rational.c: infinity in power
* rational.c (nurat_expt): return Infinity due to overflow.
[ruby-core:79686] [Bug #13242]:
rational.c: infinity in power
* rational.c (nurat_expt): return 0 due to overflow.
[ruby-core:79686] [Bug #13242]:
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@58117 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* ext/bigdecimal/bigdecimal.c (GetVpValueWithPrec): consider
non-finite float values not to raise FloatDomainError.
[ruby-core:75682] [Bug #12414]
* ext/bigdecimal/bigdecimal.c (isfinite): isfinite does not always
exist. fixed build error on Windows itroduced at r55123.
exist. fixed build error on Windows introduced at r55123.
* ext/bigdecimal/bigdecimal.c (isfinite): get rid of a warning on
cygwin. [Bug #12417][ruby-core:75691]
* include/ruby/missing.h (isfinite): move from numeric.c.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@55354 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* hash.c (rb_any_hash): use same hash values with Float#hash so
that -0.0 and +0.0 will be identical.
[ruby-core:68541] [Bug #10979]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_2@50619 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* numeric.c (flodivmod): all results are NaN if divisor is NaN.
[fix GH-692]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@47024 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
or an invalid value is returned in #coerce called by
numeric comparison operators and the exception
thrown by the caller has no information on the failure.
In the next release such exception should not be rescued or
should be the cause of the caller exception. nil is accepted
as the "no possible coercion" return value. See #7688.
* test/ruby/test_numeric.rb: Add corresponding test.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@46375 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* numeric.c (num_step_scan_args): check keyword arguments and fail
if they conflict with positional arguments.
[ruby-dev:48177] [Bug #9811]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@45861 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* numeric.c (num_step_scan_args): turn a macro into a function.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@45859 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
add links to `Object#hash` to each #`hash` methods rdocs.
[Fixes GH-567]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@45329 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* numeric.c (FLOAT_OUT_OF_RANGE): extract a macro and a helper to
raise RangeError of float.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@44611 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* numeric.c (finite): disable declaration on Windows, which can be
defined in ruby/win32.h.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@43235 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
1L << 63 on 64-bit platform is undefined, at least, according to
ISO/IEC 9899 (C99) 6.5.7.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@43222 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* numeric.c: use RB_TYPE_P() for special classes instead of switch
with TYPE().
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42909 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* numeric.c (NUM_STEP_SCAN_ARGS): remove extra class checks, which
cause the incompatibilities.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42876 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* numeric.c (NUM_STEP_GET_INF): split from NUM_STEP_SCAN_ARGS(), since
inf is not used in num_step_size().
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42829 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* numeric.c (NUM_STEP_SCAN_ARGS): On sencond thought, keep
Numeral#step backward compatible in that it raises TypeError
when nil is given as second argument.
* test/ruby/test_float.rb (TestFloat#test_num2dbl): Revert.
* test/ruby/test_numeric.rb (TestNumeric#test_step): Fix test
cases for the above change.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@42788 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
