`SecureRandom#gen_random` determines whether urandom is available or not
by trying `Random.urandom(n)`. But, when n = 0, `Random.urandom(0)`
always succeeds even if urandom is not available, which leads to a wrong
decision.
When failed, `Random.urandom` returns nil instead of returning a shorter
string than required. So the check for `ret.length != n` is not needed.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59840 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
rubygems 2.7.x depends bundler-1.15.x. This is preparation for
rubygems and bundler migration.
* lib/bundler.rb, lib/bundler/*: files of bundler-1.15.4
* spec/bundler/*: rspec examples of bundler-1.15.4. I applied patches.
* https://github.com/bundler/bundler/pull/6007
* Exclude not working examples on ruby repository.
* Fake ruby interpriter instead of installed ruby.
* Makefile.in: Added test task named `test-bundler`. This task is only
working macOS/linux yet. I'm going to support Windows environment later.
* tool/sync_default_gems.rb: Added sync task for bundler.
[Feature #12733][ruby-core:77172]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59779 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
- In common.mk, set UNICODE_VERSION to 10.0.0
- Generate and add enc/unicode/10.0.0/casefold.h and
enc/unicode/10.0.0/name2ctype.h
- Update lib/unicode_normalize/tables.rb
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59759 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/net/http/header.rb (set_field): `val` can not have `to_str`.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59695 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This version fixed strange behavior of ruby code parser.
We will list all of impromovement to Changelog when 6.0.0 releasing.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59686 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
The IMAP server of DOCOMO returns such continuation requests.
[ruby-list:50558]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59666 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/csv.rb (CSV::Table#to_a, #to_csv): use Array#push instead of
Array#concat for performance improvement. This performance improvement is
proposed by zdennis <zach.dennis@gmail.com>. The patch is from
Mau Magnaguagno <maumagnaguagno@gmail.com>.
close#946
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59657 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
It ignored all of lines when given Regexp special characters.
[Feature #9147][ruby-core:58549]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59639 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/uri/ftp.rb: [DOC] fix format of example URLs. patched by
aycabta (Code Ahss) at [ruby-core:82379]. [Bug #13814]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59598 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/rubygems/resolver.rb (sort_dependencies): use stable sort.
TestGemRequestSetLockfile#test_to_s_gem_dependency_non_default
fails because this method return unstable results.
Note that Enumerable#sort_by is unstable.
I'm not sure the "stable" nature is required for RubyGems.
The fact is that using stable sort, the test passed on
mswin64+VS2017 where the sort results was reverse (unstable) order.
Also using `-i` instead of `i` (it means forcing unstable sort)
this test fails on other platform where the test successed before.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59502 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/csv.rb: fix incompatibility introduced in r59428.
CSV.new takes options as keyword arguments.
* test/csv/test_features.rb: add a test to ensure it raises error againt
unknown options
* test/csv/test_features.rb: add a test to ensure row_sep option is properly
applied
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59437 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/csv.rb: usb keyword parameters to receive options
* test/csv/test_features.rb: remove a test for checking options
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59428 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/optparse.rb (candidate): short options are case-sensitive by
the default, should not match case-different options..
https://github.com/mernen/completion-ruby/pull/9#issuecomment-317287946
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59412 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Because progname was memoized with ||= a logger call that involved
outputting false would be nil. Example code:
logger = Logger.new(STDOUT)
logger.info(false) # => nil
Perform an explicit nil check instead of ||= so that false will be output.
patched by Gavin Miller <gavingmiller@gmail.com> [Fix GH-1667]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59380 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Based on the patch by keysen (Jérémy Carlier).
[ruby-core:81641] [Bug #13649]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59372 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* lib/optparse.rb: try Float() and Integer() instead of eval,
which does too much things.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59312 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Previously if the input started with a '0' then it will be converted
as octal even though it has been specified as a decimal. This commit
forces the number to be interpreted as a decimal.
[ruby-core:81927] [Bug #13722] [Fix GH-1665]
Author: william <william.mccumstie@outlook.com>
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59273 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
