141592653589793 # PI
*
*/
static VALUE
math_atan2(VALUE unused_obj, VALUE y, VALUE x)
{
double dx, dy;
dx = Get_Double(x);
dy = Get_Double(y);
if (dx == 0.0 && dy == 0.0) {
if (!signbit(dx))
return DBL2NUM(dy);
if (!signbit(dy))
return DBL2NUM(M_PI);
return DBL2NUM(-M_PI);
}
#ifndef ATAN2_INF_C99
if (isinf(dx) && isinf(dy)) {
/* optimization for FLONUM */
if (dx < 0.0) {
const double dz = (3.0 * M_PI / 4.0);
return (dy < 0.0) ? DBL2NUM(-dz) : DBL2NUM(dz);
}
else {
const double dz = (M_PI / 4.0);
return (dy < 0.0) ? DBL2NUM(-dz) : DBL2NUM(dz);
}
}
#endif
return DBL2NUM(atan2(dy, dx));
}
/*
* call-seq:
* Math.cos(x) -> float
*
* Returns the
* {cosine}[https://en.wikipedia.org/wiki/Sine_and_cosine] of +x+
* in {radians}[https://en.wikipedia.org/wiki/Trigonometric_functions#Radians_versus_degrees].
*
* - Domain: <tt>(-INFINITY, INFINITY)</tt>.
* - Range: <tt>[-1.0, 1.0]</tt>.
*
* Examples:
*
* cos(-PI) # => -1.0
* cos(-PI/2) # => 6.123031769111886e-17 # 0.0000000000000001
* cos(0.0) # => 1.0
* cos(PI/2) # => 6.123031769111886e-17 # 0.0000000000000001
* cos(PI) # => -1.0
*
*/
static VALUE
math_cos(VALUE unused_obj, VALUE x)
{
return DBL2NUM(cos(Get_Double(x)));
}
/*
* call-seq:
* Math.sin(x) -> float
*
* Returns the
* {sine}[https://en.wikipedia.org/wiki/Sine_and_cosine] of +x+
* in {radians}[https://en.wikipedia.org/wiki/Trigonometric_functions#Radians_versus_degrees].
*
* - Domain: <tt>(-INFINITY, INFINITY)</tt>.
* - Range: <tt>[-1.0, 1.0]</tt>.
*
* Examples:
*
* sin(-PI) # => -1.2246063538223773e-16 # -0.0000000000000001
* sin(-PI/2) # => -1.0
* sin(0.0) # => 0.0
* sin(PI/2) # => 1.0
* sin(PI) # => 1.2246063538223773e-16 # 0.0000000000000001
*
*/
static VALUE
math_sin(VALUE unused_obj, VALUE x)
{
return DBL2NUM(sin(Get_Double(x)));
}
/*
* call-seq:
* Math.tan(x) -> float
*
* Returns the
* {tangent}[https://en.wikipedia.org/wiki/Trigonometric_functions] of +x+
* in {radians}[https://en.wikipedia.org/wiki/Trigonometric_functions#Radians_versus_degrees].
*
* - Domain: <tt>(-INFINITY, INFINITY)</tt>.
* - Range: <tt>(-INFINITY, INFINITY)</tt>.
*
* Examples:
*
* tan(-PI) # => 1.2246467991473532e-16 # -0.0000000000000001
* tan(-PI/2) # => -1.633123935319537e+16 # -16331239353195370.0
* tan(0.0) # => 0.0
* tan(PI/2) # => 1.633123935319537e+16 # 16331239353195370.0
* tan(PI) # => -1.2246467991473532e-16 # -0.0000000000000001
*
*/
static VALUE
math_tan(VALUE unused_obj, VALUE x)
{
return DBL2NUM(tan(Get_Double(x)));
}
2023-05-12 06:03:19 +03:00
#define math_arc(num, func) \
double d; \
d = Get_Double((num)); \
domain_check_range(d, -1.0, 1.0, #func); \
return DBL2NUM(func(d));
/*
* call-seq:
* Math.acos(x) -> float
*
* Returns the {arc cosine}[https://en.wikipedia.org/wiki/Inverse_trigonometric_functions] of +x+.
*
* - Domain: <tt>[-1, 1]</tt>.
* - Range: <tt>[0, PI]</tt>.
*
* Examples:
*
* acos(-1.0) # => 3.141592653589793 # PI
* acos(0.0) # => 1.5707963267948966 # PI/2
* acos(1.0) # => 0.0
*
*/
static VALUE
math_acos(VALUE unused_obj, VALUE x)
{
2023-05-12 06:03:19 +03:00
math_arc(x, acos)
}
/*
* call-seq:
* Math.asin(x) -> float
*
* Returns the {arc sine}[https://en.wikipedia.org/wiki/Inverse_trigonometric_functions] of +x+.
*
* - Domain: <tt>[-1, -1]</tt>.
* - Range: <tt>[-PI/2, PI/2]</tt>.
*
* Examples:
*
* asin(-1.0) # => -1.5707963267948966 # -PI/2
* asin(0.0) # => 0.0
* asin(1.0) # => 1.5707963267948966 # PI/2
*
*/
static VALUE
math_asin(VALUE unused_obj, VALUE x)
{
2023-05-12 06:03:19 +03:00
math_arc(x, asin)
}
/*
* call-seq:
* Math.atan(x) -> Float
*
* Returns the {arc tangent}[https://en.wikipedia.org/wiki/Inverse_trigonometric_functions] of +x+.
*
* - Domain: <tt>[-INFINITY, INFINITY]</tt>.
* - Range: <tt>[-PI/2, PI/2] </tt>.
*
* Examples:
*
* atan(-INFINITY) # => -1.5707963267948966 # -PI2
* atan(-PI) # => -1.2626272556789115
* atan(-PI/2) # => -1.0038848218538872
* atan(0.0) # => 0.0
* atan(PI/2) # => 1.0038848218538872
* atan(PI) # => 1.2626272556789115
* atan(INFINITY) # => 1.5707963267948966 # PI/2
*
*/
static VALUE
math_atan(VALUE unused_obj, VALUE x)
{
return DBL2NUM(atan(Get_Double(x)));
}
#ifndef HAVE_COSH
double
cosh(double x)
{
return (exp(x) + exp(-x)) / 2;
}
#endif
/*
* call-seq:
* Math.cosh(x) -> float
*
* Returns the {hyperbolic cosine}[https://en.wikipedia.org/wiki/Hyperbolic_functions] of +x+
* in {radians}[https://en.wikipedia.org/wiki/Trigonometric_functions#Radians_versus_degrees].
*
* - Domain: <tt>[-INFINITY, INFINITY]</tt>.
* - Range: <tt>[1, INFINITY]</tt>.
*
* Examples:
*
* cosh(-INFINITY) # => Infinity
* cosh(0.0) # => 1.0
* cosh(INFINITY) # => Infinity
*
*/
static VALUE
math_cosh(VALUE unused_obj, VALUE x)
{
return DBL2NUM(cosh(Get_Double(x)));
}
#ifndef HAVE_SINH
double
sinh(double x)
{
return (exp(x) - exp(-x)) / 2;
}
#endif
/*
* call-seq:
* Math.sinh(x) -> float
*
* Returns the {hyperbolic sine}[https://en.wikipedia.org/wiki/Hyperbolic_functions] of +x+
* in {radians}[https://en.wikipedia.org/wiki/Trigonometric_functions#Radians_versus_degrees].
*
* - Domain: <tt>[-INFINITY, INFINITY]</tt>.
* - Range: <tt>[-INFINITY, INFINITY]</tt>.
*
* Examples:
*
* sinh(-INFINITY) # => -Infinity
* sinh(0.0) # => 0.0
* sinh(INFINITY) # => Infinity
*
*/
static VALUE
math_sinh(VALUE unused_obj, VALUE x)
{
return DBL2NUM(sinh(Get_Double(x)));
}
#ifndef HAVE_TANH
double
tanh(double x)
{
# if defined(HAVE_SINH) && defined(HAVE_COSH)
const double c = cosh(x);
if (!isinf(c)) return sinh(x) / c;
# else
const double e = exp(x+x);
if (!isinf(e)) return (e - 1) / (e + 1);
# endif
return x > 0 ? 1.0 : -1.0;
}
#endif
/*
* call-seq:
* Math.tanh(x) -> float
*
* Returns the {hyperbolic tangent}[https://en.wikipedia.org/wiki/Hyperbolic_functions] of +x+
* in {radians}[https://en.wikipedia.org/wiki/Trigonometric_functions#Radians_versus_degrees].
*
* - Domain: <tt>[-INFINITY, INFINITY]</tt>.
* - Range: <tt>[-1, 1]</tt>.
*
* Examples:
*
* tanh(-INFINITY) # => -1.0
* tanh(0.0) # => 0.0
* tanh(INFINITY) # => 1.0
*
*/
static VALUE
math_tanh(VALUE unused_obj, VALUE x)
{
return DBL2NUM(tanh(Get_Double(x)));
}
/*
* call-seq:
* Math.acosh(x) -> float
*
* Returns the {inverse hyperbolic cosine}[https://en.wikipedia.org/wiki/Inverse_hyperbolic_functions] of +x+.
*
* - Domain: <tt>[1, INFINITY]</tt>.
* - Range: <tt>[0, INFINITY]</tt>.
*
* Examples:
*
* acosh(1.0) # => 0.0
* acosh(INFINITY) # => Infinity
*
*/
static VALUE
math_acosh(VALUE unused_obj, VALUE x)
{
double d;
d = Get_Double(x);
2021-07-04 13:22:11 +03:00
domain_check_min(d, 1.0, "acosh");
return DBL2NUM(acosh(d));
}
/*
* call-seq:
* Math.asinh(x) -> float
*
* Returns the {inverse hyperbolic sine}[https://en.wikipedia.org/wiki/Inverse_hyperbolic_functions] of +x+.
*
* - Domain: <tt>[-INFINITY, INFINITY]</tt>.
* - Range: <tt>[-INFINITY, INFINITY]</tt>.
*
* Examples:
*
* asinh(-INFINITY) # => -Infinity
* asinh(0.0) # => 0.0
* asinh(INFINITY) # => Infinity
*
*/
static VALUE
math_asinh(VALUE unused_obj, VALUE x)
{
return DBL2NUM(asinh(Get_Double(x)));
}
/*
* call-seq:
* Math.atanh(x) -> float
*
* Returns the {inverse hyperbolic tangent}[https://en.wikipedia.org/wiki/Inverse_hyperbolic_functions] of +x+.
*
* - Domain: <tt>[-1, 1]</tt>.
* - Range: <tt>[-INFINITY, INFINITY]</tt>.
*
* Examples:
*
* atanh(-1.0) # => -Infinity
* atanh(0.0) # => 0.0
* atanh(1.0) # => Infinity
*
*/
static VALUE
math_atanh(VALUE unused_obj, VALUE x)
{
double d;
d = Get_Double(x);
2021-07-04 13:22:11 +03:00
domain_check_range(d, -1.0, +1.0, "atanh");
/* check for pole error */
if (d == -1.0) return DBL2NUM(-HUGE_VAL);
if (d == +1.0) return DBL2NUM(+HUGE_VAL);
return DBL2NUM(atanh(d));
}
/*
* call-seq:
* Math.exp(x) -> float
*
* Returns +e+ raised to the +x+ power.
*
* - Domain: <tt>[-INFINITY, INFINITY]</tt>.
* - Range: <tt>[0, INFINITY]</tt>.
*
* Examples:
*
* exp(-INFINITY) # => 0.0
* exp(-1.0) # => 0.36787944117144233 # 1.0/E
* exp(0.0) # => 1.0
* exp(0.5) # => 1.6487212707001282 # sqrt(E)
* exp(1.0) # => 2.718281828459045 # E
* exp(2.0) # => 7.38905609893065 # E**2
* exp(INFINITY) # => Infinity
*
*/
static VALUE
math_exp(VALUE unused_obj, VALUE x)
{
return DBL2NUM(exp(Get_Double(x)));
}
#if defined __CYGWIN__
# include <cygwin/version.h>
# if CYGWIN_VERSION_DLL_MAJOR < 1005
# define nan(x) nan()
# endif
# define log(x) ((x) < 0.0 ? nan("") : log(x))
# define log10(x) ((x) < 0.0 ? nan("") : log10(x))
#endif
#ifndef M_LN2
# define M_LN2 0.693147180559945309417232121458176568
#endif
#ifndef M_LN10
# define M_LN10 2.30258509299404568401799145468436421
#endif
FUNC_MINIMIZED(static VALUE math_log(int, const VALUE *, VALUE));
/*
* call-seq:
* Math.log(x, base = Math::E) -> Float
*
* Returns the base +base+ {logarithm}[https://en.wikipedia.org/wiki/Logarithm] of +x+.
*
* - Domain: <tt>[0, INFINITY]</tt>.
* - Range: <tt>[-INFINITY, INFINITY)]</tt>.
*
* Examples:
*
* log(0.0) # => -Infinity
* log(1.0) # => 0.0
* log(E) # => 1.0
* log(INFINITY) # => Infinity
*
* log(0.0, 2.0) # => -Infinity
* log(1.0, 2.0) # => 0.0
* log(2.0, 2.0) # => 1.0
*
* log(0.0, 10.0) # => -Infinity
* log(1.0, 10.0) # => 0.0
* log(10.0, 10.0) # => 1.0
*
*/
static VALUE
math_log(int argc, const VALUE *argv, VALUE unused_obj)
{
return rb_math_log(argc, argv);
}
static double
get_double_rshift(VALUE x, size_t *pnumbits)
{
size_t numbits;
if (RB_BIGNUM_TYPE_P(x) && BIGNUM_POSITIVE_P(x) &&
DBL_MAX_EXP <= (numbits = rb_absint_numwords(x, 1, NULL))) {
numbits -= DBL_MANT_DIG;
x = rb_big_rshift(x, SIZET2NUM(numbits));
}
else {
numbits = 0;
}
*pnumbits = numbits;
return Get_Double(x);
}
static double
math_log_split(VALUE x, size_t *numbits)
{
double d = get_double_rshift(x, numbits);
2021-07-04 13:22:11 +03:00
domain_check_min(d, 0.0, "log");
return d;
}
#if defined(log2) || defined(HAVE_LOG2)
# define log_intermediate log2
#else
# define log_intermediate log10
double log2(double x);
#endif
VALUE
rb_math_log(int argc, const VALUE *argv)
{
VALUE x, base;
double d;
size_t numbits;
argc = rb_scan_args(argc, argv, "11", &x, &base);
d = math_log_split(x, &numbits);
if (argc == 2) {
size_t numbits_2;
double b = math_log_split(base, &numbits_2);
/* check for pole error */
if (d == 0.0) {
2023-07-17 17:57:11 +03:00
// Already DomainError if b < 0.0
return b ? DBL2NUM(-HUGE_VAL) : DBL2NUM(NAN);
}
else if (b == 0.0) {
return DBL2NUM(-0.0);
}
d = log_intermediate(d) / log_intermediate(b);
d += (numbits - numbits_2) / log2(b);
}
else {
/* check for pole error */
if (d == 0.0) return DBL2NUM(-HUGE_VAL);
d = log(d);
d += numbits * M_LN2;
}
return DBL2NUM(d);
* sprintf.c (rb_str_format): allow %c to print one character string (e.g. ?x). * lib/tempfile.rb (Tempfile::make_tmpname): put dot between basename and pid. [ruby-talk:196272] * parse.y (do_block): remove -> style block. * parse.y (parser_yylex): remove tLAMBDA_ARG. * eval.c (rb_call0): binding for the return event hook should have consistent scope. [ruby-core:07928] * eval.c (proc_invoke): return behavior should depend whether it is surrounded by a lambda or a mere block. * eval.c (formal_assign): handles post splat arguments. * eval.c (rb_call0): ditto. * st.c (strhash): use FNV-1a hash. * parse.y (parser_yylex): removed experimental ';;' terminator. * eval.c (rb_node_arity): should be aware of post splat arguments. * eval.c (rb_proc_arity): ditto. * parse.y (f_args): syntax rule enhanced to support arguments after the splat. * parse.y (block_param): ditto for block parameters. * parse.y (f_post_arg): mandatory formal arguments after the splat argument. * parse.y (new_args_gen): generate nodes for mandatory formal arguments after the splat argument. * eval.c (rb_eval): dispatch mandatory formal arguments after the splat argument. * parse.y (args): allow more than one splat in the argument list. * parse.y (method_call): allow aref [] to accept all kind of method argument, including assocs, splat, and block argument. * eval.c (SETUP_ARGS0): prepare block argument as well. * lib/mathn.rb (Integer): remove Integer#gcd2. [ruby-core:07931] * eval.c (error_line): print receivers true/false/nil specially. * eval.c (rb_proc_yield): handles parameters in yield semantics. * eval.c (nil_yield): gives LocalJumpError to denote no block error. * io.c (rb_io_getc): now takes one-character string. * string.c (rb_str_hash): use FNV-1a hash from Fowler/Noll/Vo hashing algorithm. * string.c (rb_str_aref): str[0] now returns 1 character string, instead of a fixnum. [Ruby2] * parse.y (parser_yylex): ?c now returns 1 character string, instead of a fixnum. [Ruby2] * string.c (rb_str_aset): no longer support fixnum insertion. * eval.c (umethod_bind): should not update original class. [ruby-dev:28636] * eval.c (ev_const_get): should support constant access from within instance_eval(). [ruby-dev:28327] * time.c (time_timeval): should round for usec floating number. [ruby-core:07896] * time.c (time_add): ditto. * dir.c (sys_warning): should not call a vararg function rb_sys_warning() indirectly. [ruby-core:07886] * numeric.c (flo_divmod): the first element of Float#divmod should be an integer. [ruby-dev:28589] * test/ruby/test_float.rb: add tests for divmod, div, modulo and remainder. * re.c (rb_reg_initialize): should not allow modifying literal regexps. frozen check moved from rb_reg_initialize_m as well. * re.c (rb_reg_initialize): should not modify untainted objects in safe levels higher than 3. * re.c (rb_memcmp): type change from char* to const void*. * dir.c (dir_close): should not close untainted dir stream. * dir.c (GetDIR): add tainted/frozen check for each dir operation. * lib/rdoc/parsers/parse_rb.rb (RDoc::RubyParser::parse_symbol_arg): typo fixed. a patch from Florian Gross <florg at florg.net>. * eval.c (EXEC_EVENT_HOOK): trace_func may remove itself from event_hooks. no guarantee for arbitrary hook deletion. [ruby-dev:28632] * util.c (ruby_strtod): differ addition to minimize error. [ruby-dev:28619] * util.c (ruby_strtod): should not raise ERANGE when the input string does not have any digits. [ruby-dev:28629] * eval.c (proc_invoke): should restore old ruby_frame->block. thanks to ts <decoux at moulon.inra.fr>. [ruby-core:07833] also fix [ruby-dev:28614] as well. * signal.c (trap): sig should be less then NSIG. Coverity found this bug. a patch from Kevin Tew <tewk at tewk.com>. [ruby-core:07823] * math.c (math_log2): add new method inspired by [ruby-talk:191237]. * math.c (math_log): add optional base argument to Math::log(). [ruby-talk:191308] * ext/syck/emitter.c (syck_scan_scalar): avoid accessing uninitialized array element. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07809] * array.c (rb_ary_fill): initialize local variables first. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07810] * ext/syck/yaml2byte.c (syck_yaml2byte_handler): need to free type_tag. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07808] * ext/socket/socket.c (make_hostent_internal): accept ai_family check from Sam Roberts <sroberts at uniserve.com>. [ruby-core:07691] * util.c (ruby_strtod): should not cut off 18 digits for no reason. [ruby-core:07796] * array.c (rb_ary_fill): internalize local variable "beg" to pacify Coverity. [ruby-core:07770] * pack.c (pack_unpack): now supports CRLF newlines. a patch from <tommy at tmtm.org>. [ruby-dev:28601] * applied code clean-up patch from Stefan Huehner <stefan at huehner.org>. [ruby-core:07764] * lib/jcode.rb (String::tr_s): should have translated non squeezing character sequence (i.e. a character) as well. thanks to Hiroshi Ichikawa <gimite at gimite.ddo.jp> [ruby-list:42090] * ext/socket/socket.c: document update patch from Sam Roberts <sroberts at uniserve.com>. [ruby-core:07701] * lib/mathn.rb (Integer): need not to remove gcd2. a patch from NARUSE, Yui <naruse at airemix.com>. [ruby-dev:28570] * parse.y (arg): too much NEW_LIST() * eval.c (SETUP_ARGS0): remove unnecessary access to nd_alen. * eval.c (rb_eval): use ARGSCAT for NODE_OP_ASGN1. [ruby-dev:28585] * parse.y (arg): use NODE_ARGSCAT for placeholder. * lib/getoptlong.rb (GetoptLong::get): RDoc update patch from mathew <meta at pobox.com>. [ruby-core:07738] * variable.c (rb_const_set): raise error when no target klass is supplied. [ruby-dev:28582] * prec.c (prec_prec_f): documentation patch from <gerardo.santana at gmail.com>. [ruby-core:07689] * bignum.c (rb_big_pow): second operand may be too big even if it's a Fixnum. [ruby-talk:187984] * README.EXT: update symbol description. [ruby-talk:188104] * COPYING: explicitly note GPLv2. [ruby-talk:187922] * parse.y: remove some obsolete syntax rules (unparenthesized method calls in argument list). * eval.c (rb_call0): insecure calling should be checked for non NODE_SCOPE method invocations too. * eval.c (rb_alias): should preserve the current safe level as well as method definition. * process.c (rb_f_sleep): remove RDoc description about SIGALRM which is not valid on the current implementation. [ruby-dev:28464] Thu Mar 23 21:40:47 2006 K.Kosako <sndgk393 AT ybb.ne.jp> * eval.c (method_missing): should support argument splat in super. a bug in combination of super, splat and method_missing. [ruby-talk:185438] * configure.in: Solaris SunPro compiler -rapth patch from <kuwa at labs.fujitsu.com>. [ruby-dev:28443] * configure.in: remove enable_rpath=no for Solaris. [ruby-dev:28440] * ext/win32ole/win32ole.c (ole_val2olevariantdata): change behavior of converting OLE Variant object with VT_ARRAY|VT_UI1 and Ruby String object. * ruby.1: a clarification patch from David Lutterkort <dlutter at redhat.com>. [ruby-core:7508] * lib/rdoc/ri/ri_paths.rb (RI::Paths): adding paths from rubygems directories. a patch from Eric Hodel <drbrain at segment7.net>. [ruby-core:07423] * eval.c (rb_clear_cache_by_class): clearing wrong cache. * ext/extmk.rb: use :remove_destination to install extension libraries to avoid SEGV. [ruby-dev:28417] * eval.c (rb_thread_fd_writable): should not re-schedule output from KILLED thread (must be error printing). * array.c (rb_ary_flatten_bang): allow specifying recursion level. [ruby-talk:182170] * array.c (rb_ary_flatten): ditto. * gc.c (add_heap): a heap_slots may overflow. a patch from Stefan Weil <weil at mail.berlios.de>. * eval.c (rb_call): use separate cache for fcall/vcall invocation. * eval.c (rb_eval): NODE_FCALL, NODE_VCALL can call local functions. * eval.c (rb_mod_local): a new method to specify newly added visibility "local". * eval.c (search_method): search for local methods which are visible only from the current class. * class.c (rb_class_local_methods): a method to list local methods. * object.c (Init_Object): add BasicObject class as a top level BlankSlate class. * ruby.h (SYM2ID): should not cast to signed long. [ruby-core:07414] * class.c (rb_include_module): allow module duplication. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@10235 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2006-06-10 01:20:17 +04:00
}
#ifndef log2
* sprintf.c (rb_str_format): allow %c to print one character string (e.g. ?x). * lib/tempfile.rb (Tempfile::make_tmpname): put dot between basename and pid. [ruby-talk:196272] * parse.y (do_block): remove -> style block. * parse.y (parser_yylex): remove tLAMBDA_ARG. * eval.c (rb_call0): binding for the return event hook should have consistent scope. [ruby-core:07928] * eval.c (proc_invoke): return behavior should depend whether it is surrounded by a lambda or a mere block. * eval.c (formal_assign): handles post splat arguments. * eval.c (rb_call0): ditto. * st.c (strhash): use FNV-1a hash. * parse.y (parser_yylex): removed experimental ';;' terminator. * eval.c (rb_node_arity): should be aware of post splat arguments. * eval.c (rb_proc_arity): ditto. * parse.y (f_args): syntax rule enhanced to support arguments after the splat. * parse.y (block_param): ditto for block parameters. * parse.y (f_post_arg): mandatory formal arguments after the splat argument. * parse.y (new_args_gen): generate nodes for mandatory formal arguments after the splat argument. * eval.c (rb_eval): dispatch mandatory formal arguments after the splat argument. * parse.y (args): allow more than one splat in the argument list. * parse.y (method_call): allow aref [] to accept all kind of method argument, including assocs, splat, and block argument. * eval.c (SETUP_ARGS0): prepare block argument as well. * lib/mathn.rb (Integer): remove Integer#gcd2. [ruby-core:07931] * eval.c (error_line): print receivers true/false/nil specially. * eval.c (rb_proc_yield): handles parameters in yield semantics. * eval.c (nil_yield): gives LocalJumpError to denote no block error. * io.c (rb_io_getc): now takes one-character string. * string.c (rb_str_hash): use FNV-1a hash from Fowler/Noll/Vo hashing algorithm. * string.c (rb_str_aref): str[0] now returns 1 character string, instead of a fixnum. [Ruby2] * parse.y (parser_yylex): ?c now returns 1 character string, instead of a fixnum. [Ruby2] * string.c (rb_str_aset): no longer support fixnum insertion. * eval.c (umethod_bind): should not update original class. [ruby-dev:28636] * eval.c (ev_const_get): should support constant access from within instance_eval(). [ruby-dev:28327] * time.c (time_timeval): should round for usec floating number. [ruby-core:07896] * time.c (time_add): ditto. * dir.c (sys_warning): should not call a vararg function rb_sys_warning() indirectly. [ruby-core:07886] * numeric.c (flo_divmod): the first element of Float#divmod should be an integer. [ruby-dev:28589] * test/ruby/test_float.rb: add tests for divmod, div, modulo and remainder. * re.c (rb_reg_initialize): should not allow modifying literal regexps. frozen check moved from rb_reg_initialize_m as well. * re.c (rb_reg_initialize): should not modify untainted objects in safe levels higher than 3. * re.c (rb_memcmp): type change from char* to const void*. * dir.c (dir_close): should not close untainted dir stream. * dir.c (GetDIR): add tainted/frozen check for each dir operation. * lib/rdoc/parsers/parse_rb.rb (RDoc::RubyParser::parse_symbol_arg): typo fixed. a patch from Florian Gross <florg at florg.net>. * eval.c (EXEC_EVENT_HOOK): trace_func may remove itself from event_hooks. no guarantee for arbitrary hook deletion. [ruby-dev:28632] * util.c (ruby_strtod): differ addition to minimize error. [ruby-dev:28619] * util.c (ruby_strtod): should not raise ERANGE when the input string does not have any digits. [ruby-dev:28629] * eval.c (proc_invoke): should restore old ruby_frame->block. thanks to ts <decoux at moulon.inra.fr>. [ruby-core:07833] also fix [ruby-dev:28614] as well. * signal.c (trap): sig should be less then NSIG. Coverity found this bug. a patch from Kevin Tew <tewk at tewk.com>. [ruby-core:07823] * math.c (math_log2): add new method inspired by [ruby-talk:191237]. * math.c (math_log): add optional base argument to Math::log(). [ruby-talk:191308] * ext/syck/emitter.c (syck_scan_scalar): avoid accessing uninitialized array element. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07809] * array.c (rb_ary_fill): initialize local variables first. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07810] * ext/syck/yaml2byte.c (syck_yaml2byte_handler): need to free type_tag. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07808] * ext/socket/socket.c (make_hostent_internal): accept ai_family check from Sam Roberts <sroberts at uniserve.com>. [ruby-core:07691] * util.c (ruby_strtod): should not cut off 18 digits for no reason. [ruby-core:07796] * array.c (rb_ary_fill): internalize local variable "beg" to pacify Coverity. [ruby-core:07770] * pack.c (pack_unpack): now supports CRLF newlines. a patch from <tommy at tmtm.org>. [ruby-dev:28601] * applied code clean-up patch from Stefan Huehner <stefan at huehner.org>. [ruby-core:07764] * lib/jcode.rb (String::tr_s): should have translated non squeezing character sequence (i.e. a character) as well. thanks to Hiroshi Ichikawa <gimite at gimite.ddo.jp> [ruby-list:42090] * ext/socket/socket.c: document update patch from Sam Roberts <sroberts at uniserve.com>. [ruby-core:07701] * lib/mathn.rb (Integer): need not to remove gcd2. a patch from NARUSE, Yui <naruse at airemix.com>. [ruby-dev:28570] * parse.y (arg): too much NEW_LIST() * eval.c (SETUP_ARGS0): remove unnecessary access to nd_alen. * eval.c (rb_eval): use ARGSCAT for NODE_OP_ASGN1. [ruby-dev:28585] * parse.y (arg): use NODE_ARGSCAT for placeholder. * lib/getoptlong.rb (GetoptLong::get): RDoc update patch from mathew <meta at pobox.com>. [ruby-core:07738] * variable.c (rb_const_set): raise error when no target klass is supplied. [ruby-dev:28582] * prec.c (prec_prec_f): documentation patch from <gerardo.santana at gmail.com>. [ruby-core:07689] * bignum.c (rb_big_pow): second operand may be too big even if it's a Fixnum. [ruby-talk:187984] * README.EXT: update symbol description. [ruby-talk:188104] * COPYING: explicitly note GPLv2. [ruby-talk:187922] * parse.y: remove some obsolete syntax rules (unparenthesized method calls in argument list). * eval.c (rb_call0): insecure calling should be checked for non NODE_SCOPE method invocations too. * eval.c (rb_alias): should preserve the current safe level as well as method definition. * process.c (rb_f_sleep): remove RDoc description about SIGALRM which is not valid on the current implementation. [ruby-dev:28464] Thu Mar 23 21:40:47 2006 K.Kosako <sndgk393 AT ybb.ne.jp> * eval.c (method_missing): should support argument splat in super. a bug in combination of super, splat and method_missing. [ruby-talk:185438] * configure.in: Solaris SunPro compiler -rapth patch from <kuwa at labs.fujitsu.com>. [ruby-dev:28443] * configure.in: remove enable_rpath=no for Solaris. [ruby-dev:28440] * ext/win32ole/win32ole.c (ole_val2olevariantdata): change behavior of converting OLE Variant object with VT_ARRAY|VT_UI1 and Ruby String object. * ruby.1: a clarification patch from David Lutterkort <dlutter at redhat.com>. [ruby-core:7508] * lib/rdoc/ri/ri_paths.rb (RI::Paths): adding paths from rubygems directories. a patch from Eric Hodel <drbrain at segment7.net>. [ruby-core:07423] * eval.c (rb_clear_cache_by_class): clearing wrong cache. * ext/extmk.rb: use :remove_destination to install extension libraries to avoid SEGV. [ruby-dev:28417] * eval.c (rb_thread_fd_writable): should not re-schedule output from KILLED thread (must be error printing). * array.c (rb_ary_flatten_bang): allow specifying recursion level. [ruby-talk:182170] * array.c (rb_ary_flatten): ditto. * gc.c (add_heap): a heap_slots may overflow. a patch from Stefan Weil <weil at mail.berlios.de>. * eval.c (rb_call): use separate cache for fcall/vcall invocation. * eval.c (rb_eval): NODE_FCALL, NODE_VCALL can call local functions. * eval.c (rb_mod_local): a new method to specify newly added visibility "local". * eval.c (search_method): search for local methods which are visible only from the current class. * class.c (rb_class_local_methods): a method to list local methods. * object.c (Init_Object): add BasicObject class as a top level BlankSlate class. * ruby.h (SYM2ID): should not cast to signed long. [ruby-core:07414] * class.c (rb_include_module): allow module duplication. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@10235 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2006-06-10 01:20:17 +04:00
#ifndef HAVE_LOG2
double
log2(double x)
{
return log10(x)/log10(2.0);
}
#else
extern double log2(double);
#endif
* sprintf.c (rb_str_format): allow %c to print one character string (e.g. ?x). * lib/tempfile.rb (Tempfile::make_tmpname): put dot between basename and pid. [ruby-talk:196272] * parse.y (do_block): remove -> style block. * parse.y (parser_yylex): remove tLAMBDA_ARG. * eval.c (rb_call0): binding for the return event hook should have consistent scope. [ruby-core:07928] * eval.c (proc_invoke): return behavior should depend whether it is surrounded by a lambda or a mere block. * eval.c (formal_assign): handles post splat arguments. * eval.c (rb_call0): ditto. * st.c (strhash): use FNV-1a hash. * parse.y (parser_yylex): removed experimental ';;' terminator. * eval.c (rb_node_arity): should be aware of post splat arguments. * eval.c (rb_proc_arity): ditto. * parse.y (f_args): syntax rule enhanced to support arguments after the splat. * parse.y (block_param): ditto for block parameters. * parse.y (f_post_arg): mandatory formal arguments after the splat argument. * parse.y (new_args_gen): generate nodes for mandatory formal arguments after the splat argument. * eval.c (rb_eval): dispatch mandatory formal arguments after the splat argument. * parse.y (args): allow more than one splat in the argument list. * parse.y (method_call): allow aref [] to accept all kind of method argument, including assocs, splat, and block argument. * eval.c (SETUP_ARGS0): prepare block argument as well. * lib/mathn.rb (Integer): remove Integer#gcd2. [ruby-core:07931] * eval.c (error_line): print receivers true/false/nil specially. * eval.c (rb_proc_yield): handles parameters in yield semantics. * eval.c (nil_yield): gives LocalJumpError to denote no block error. * io.c (rb_io_getc): now takes one-character string. * string.c (rb_str_hash): use FNV-1a hash from Fowler/Noll/Vo hashing algorithm. * string.c (rb_str_aref): str[0] now returns 1 character string, instead of a fixnum. [Ruby2] * parse.y (parser_yylex): ?c now returns 1 character string, instead of a fixnum. [Ruby2] * string.c (rb_str_aset): no longer support fixnum insertion. * eval.c (umethod_bind): should not update original class. [ruby-dev:28636] * eval.c (ev_const_get): should support constant access from within instance_eval(). [ruby-dev:28327] * time.c (time_timeval): should round for usec floating number. [ruby-core:07896] * time.c (time_add): ditto. * dir.c (sys_warning): should not call a vararg function rb_sys_warning() indirectly. [ruby-core:07886] * numeric.c (flo_divmod): the first element of Float#divmod should be an integer. [ruby-dev:28589] * test/ruby/test_float.rb: add tests for divmod, div, modulo and remainder. * re.c (rb_reg_initialize): should not allow modifying literal regexps. frozen check moved from rb_reg_initialize_m as well. * re.c (rb_reg_initialize): should not modify untainted objects in safe levels higher than 3. * re.c (rb_memcmp): type change from char* to const void*. * dir.c (dir_close): should not close untainted dir stream. * dir.c (GetDIR): add tainted/frozen check for each dir operation. * lib/rdoc/parsers/parse_rb.rb (RDoc::RubyParser::parse_symbol_arg): typo fixed. a patch from Florian Gross <florg at florg.net>. * eval.c (EXEC_EVENT_HOOK): trace_func may remove itself from event_hooks. no guarantee for arbitrary hook deletion. [ruby-dev:28632] * util.c (ruby_strtod): differ addition to minimize error. [ruby-dev:28619] * util.c (ruby_strtod): should not raise ERANGE when the input string does not have any digits. [ruby-dev:28629] * eval.c (proc_invoke): should restore old ruby_frame->block. thanks to ts <decoux at moulon.inra.fr>. [ruby-core:07833] also fix [ruby-dev:28614] as well. * signal.c (trap): sig should be less then NSIG. Coverity found this bug. a patch from Kevin Tew <tewk at tewk.com>. [ruby-core:07823] * math.c (math_log2): add new method inspired by [ruby-talk:191237]. * math.c (math_log): add optional base argument to Math::log(). [ruby-talk:191308] * ext/syck/emitter.c (syck_scan_scalar): avoid accessing uninitialized array element. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07809] * array.c (rb_ary_fill): initialize local variables first. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07810] * ext/syck/yaml2byte.c (syck_yaml2byte_handler): need to free type_tag. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07808] * ext/socket/socket.c (make_hostent_internal): accept ai_family check from Sam Roberts <sroberts at uniserve.com>. [ruby-core:07691] * util.c (ruby_strtod): should not cut off 18 digits for no reason. [ruby-core:07796] * array.c (rb_ary_fill): internalize local variable "beg" to pacify Coverity. [ruby-core:07770] * pack.c (pack_unpack): now supports CRLF newlines. a patch from <tommy at tmtm.org>. [ruby-dev:28601] * applied code clean-up patch from Stefan Huehner <stefan at huehner.org>. [ruby-core:07764] * lib/jcode.rb (String::tr_s): should have translated non squeezing character sequence (i.e. a character) as well. thanks to Hiroshi Ichikawa <gimite at gimite.ddo.jp> [ruby-list:42090] * ext/socket/socket.c: document update patch from Sam Roberts <sroberts at uniserve.com>. [ruby-core:07701] * lib/mathn.rb (Integer): need not to remove gcd2. a patch from NARUSE, Yui <naruse at airemix.com>. [ruby-dev:28570] * parse.y (arg): too much NEW_LIST() * eval.c (SETUP_ARGS0): remove unnecessary access to nd_alen. * eval.c (rb_eval): use ARGSCAT for NODE_OP_ASGN1. [ruby-dev:28585] * parse.y (arg): use NODE_ARGSCAT for placeholder. * lib/getoptlong.rb (GetoptLong::get): RDoc update patch from mathew <meta at pobox.com>. [ruby-core:07738] * variable.c (rb_const_set): raise error when no target klass is supplied. [ruby-dev:28582] * prec.c (prec_prec_f): documentation patch from <gerardo.santana at gmail.com>. [ruby-core:07689] * bignum.c (rb_big_pow): second operand may be too big even if it's a Fixnum. [ruby-talk:187984] * README.EXT: update symbol description. [ruby-talk:188104] * COPYING: explicitly note GPLv2. [ruby-talk:187922] * parse.y: remove some obsolete syntax rules (unparenthesized method calls in argument list). * eval.c (rb_call0): insecure calling should be checked for non NODE_SCOPE method invocations too. * eval.c (rb_alias): should preserve the current safe level as well as method definition. * process.c (rb_f_sleep): remove RDoc description about SIGALRM which is not valid on the current implementation. [ruby-dev:28464] Thu Mar 23 21:40:47 2006 K.Kosako <sndgk393 AT ybb.ne.jp> * eval.c (method_missing): should support argument splat in super. a bug in combination of super, splat and method_missing. [ruby-talk:185438] * configure.in: Solaris SunPro compiler -rapth patch from <kuwa at labs.fujitsu.com>. [ruby-dev:28443] * configure.in: remove enable_rpath=no for Solaris. [ruby-dev:28440] * ext/win32ole/win32ole.c (ole_val2olevariantdata): change behavior of converting OLE Variant object with VT_ARRAY|VT_UI1 and Ruby String object. * ruby.1: a clarification patch from David Lutterkort <dlutter at redhat.com>. [ruby-core:7508] * lib/rdoc/ri/ri_paths.rb (RI::Paths): adding paths from rubygems directories. a patch from Eric Hodel <drbrain at segment7.net>. [ruby-core:07423] * eval.c (rb_clear_cache_by_class): clearing wrong cache. * ext/extmk.rb: use :remove_destination to install extension libraries to avoid SEGV. [ruby-dev:28417] * eval.c (rb_thread_fd_writable): should not re-schedule output from KILLED thread (must be error printing). * array.c (rb_ary_flatten_bang): allow specifying recursion level. [ruby-talk:182170] * array.c (rb_ary_flatten): ditto. * gc.c (add_heap): a heap_slots may overflow. a patch from Stefan Weil <weil at mail.berlios.de>. * eval.c (rb_call): use separate cache for fcall/vcall invocation. * eval.c (rb_eval): NODE_FCALL, NODE_VCALL can call local functions. * eval.c (rb_mod_local): a new method to specify newly added visibility "local". * eval.c (search_method): search for local methods which are visible only from the current class. * class.c (rb_class_local_methods): a method to list local methods. * object.c (Init_Object): add BasicObject class as a top level BlankSlate class. * ruby.h (SYM2ID): should not cast to signed long. [ruby-core:07414] * class.c (rb_include_module): allow module duplication. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@10235 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2006-06-10 01:20:17 +04:00
#endif
/*
* call-seq:
* Math.log2(x) -> float
*
* Returns the base 2 {logarithm}[https://en.wikipedia.org/wiki/Logarithm] of +x+.
*
* - Domain: <tt>[0, INFINITY]</tt>.
* - Range: <tt>[-INFINITY, INFINITY]</tt>.
*
* Examples:
*
* log2(0.0) # => -Infinity
* log2(1.0) # => 0.0
* log2(2.0) # => 1.0
* log2(INFINITY) # => Infinity
*
* sprintf.c (rb_str_format): allow %c to print one character string (e.g. ?x). * lib/tempfile.rb (Tempfile::make_tmpname): put dot between basename and pid. [ruby-talk:196272] * parse.y (do_block): remove -> style block. * parse.y (parser_yylex): remove tLAMBDA_ARG. * eval.c (rb_call0): binding for the return event hook should have consistent scope. [ruby-core:07928] * eval.c (proc_invoke): return behavior should depend whether it is surrounded by a lambda or a mere block. * eval.c (formal_assign): handles post splat arguments. * eval.c (rb_call0): ditto. * st.c (strhash): use FNV-1a hash. * parse.y (parser_yylex): removed experimental ';;' terminator. * eval.c (rb_node_arity): should be aware of post splat arguments. * eval.c (rb_proc_arity): ditto. * parse.y (f_args): syntax rule enhanced to support arguments after the splat. * parse.y (block_param): ditto for block parameters. * parse.y (f_post_arg): mandatory formal arguments after the splat argument. * parse.y (new_args_gen): generate nodes for mandatory formal arguments after the splat argument. * eval.c (rb_eval): dispatch mandatory formal arguments after the splat argument. * parse.y (args): allow more than one splat in the argument list. * parse.y (method_call): allow aref [] to accept all kind of method argument, including assocs, splat, and block argument. * eval.c (SETUP_ARGS0): prepare block argument as well. * lib/mathn.rb (Integer): remove Integer#gcd2. [ruby-core:07931] * eval.c (error_line): print receivers true/false/nil specially. * eval.c (rb_proc_yield): handles parameters in yield semantics. * eval.c (nil_yield): gives LocalJumpError to denote no block error. * io.c (rb_io_getc): now takes one-character string. * string.c (rb_str_hash): use FNV-1a hash from Fowler/Noll/Vo hashing algorithm. * string.c (rb_str_aref): str[0] now returns 1 character string, instead of a fixnum. [Ruby2] * parse.y (parser_yylex): ?c now returns 1 character string, instead of a fixnum. [Ruby2] * string.c (rb_str_aset): no longer support fixnum insertion. * eval.c (umethod_bind): should not update original class. [ruby-dev:28636] * eval.c (ev_const_get): should support constant access from within instance_eval(). [ruby-dev:28327] * time.c (time_timeval): should round for usec floating number. [ruby-core:07896] * time.c (time_add): ditto. * dir.c (sys_warning): should not call a vararg function rb_sys_warning() indirectly. [ruby-core:07886] * numeric.c (flo_divmod): the first element of Float#divmod should be an integer. [ruby-dev:28589] * test/ruby/test_float.rb: add tests for divmod, div, modulo and remainder. * re.c (rb_reg_initialize): should not allow modifying literal regexps. frozen check moved from rb_reg_initialize_m as well. * re.c (rb_reg_initialize): should not modify untainted objects in safe levels higher than 3. * re.c (rb_memcmp): type change from char* to const void*. * dir.c (dir_close): should not close untainted dir stream. * dir.c (GetDIR): add tainted/frozen check for each dir operation. * lib/rdoc/parsers/parse_rb.rb (RDoc::RubyParser::parse_symbol_arg): typo fixed. a patch from Florian Gross <florg at florg.net>. * eval.c (EXEC_EVENT_HOOK): trace_func may remove itself from event_hooks. no guarantee for arbitrary hook deletion. [ruby-dev:28632] * util.c (ruby_strtod): differ addition to minimize error. [ruby-dev:28619] * util.c (ruby_strtod): should not raise ERANGE when the input string does not have any digits. [ruby-dev:28629] * eval.c (proc_invoke): should restore old ruby_frame->block. thanks to ts <decoux at moulon.inra.fr>. [ruby-core:07833] also fix [ruby-dev:28614] as well. * signal.c (trap): sig should be less then NSIG. Coverity found this bug. a patch from Kevin Tew <tewk at tewk.com>. [ruby-core:07823] * math.c (math_log2): add new method inspired by [ruby-talk:191237]. * math.c (math_log): add optional base argument to Math::log(). [ruby-talk:191308] * ext/syck/emitter.c (syck_scan_scalar): avoid accessing uninitialized array element. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07809] * array.c (rb_ary_fill): initialize local variables first. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07810] * ext/syck/yaml2byte.c (syck_yaml2byte_handler): need to free type_tag. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07808] * ext/socket/socket.c (make_hostent_internal): accept ai_family check from Sam Roberts <sroberts at uniserve.com>. [ruby-core:07691] * util.c (ruby_strtod): should not cut off 18 digits for no reason. [ruby-core:07796] * array.c (rb_ary_fill): internalize local variable "beg" to pacify Coverity. [ruby-core:07770] * pack.c (pack_unpack): now supports CRLF newlines. a patch from <tommy at tmtm.org>. [ruby-dev:28601] * applied code clean-up patch from Stefan Huehner <stefan at huehner.org>. [ruby-core:07764] * lib/jcode.rb (String::tr_s): should have translated non squeezing character sequence (i.e. a character) as well. thanks to Hiroshi Ichikawa <gimite at gimite.ddo.jp> [ruby-list:42090] * ext/socket/socket.c: document update patch from Sam Roberts <sroberts at uniserve.com>. [ruby-core:07701] * lib/mathn.rb (Integer): need not to remove gcd2. a patch from NARUSE, Yui <naruse at airemix.com>. [ruby-dev:28570] * parse.y (arg): too much NEW_LIST() * eval.c (SETUP_ARGS0): remove unnecessary access to nd_alen. * eval.c (rb_eval): use ARGSCAT for NODE_OP_ASGN1. [ruby-dev:28585] * parse.y (arg): use NODE_ARGSCAT for placeholder. * lib/getoptlong.rb (GetoptLong::get): RDoc update patch from mathew <meta at pobox.com>. [ruby-core:07738] * variable.c (rb_const_set): raise error when no target klass is supplied. [ruby-dev:28582] * prec.c (prec_prec_f): documentation patch from <gerardo.santana at gmail.com>. [ruby-core:07689] * bignum.c (rb_big_pow): second operand may be too big even if it's a Fixnum. [ruby-talk:187984] * README.EXT: update symbol description. [ruby-talk:188104] * COPYING: explicitly note GPLv2. [ruby-talk:187922] * parse.y: remove some obsolete syntax rules (unparenthesized method calls in argument list). * eval.c (rb_call0): insecure calling should be checked for non NODE_SCOPE method invocations too. * eval.c (rb_alias): should preserve the current safe level as well as method definition. * process.c (rb_f_sleep): remove RDoc description about SIGALRM which is not valid on the current implementation. [ruby-dev:28464] Thu Mar 23 21:40:47 2006 K.Kosako <sndgk393 AT ybb.ne.jp> * eval.c (method_missing): should support argument splat in super. a bug in combination of super, splat and method_missing. [ruby-talk:185438] * configure.in: Solaris SunPro compiler -rapth patch from <kuwa at labs.fujitsu.com>. [ruby-dev:28443] * configure.in: remove enable_rpath=no for Solaris. [ruby-dev:28440] * ext/win32ole/win32ole.c (ole_val2olevariantdata): change behavior of converting OLE Variant object with VT_ARRAY|VT_UI1 and Ruby String object. * ruby.1: a clarification patch from David Lutterkort <dlutter at redhat.com>. [ruby-core:7508] * lib/rdoc/ri/ri_paths.rb (RI::Paths): adding paths from rubygems directories. a patch from Eric Hodel <drbrain at segment7.net>. [ruby-core:07423] * eval.c (rb_clear_cache_by_class): clearing wrong cache. * ext/extmk.rb: use :remove_destination to install extension libraries to avoid SEGV. [ruby-dev:28417] * eval.c (rb_thread_fd_writable): should not re-schedule output from KILLED thread (must be error printing). * array.c (rb_ary_flatten_bang): allow specifying recursion level. [ruby-talk:182170] * array.c (rb_ary_flatten): ditto. * gc.c (add_heap): a heap_slots may overflow. a patch from Stefan Weil <weil at mail.berlios.de>. * eval.c (rb_call): use separate cache for fcall/vcall invocation. * eval.c (rb_eval): NODE_FCALL, NODE_VCALL can call local functions. * eval.c (rb_mod_local): a new method to specify newly added visibility "local". * eval.c (search_method): search for local methods which are visible only from the current class. * class.c (rb_class_local_methods): a method to list local methods. * object.c (Init_Object): add BasicObject class as a top level BlankSlate class. * ruby.h (SYM2ID): should not cast to signed long. [ruby-core:07414] * class.c (rb_include_module): allow module duplication. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@10235 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2006-06-10 01:20:17 +04:00
*/
static VALUE
math_log2(VALUE unused_obj, VALUE x)
* sprintf.c (rb_str_format): allow %c to print one character string (e.g. ?x). * lib/tempfile.rb (Tempfile::make_tmpname): put dot between basename and pid. [ruby-talk:196272] * parse.y (do_block): remove -> style block. * parse.y (parser_yylex): remove tLAMBDA_ARG. * eval.c (rb_call0): binding for the return event hook should have consistent scope. [ruby-core:07928] * eval.c (proc_invoke): return behavior should depend whether it is surrounded by a lambda or a mere block. * eval.c (formal_assign): handles post splat arguments. * eval.c (rb_call0): ditto. * st.c (strhash): use FNV-1a hash. * parse.y (parser_yylex): removed experimental ';;' terminator. * eval.c (rb_node_arity): should be aware of post splat arguments. * eval.c (rb_proc_arity): ditto. * parse.y (f_args): syntax rule enhanced to support arguments after the splat. * parse.y (block_param): ditto for block parameters. * parse.y (f_post_arg): mandatory formal arguments after the splat argument. * parse.y (new_args_gen): generate nodes for mandatory formal arguments after the splat argument. * eval.c (rb_eval): dispatch mandatory formal arguments after the splat argument. * parse.y (args): allow more than one splat in the argument list. * parse.y (method_call): allow aref [] to accept all kind of method argument, including assocs, splat, and block argument. * eval.c (SETUP_ARGS0): prepare block argument as well. * lib/mathn.rb (Integer): remove Integer#gcd2. [ruby-core:07931] * eval.c (error_line): print receivers true/false/nil specially. * eval.c (rb_proc_yield): handles parameters in yield semantics. * eval.c (nil_yield): gives LocalJumpError to denote no block error. * io.c (rb_io_getc): now takes one-character string. * string.c (rb_str_hash): use FNV-1a hash from Fowler/Noll/Vo hashing algorithm. * string.c (rb_str_aref): str[0] now returns 1 character string, instead of a fixnum. [Ruby2] * parse.y (parser_yylex): ?c now returns 1 character string, instead of a fixnum. [Ruby2] * string.c (rb_str_aset): no longer support fixnum insertion. * eval.c (umethod_bind): should not update original class. [ruby-dev:28636] * eval.c (ev_const_get): should support constant access from within instance_eval(). [ruby-dev:28327] * time.c (time_timeval): should round for usec floating number. [ruby-core:07896] * time.c (time_add): ditto. * dir.c (sys_warning): should not call a vararg function rb_sys_warning() indirectly. [ruby-core:07886] * numeric.c (flo_divmod): the first element of Float#divmod should be an integer. [ruby-dev:28589] * test/ruby/test_float.rb: add tests for divmod, div, modulo and remainder. * re.c (rb_reg_initialize): should not allow modifying literal regexps. frozen check moved from rb_reg_initialize_m as well. * re.c (rb_reg_initialize): should not modify untainted objects in safe levels higher than 3. * re.c (rb_memcmp): type change from char* to const void*. * dir.c (dir_close): should not close untainted dir stream. * dir.c (GetDIR): add tainted/frozen check for each dir operation. * lib/rdoc/parsers/parse_rb.rb (RDoc::RubyParser::parse_symbol_arg): typo fixed. a patch from Florian Gross <florg at florg.net>. * eval.c (EXEC_EVENT_HOOK): trace_func may remove itself from event_hooks. no guarantee for arbitrary hook deletion. [ruby-dev:28632] * util.c (ruby_strtod): differ addition to minimize error. [ruby-dev:28619] * util.c (ruby_strtod): should not raise ERANGE when the input string does not have any digits. [ruby-dev:28629] * eval.c (proc_invoke): should restore old ruby_frame->block. thanks to ts <decoux at moulon.inra.fr>. [ruby-core:07833] also fix [ruby-dev:28614] as well. * signal.c (trap): sig should be less then NSIG. Coverity found this bug. a patch from Kevin Tew <tewk at tewk.com>. [ruby-core:07823] * math.c (math_log2): add new method inspired by [ruby-talk:191237]. * math.c (math_log): add optional base argument to Math::log(). [ruby-talk:191308] * ext/syck/emitter.c (syck_scan_scalar): avoid accessing uninitialized array element. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07809] * array.c (rb_ary_fill): initialize local variables first. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07810] * ext/syck/yaml2byte.c (syck_yaml2byte_handler): need to free type_tag. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07808] * ext/socket/socket.c (make_hostent_internal): accept ai_family check from Sam Roberts <sroberts at uniserve.com>. [ruby-core:07691] * util.c (ruby_strtod): should not cut off 18 digits for no reason. [ruby-core:07796] * array.c (rb_ary_fill): internalize local variable "beg" to pacify Coverity. [ruby-core:07770] * pack.c (pack_unpack): now supports CRLF newlines. a patch from <tommy at tmtm.org>. [ruby-dev:28601] * applied code clean-up patch from Stefan Huehner <stefan at huehner.org>. [ruby-core:07764] * lib/jcode.rb (String::tr_s): should have translated non squeezing character sequence (i.e. a character) as well. thanks to Hiroshi Ichikawa <gimite at gimite.ddo.jp> [ruby-list:42090] * ext/socket/socket.c: document update patch from Sam Roberts <sroberts at uniserve.com>. [ruby-core:07701] * lib/mathn.rb (Integer): need not to remove gcd2. a patch from NARUSE, Yui <naruse at airemix.com>. [ruby-dev:28570] * parse.y (arg): too much NEW_LIST() * eval.c (SETUP_ARGS0): remove unnecessary access to nd_alen. * eval.c (rb_eval): use ARGSCAT for NODE_OP_ASGN1. [ruby-dev:28585] * parse.y (arg): use NODE_ARGSCAT for placeholder. * lib/getoptlong.rb (GetoptLong::get): RDoc update patch from mathew <meta at pobox.com>. [ruby-core:07738] * variable.c (rb_const_set): raise error when no target klass is supplied. [ruby-dev:28582] * prec.c (prec_prec_f): documentation patch from <gerardo.santana at gmail.com>. [ruby-core:07689] * bignum.c (rb_big_pow): second operand may be too big even if it's a Fixnum. [ruby-talk:187984] * README.EXT: update symbol description. [ruby-talk:188104] * COPYING: explicitly note GPLv2. [ruby-talk:187922] * parse.y: remove some obsolete syntax rules (unparenthesized method calls in argument list). * eval.c (rb_call0): insecure calling should be checked for non NODE_SCOPE method invocations too. * eval.c (rb_alias): should preserve the current safe level as well as method definition. * process.c (rb_f_sleep): remove RDoc description about SIGALRM which is not valid on the current implementation. [ruby-dev:28464] Thu Mar 23 21:40:47 2006 K.Kosako <sndgk393 AT ybb.ne.jp> * eval.c (method_missing): should support argument splat in super. a bug in combination of super, splat and method_missing. [ruby-talk:185438] * configure.in: Solaris SunPro compiler -rapth patch from <kuwa at labs.fujitsu.com>. [ruby-dev:28443] * configure.in: remove enable_rpath=no for Solaris. [ruby-dev:28440] * ext/win32ole/win32ole.c (ole_val2olevariantdata): change behavior of converting OLE Variant object with VT_ARRAY|VT_UI1 and Ruby String object. * ruby.1: a clarification patch from David Lutterkort <dlutter at redhat.com>. [ruby-core:7508] * lib/rdoc/ri/ri_paths.rb (RI::Paths): adding paths from rubygems directories. a patch from Eric Hodel <drbrain at segment7.net>. [ruby-core:07423] * eval.c (rb_clear_cache_by_class): clearing wrong cache. * ext/extmk.rb: use :remove_destination to install extension libraries to avoid SEGV. [ruby-dev:28417] * eval.c (rb_thread_fd_writable): should not re-schedule output from KILLED thread (must be error printing). * array.c (rb_ary_flatten_bang): allow specifying recursion level. [ruby-talk:182170] * array.c (rb_ary_flatten): ditto. * gc.c (add_heap): a heap_slots may overflow. a patch from Stefan Weil <weil at mail.berlios.de>. * eval.c (rb_call): use separate cache for fcall/vcall invocation. * eval.c (rb_eval): NODE_FCALL, NODE_VCALL can call local functions. * eval.c (rb_mod_local): a new method to specify newly added visibility "local". * eval.c (search_method): search for local methods which are visible only from the current class. * class.c (rb_class_local_methods): a method to list local methods. * object.c (Init_Object): add BasicObject class as a top level BlankSlate class. * ruby.h (SYM2ID): should not cast to signed long. [ruby-core:07414] * class.c (rb_include_module): allow module duplication. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@10235 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2006-06-10 01:20:17 +04:00
{
size_t numbits;
double d = get_double_rshift(x, &numbits);
2021-07-04 13:22:11 +03:00
domain_check_min(d, 0.0, "log2");
/* check for pole error */
if (d == 0.0) return DBL2NUM(-HUGE_VAL);
return DBL2NUM(log2(d) + numbits); /* log2(d * 2 ** numbits) */
}
/*
* call-seq:
* Math.log10(x) -> float
*
* Returns the base 10 {logarithm}[https://en.wikipedia.org/wiki/Logarithm] of +x+.
*
* - Domain: <tt>[0, INFINITY]</tt>.
* - Range: <tt>[-INFINITY, INFINITY]</tt>.
*
* Examples:
*
* log10(0.0) # => -Infinity
* log10(1.0) # => 0.0
* log10(10.0) # => 1.0
* log10(INFINITY) # => Infinity
*
*/
static VALUE
math_log10(VALUE unused_obj, VALUE x)
{
size_t numbits;
double d = get_double_rshift(x, &numbits);
2021-07-04 13:22:11 +03:00
domain_check_min(d, 0.0, "log10");
/* check for pole error */
if (d == 0.0) return DBL2NUM(-HUGE_VAL);
return DBL2NUM(log10(d) + numbits * log10(2)); /* log10(d * 2 ** numbits) */
}
static VALUE rb_math_sqrt(VALUE x);
/*
* call-seq:
* Math.sqrt(x) -> float
*
* Returns the principal (non-negative) {square root}[https://en.wikipedia.org/wiki/Square_root] of +x+.
*
* - Domain: <tt>[0, INFINITY]</tt>.
* - Range: <tt>[0, INFINITY]</tt>.
*
* Examples:
*
* sqrt(0.0) # => 0.0
* sqrt(0.5) # => 0.7071067811865476
* sqrt(1.0) # => 1.0
* sqrt(2.0) # => 1.4142135623730951
* sqrt(4.0) # => 2.0
* sqrt(9.0) # => 3.0
* sqrt(INFINITY) # => Infinity
*
*/
static VALUE
math_sqrt(VALUE unused_obj, VALUE x)
{
return rb_math_sqrt(x);
}
inline static VALUE
f_negative_p(VALUE x)
{
if (FIXNUM_P(x))
return RBOOL(FIX2LONG(x) < 0);
return rb_funcall(x, '<', 1, INT2FIX(0));
}
inline static VALUE
f_signbit(VALUE x)
{
2021-09-11 03:56:59 +03:00
if (RB_FLOAT_TYPE_P(x)) {
double f = RFLOAT_VALUE(x);
return RBOOL(!isnan(f) && signbit(f));
}
return f_negative_p(x);
}
static VALUE
rb_math_sqrt(VALUE x)
{
double d;
if (RB_TYPE_P(x, T_COMPLEX)) {
VALUE neg = f_signbit(RCOMPLEX(x)->imag);
double re = Get_Double(RCOMPLEX(x)->real), im;
d = Get_Double(rb_complex_abs(x));
im = sqrt((d - re) / 2.0);
re = sqrt((d + re) / 2.0);
if (neg) im = -im;
return rb_complex_new(DBL2NUM(re), DBL2NUM(im));
}
d = Get_Double(x);
2021-07-04 13:22:11 +03:00
domain_check_min(d, 0.0, "sqrt");
if (d == 0.0) return DBL2NUM(0.0);
return DBL2NUM(sqrt(d));
}
/*
* call-seq:
* Math.cbrt(x) -> float
*
* Returns the {cube root}[https://en.wikipedia.org/wiki/Cube_root] of +x+.
*
* - Domain: <tt>[-INFINITY, INFINITY]</tt>.
* - Range: <tt>[-INFINITY, INFINITY]</tt>.
*
* Examples:
*
* cbrt(-INFINITY) # => -Infinity
* cbrt(-27.0) # => -3.0
* cbrt(-8.0) # => -2.0
* cbrt(-2.0) # => -1.2599210498948732
* cbrt(1.0) # => 1.0
* cbrt(0.0) # => 0.0
* cbrt(1.0) # => 1.0
2023-09-27 10:18:05 +03:00
* cbrt(2.0) # => 1.2599210498948732
* cbrt(8.0) # => 2.0
* cbrt(27.0) # => 3.0
* cbrt(INFINITY) # => Infinity
*
*/
static VALUE
math_cbrt(VALUE unused_obj, VALUE x)
{
double f = Get_Double(x);
double r = cbrt(f);
#if defined __GLIBC__
if (isfinite(r) && !(f == 0.0 && r == 0.0)) {
r = (2.0 * r + (f / r / r)) / 3.0;
}
#endif
return DBL2NUM(r);
}
/*
* call-seq:
* Math.frexp(x) -> [fraction, exponent]
*
* Returns a 2-element array containing the normalized signed float +fraction+
* and integer +exponent+ of +x+ such that:
*
* x = fraction * 2**exponent
*
* See {IEEE 754 double-precision binary floating-point format: binary64}[https://en.wikipedia.org/wiki/Double-precision_floating-point_format#IEEE_754_double-precision_binary_floating-point_format:_binary64].
*
* - Domain: <tt>[-INFINITY, INFINITY]</tt>.
* - Range <tt>[-INFINITY, INFINITY]</tt>.
*
* Examples:
*
* frexp(-INFINITY) # => [-Infinity, -1]
* frexp(-2.0) # => [-0.5, 2]
* frexp(-1.0) # => [-0.5, 1]
* frexp(0.0) # => [0.0, 0]
* frexp(1.0) # => [0.5, 1]
* frexp(2.0) # => [0.5, 2]
* frexp(INFINITY) # => [Infinity, -1]
*
* Related: Math.ldexp (inverse of Math.frexp).
*
*/
static VALUE
math_frexp(VALUE unused_obj, VALUE x)
{
double d;
int exp;
d = frexp(Get_Double(x), &exp);
return rb_assoc_new(DBL2NUM(d), INT2NUM(exp));
}
/*
* call-seq:
* Math.ldexp(fraction, exponent) -> float
*
* Returns the value of <tt>fraction * 2**exponent</tt>.
*
* - Domain of +fraction+: <tt>[0.0, 1.0)</tt>.
* - Domain of +exponent+: <tt>[0, 1024]</tt>
* (larger values are equivalent to 1024).
*
* See {IEEE 754 double-precision binary floating-point format: binary64}[https://en.wikipedia.org/wiki/Double-precision_floating-point_format#IEEE_754_double-precision_binary_floating-point_format:_binary64].
*
* Examples:
*
* ldexp(-INFINITY, -1) # => -Infinity
* ldexp(-0.5, 2) # => -2.0
* ldexp(-0.5, 1) # => -1.0
* ldexp(0.0, 0) # => 0.0
* ldexp(-0.5, 1) # => 1.0
* ldexp(-0.5, 2) # => 2.0
* ldexp(INFINITY, -1) # => Infinity
*
* Related: Math.frexp (inverse of Math.ldexp).
*
*/
static VALUE
math_ldexp(VALUE unused_obj, VALUE x, VALUE n)
{
return DBL2NUM(ldexp(Get_Double(x), NUM2INT(n)));
}
/*
* call-seq:
* Math.hypot(a, b) -> float
*
* Returns <tt>sqrt(a**2 + b**2)</tt>,
* which is the length of the longest side +c+ (the hypotenuse)
* of the right triangle whose other sides have lengths +a+ and +b+.
*
* - Domain of +a+: <tt>[-INFINITY, INFINITY]</tt>.
* - Domain of +ab: <tt>[-INFINITY, INFINITY]</tt>.
* - Range: <tt>[0, INFINITY]</tt>.
*
* Examples:
*
* hypot(0.0, 1.0) # => 1.0
* hypot(1.0, 1.0) # => 1.4142135623730951 # sqrt(2.0)
* hypot(3.0, 4.0) # => 5.0
* hypot(5.0, 12.0) # => 13.0
* hypot(1.0, sqrt(3.0)) # => 1.9999999999999998 # Near 2.0
*
* Note that if either argument is +INFINITY+ or <tt>-INFINITY</tt>,
* the result is +Infinity+.
*
*/
static VALUE
math_hypot(VALUE unused_obj, VALUE x, VALUE y)
{
return DBL2NUM(hypot(Get_Double(x), Get_Double(y)));
}
/*
* call-seq:
* Math.erf(x) -> float
*
* Returns the value of the {Gauss error function}[https://en.wikipedia.org/wiki/Error_function] for +x+.
*
* - Domain: <tt>[-INFINITY, INFINITY]</tt>.
* - Range: <tt>[-1, 1]</tt>.
*
* Examples:
*
* erf(-INFINITY) # => -1.0
* erf(0.0) # => 0.0
* erf(INFINITY) # => 1.0
*
* Related: Math.erfc.
*
*/
static VALUE
math_erf(VALUE unused_obj, VALUE x)
{
return DBL2NUM(erf(Get_Double(x)));
}
/*
* call-seq:
* Math.erfc(x) -> Float
*
* Returns the value of the {complementary error function}[https://en.wikipedia.org/wiki/Error_function#Complementary_error_function] for +x+.
*
* - Domain: <tt>[-INFINITY, INFINITY]</tt>.
* - Range: <tt>[0, 2]</tt>.
*
* Examples:
*
* erfc(-INFINITY) # => 2.0
* erfc(0.0) # => 1.0
* erfc(INFINITY) # => 0.0
*
* Related: Math.erf.
*
*/
static VALUE
math_erfc(VALUE unused_obj, VALUE x)
{
return DBL2NUM(erfc(Get_Double(x)));
}
/*
* call-seq:
* Math.gamma(x) -> float
*
* Returns the value of the {gamma function}[https://en.wikipedia.org/wiki/Gamma_function] for +x+.
*
* - Domain: <tt>(-INFINITY, INFINITY]</tt> excluding negative integers.
* - Range: <tt>[-INFINITY, INFINITY]</tt>.
*
* Examples:
*
* gamma(-2.5) # => -0.9453087204829431
* gamma(-1.5) # => 2.3632718012073513
* gamma(-0.5) # => -3.5449077018110375
* gamma(0.0) # => Infinity
* gamma(1.0) # => 1.0
* gamma(2.0) # => 1.0
* gamma(3.0) # => 2.0
* gamma(4.0) # => 6.0
* gamma(5.0) # => 24.0
*
* Related: Math.lgamma.
*
*/
static VALUE
math_gamma(VALUE unused_obj, VALUE x)
{
static const double fact_table[] = {
/* fact(0) */ 1.0,
/* fact(1) */ 1.0,
/* fact(2) */ 2.0,
/* fact(3) */ 6.0,
/* fact(4) */ 24.0,
/* fact(5) */ 120.0,
/* fact(6) */ 720.0,
/* fact(7) */ 5040.0,
/* fact(8) */ 40320.0,
/* fact(9) */ 362880.0,
/* fact(10) */ 3628800.0,
/* fact(11) */ 39916800.0,
/* fact(12) */ 479001600.0,
/* fact(13) */ 6227020800.0,
/* fact(14) */ 87178291200.0,
/* fact(15) */ 1307674368000.0,
/* fact(16) */ 20922789888000.0,
/* fact(17) */ 355687428096000.0,
/* fact(18) */ 6402373705728000.0,
/* fact(19) */ 121645100408832000.0,
/* fact(20) */ 2432902008176640000.0,
/* fact(21) */ 51090942171709440000.0,
/* fact(22) */ 1124000727777607680000.0,
/* fact(23)=25852016738884976640000 needs 56bit mantissa which is
* impossible to represent exactly in IEEE 754 double which have
* 53bit mantissa. */
};
enum {NFACT_TABLE = numberof(fact_table)};
double d;
d = Get_Double(x);
/* check for domain error */
if (isinf(d)) {
if (signbit(d)) domain_error("gamma");
return DBL2NUM(HUGE_VAL);
}
if (d == 0.0) {
return signbit(d) ? DBL2NUM(-HUGE_VAL) : DBL2NUM(HUGE_VAL);
}
if (d == floor(d)) {
2021-07-04 13:22:11 +03:00
domain_check_min(d, 0.0, "gamma");
if (1.0 <= d && d <= (double)NFACT_TABLE) {
return DBL2NUM(fact_table[(int)d - 1]);
}
}
return DBL2NUM(tgamma(d));
}
/*
* call-seq:
* Math.lgamma(x) -> [float, -1 or 1]
*
* Returns a 2-element array equivalent to:
*
* [Math.log(Math.gamma(x).abs), Math.gamma(x) < 0 ? -1 : 1]
*
* See {logarithmic gamma function}[https://en.wikipedia.org/wiki/Gamma_function#The_log-gamma_function].
*
* - Domain: <tt>(-INFINITY, INFINITY]</tt>.
* - Range of first element: <tt>(-INFINITY, INFINITY]</tt>.
* - Second element is -1 or 1.
*
* Examples:
*
* lgamma(-4.0) # => [Infinity, -1]
* lgamma(-3.0) # => [Infinity, -1]
* lgamma(-2.0) # => [Infinity, -1]
* lgamma(-1.0) # => [Infinity, -1]
* lgamma(0.0) # => [Infinity, 1]
*
* lgamma(1.0) # => [0.0, 1]
* lgamma(2.0) # => [0.0, 1]
* lgamma(3.0) # => [0.6931471805599436, 1]
* lgamma(4.0) # => [1.7917594692280545, 1]
*
* lgamma(-2.5) # => [-0.05624371649767279, -1]
* lgamma(-1.5) # => [0.8600470153764797, 1]
* lgamma(-0.5) # => [1.265512123484647, -1]
* lgamma(0.5) # => [0.5723649429247004, 1]
* lgamma(1.5) # => [-0.12078223763524676, 1]
* lgamma(2.5) # => [0.2846828704729205, 1]
*
* Related: Math.gamma.
*
*/
static VALUE
math_lgamma(VALUE unused_obj, VALUE x)
{
double d;
int sign=1;
VALUE v;
d = Get_Double(x);
/* check for domain error */
if (isinf(d)) {
if (signbit(d)) domain_error("lgamma");
return rb_assoc_new(DBL2NUM(HUGE_VAL), INT2FIX(1));
}
if (d == 0.0) {
VALUE vsign = signbit(d) ? INT2FIX(-1) : INT2FIX(+1);
return rb_assoc_new(DBL2NUM(HUGE_VAL), vsign);
}
v = DBL2NUM(lgamma_r(d, &sign));
return rb_assoc_new(v, INT2FIX(sign));
}
#define exp1(n) \
VALUE \
rb_math_##n(VALUE x)\
{\
return math_##n(0, x);\
}
#define exp2(n) \
VALUE \
rb_math_##n(VALUE x, VALUE y)\
{\
return math_##n(0, x, y);\
}
exp2(atan2)
exp1(cos)
exp1(cosh)
exp1(exp)
exp2(hypot)
exp1(sin)
exp1(sinh)
#if 0
exp1(sqrt)
#endif
/*
* Document-class: Math::DomainError
*
* Raised when a mathematical function is evaluated outside of its
* domain of definition.
*
* For example, since +cos+ returns values in the range -1..1,
* its inverse function +acos+ is only defined on that interval:
*
* Math.acos(42)
*
* <em>produces:</em>
*
* Math::DomainError: Numerical argument is out of domain - "acos"
*/
/*
* Document-class: Math
*
* :include: doc/math/math.rdoc
*
*/
void
InitVM_Math(void)
{
rb_mMath = rb_define_module("Math");
rb_eMathDomainError = rb_define_class_under(rb_mMath, "DomainError", rb_eStandardError);
/* Definition of the mathematical constant PI as a Float number. */
rb_define_const(rb_mMath, "PI", DBL2NUM(M_PI));
#ifdef M_E
/* Definition of the mathematical constant E for Euler's number (e) as a Float number. */
rb_define_const(rb_mMath, "E", DBL2NUM(M_E));
#else
rb_define_const(rb_mMath, "E", DBL2NUM(exp(1.0)));
#endif
rb_define_module_function(rb_mMath, "atan2", math_atan2, 2);
rb_define_module_function(rb_mMath, "cos", math_cos, 1);
rb_define_module_function(rb_mMath, "sin", math_sin, 1);
rb_define_module_function(rb_mMath, "tan", math_tan, 1);
rb_define_module_function(rb_mMath, "acos", math_acos, 1);
rb_define_module_function(rb_mMath, "asin", math_asin, 1);
rb_define_module_function(rb_mMath, "atan", math_atan, 1);
rb_define_module_function(rb_mMath, "cosh", math_cosh, 1);
rb_define_module_function(rb_mMath, "sinh", math_sinh, 1);
rb_define_module_function(rb_mMath, "tanh", math_tanh, 1);
rb_define_module_function(rb_mMath, "acosh", math_acosh, 1);
rb_define_module_function(rb_mMath, "asinh", math_asinh, 1);
rb_define_module_function(rb_mMath, "atanh", math_atanh, 1);
rb_define_module_function(rb_mMath, "exp", math_exp, 1);
* sprintf.c (rb_str_format): allow %c to print one character string (e.g. ?x). * lib/tempfile.rb (Tempfile::make_tmpname): put dot between basename and pid. [ruby-talk:196272] * parse.y (do_block): remove -> style block. * parse.y (parser_yylex): remove tLAMBDA_ARG. * eval.c (rb_call0): binding for the return event hook should have consistent scope. [ruby-core:07928] * eval.c (proc_invoke): return behavior should depend whether it is surrounded by a lambda or a mere block. * eval.c (formal_assign): handles post splat arguments. * eval.c (rb_call0): ditto. * st.c (strhash): use FNV-1a hash. * parse.y (parser_yylex): removed experimental ';;' terminator. * eval.c (rb_node_arity): should be aware of post splat arguments. * eval.c (rb_proc_arity): ditto. * parse.y (f_args): syntax rule enhanced to support arguments after the splat. * parse.y (block_param): ditto for block parameters. * parse.y (f_post_arg): mandatory formal arguments after the splat argument. * parse.y (new_args_gen): generate nodes for mandatory formal arguments after the splat argument. * eval.c (rb_eval): dispatch mandatory formal arguments after the splat argument. * parse.y (args): allow more than one splat in the argument list. * parse.y (method_call): allow aref [] to accept all kind of method argument, including assocs, splat, and block argument. * eval.c (SETUP_ARGS0): prepare block argument as well. * lib/mathn.rb (Integer): remove Integer#gcd2. [ruby-core:07931] * eval.c (error_line): print receivers true/false/nil specially. * eval.c (rb_proc_yield): handles parameters in yield semantics. * eval.c (nil_yield): gives LocalJumpError to denote no block error. * io.c (rb_io_getc): now takes one-character string. * string.c (rb_str_hash): use FNV-1a hash from Fowler/Noll/Vo hashing algorithm. * string.c (rb_str_aref): str[0] now returns 1 character string, instead of a fixnum. [Ruby2] * parse.y (parser_yylex): ?c now returns 1 character string, instead of a fixnum. [Ruby2] * string.c (rb_str_aset): no longer support fixnum insertion. * eval.c (umethod_bind): should not update original class. [ruby-dev:28636] * eval.c (ev_const_get): should support constant access from within instance_eval(). [ruby-dev:28327] * time.c (time_timeval): should round for usec floating number. [ruby-core:07896] * time.c (time_add): ditto. * dir.c (sys_warning): should not call a vararg function rb_sys_warning() indirectly. [ruby-core:07886] * numeric.c (flo_divmod): the first element of Float#divmod should be an integer. [ruby-dev:28589] * test/ruby/test_float.rb: add tests for divmod, div, modulo and remainder. * re.c (rb_reg_initialize): should not allow modifying literal regexps. frozen check moved from rb_reg_initialize_m as well. * re.c (rb_reg_initialize): should not modify untainted objects in safe levels higher than 3. * re.c (rb_memcmp): type change from char* to const void*. * dir.c (dir_close): should not close untainted dir stream. * dir.c (GetDIR): add tainted/frozen check for each dir operation. * lib/rdoc/parsers/parse_rb.rb (RDoc::RubyParser::parse_symbol_arg): typo fixed. a patch from Florian Gross <florg at florg.net>. * eval.c (EXEC_EVENT_HOOK): trace_func may remove itself from event_hooks. no guarantee for arbitrary hook deletion. [ruby-dev:28632] * util.c (ruby_strtod): differ addition to minimize error. [ruby-dev:28619] * util.c (ruby_strtod): should not raise ERANGE when the input string does not have any digits. [ruby-dev:28629] * eval.c (proc_invoke): should restore old ruby_frame->block. thanks to ts <decoux at moulon.inra.fr>. [ruby-core:07833] also fix [ruby-dev:28614] as well. * signal.c (trap): sig should be less then NSIG. Coverity found this bug. a patch from Kevin Tew <tewk at tewk.com>. [ruby-core:07823] * math.c (math_log2): add new method inspired by [ruby-talk:191237]. * math.c (math_log): add optional base argument to Math::log(). [ruby-talk:191308] * ext/syck/emitter.c (syck_scan_scalar): avoid accessing uninitialized array element. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07809] * array.c (rb_ary_fill): initialize local variables first. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07810] * ext/syck/yaml2byte.c (syck_yaml2byte_handler): need to free type_tag. a patch from Pat Eyler <rubypate at gmail.com>. [ruby-core:07808] * ext/socket/socket.c (make_hostent_internal): accept ai_family check from Sam Roberts <sroberts at uniserve.com>. [ruby-core:07691] * util.c (ruby_strtod): should not cut off 18 digits for no reason. [ruby-core:07796] * array.c (rb_ary_fill): internalize local variable "beg" to pacify Coverity. [ruby-core:07770] * pack.c (pack_unpack): now supports CRLF newlines. a patch from <tommy at tmtm.org>. [ruby-dev:28601] * applied code clean-up patch from Stefan Huehner <stefan at huehner.org>. [ruby-core:07764] * lib/jcode.rb (String::tr_s): should have translated non squeezing character sequence (i.e. a character) as well. thanks to Hiroshi Ichikawa <gimite at gimite.ddo.jp> [ruby-list:42090] * ext/socket/socket.c: document update patch from Sam Roberts <sroberts at uniserve.com>. [ruby-core:07701] * lib/mathn.rb (Integer): need not to remove gcd2. a patch from NARUSE, Yui <naruse at airemix.com>. [ruby-dev:28570] * parse.y (arg): too much NEW_LIST() * eval.c (SETUP_ARGS0): remove unnecessary access to nd_alen. * eval.c (rb_eval): use ARGSCAT for NODE_OP_ASGN1. [ruby-dev:28585] * parse.y (arg): use NODE_ARGSCAT for placeholder. * lib/getoptlong.rb (GetoptLong::get): RDoc update patch from mathew <meta at pobox.com>. [ruby-core:07738] * variable.c (rb_const_set): raise error when no target klass is supplied. [ruby-dev:28582] * prec.c (prec_prec_f): documentation patch from <gerardo.santana at gmail.com>. [ruby-core:07689] * bignum.c (rb_big_pow): second operand may be too big even if it's a Fixnum. [ruby-talk:187984] * README.EXT: update symbol description. [ruby-talk:188104] * COPYING: explicitly note GPLv2. [ruby-talk:187922] * parse.y: remove some obsolete syntax rules (unparenthesized method calls in argument list). * eval.c (rb_call0): insecure calling should be checked for non NODE_SCOPE method invocations too. * eval.c (rb_alias): should preserve the current safe level as well as method definition. * process.c (rb_f_sleep): remove RDoc description about SIGALRM which is not valid on the current implementation. [ruby-dev:28464] Thu Mar 23 21:40:47 2006 K.Kosako <sndgk393 AT ybb.ne.jp> * eval.c (method_missing): should support argument splat in super. a bug in combination of super, splat and method_missing. [ruby-talk:185438] * configure.in: Solaris SunPro compiler -rapth patch from <kuwa at labs.fujitsu.com>. [ruby-dev:28443] * configure.in: remove enable_rpath=no for Solaris. [ruby-dev:28440] * ext/win32ole/win32ole.c (ole_val2olevariantdata): change behavior of converting OLE Variant object with VT_ARRAY|VT_UI1 and Ruby String object. * ruby.1: a clarification patch from David Lutterkort <dlutter at redhat.com>. [ruby-core:7508] * lib/rdoc/ri/ri_paths.rb (RI::Paths): adding paths from rubygems directories. a patch from Eric Hodel <drbrain at segment7.net>. [ruby-core:07423] * eval.c (rb_clear_cache_by_class): clearing wrong cache. * ext/extmk.rb: use :remove_destination to install extension libraries to avoid SEGV. [ruby-dev:28417] * eval.c (rb_thread_fd_writable): should not re-schedule output from KILLED thread (must be error printing). * array.c (rb_ary_flatten_bang): allow specifying recursion level. [ruby-talk:182170] * array.c (rb_ary_flatten): ditto. * gc.c (add_heap): a heap_slots may overflow. a patch from Stefan Weil <weil at mail.berlios.de>. * eval.c (rb_call): use separate cache for fcall/vcall invocation. * eval.c (rb_eval): NODE_FCALL, NODE_VCALL can call local functions. * eval.c (rb_mod_local): a new method to specify newly added visibility "local". * eval.c (search_method): search for local methods which are visible only from the current class. * class.c (rb_class_local_methods): a method to list local methods. * object.c (Init_Object): add BasicObject class as a top level BlankSlate class. * ruby.h (SYM2ID): should not cast to signed long. [ruby-core:07414] * class.c (rb_include_module): allow module duplication. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@10235 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2006-06-10 01:20:17 +04:00
rb_define_module_function(rb_mMath, "log", math_log, -1);
rb_define_module_function(rb_mMath, "log2", math_log2, 1);
rb_define_module_function(rb_mMath, "log10", math_log10, 1);
rb_define_module_function(rb_mMath, "sqrt", math_sqrt, 1);
rb_define_module_function(rb_mMath, "cbrt", math_cbrt, 1);
rb_define_module_function(rb_mMath, "frexp", math_frexp, 1);
rb_define_module_function(rb_mMath, "ldexp", math_ldexp, 2);
rb_define_module_function(rb_mMath, "hypot", math_hypot, 2);
rb_define_module_function(rb_mMath, "erf", math_erf, 1);
rb_define_module_function(rb_mMath, "erfc", math_erfc, 1);
rb_define_module_function(rb_mMath, "gamma", math_gamma, 1);
rb_define_module_function(rb_mMath, "lgamma", math_lgamma, 1);
}
void
Init_Math(void)
{
InitVM(Math);
}