| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
|
|
timeout of unicorn process to 60s instead of 30s. Problems with the lower
timeouts will become apparent if you have a local mail server throttling
connections and you do something like update >=10 issues simultaneously. Install
some helper shell scripts to ease with Redmine configuration and management.
Install a sample secrets.yml file for usage with the application. This package
will be renamed to ruby-redmine after the freeze and its files installed with
${RUBY_PKGPREFIX} to enable the usage and testing of different Ruby versions
with Redmine. This package has been tested heavily in production and is known to
work well in its current state; though, we will want to change the way it uses
gems when pkgsrc gets rails-4.x. Discussed with maintainter previously.
From Changelog:
3.0.3 (2015-05-10)
Defect #18580: Can't bulk edit own time entries with "Edit own time entries"
Defect #19731: Issue validation fails if % done field is deactivated
Defect #19735: Email addresses with slashes are not linked correctly
Patch #19655: Set a back_url when forcing new login after session expiration
Patch #19706: Issue show : optimizations
Patch #19793: Adding flash messages to files_controller#create
3.0.2 (2015-04-26)
Defect #19297: Custom fields with hidden/read-only combination displayed in Issue Edit Form
Defect #19400: Possibility of having 2 (or more) repositories with empty identifier
Defect #19444: Fix typo in wiki_syntax_detailed.html
Defect #19538: Keywords in commit messages: journal entries are created even if nothing was changed
Defect #19569: Field permissions not working properly with inherited memberships
Defect #19580: "Required" and "Read-only" rules on "Fields Permissions" screen are not colored
Defect #13583: Space between lines in nested lists not equal
Defect #19161: 500 Internal error: sorting for column mail at Administration/User
Defect #19163: Bulk edit form shows additional custom fields
Defect #19168: Activity: changes made to tickets are shown multiple times
Defect #19185: Update Install/Upgrade guide for 3.x version and get gid of DEPRECATION WARNING: You didn't set config.secret_key_base
Defect #19276: Creating new issues with invalid project_id should return 422 instead of 403 error
Defect #19405: Setting config.logger.level in additional_environment.rb has no effect
Defect #19464: Possible to log time on project without time tracking
Defect #19482: Custom field (long text format) displayed even if empty
Defect #19537: Broken HTML sanitizer refence breaks email receiving
Defect #19544: Malformed SQL query with SQLServer when grouping issues
Defect #19553: When create by copying the issue, status can not be changed to default
Defect #19558: Mail handler should not ignore emails with x-auto-response-suppress header
Defect #19606: Issue Estimated Time not updated on tracker change
Feature #19437: Upgrade to Rails 4.2.1
Feature #19489: Translation for Spanish Panama
Patch #19570: Spanish translation updated
|
|
|
|
|
|
of Solaris.
Change during pkgsrc freeze approved by Jonathan Perkin.
|
|
Use PERL5_MODULE_TYPE instead.
Approved by wiz@.
|
|
in joyent/pkgsrc#269.
|
|
|
|
build on SunOS.
|
|
markd reports this fixes the build on NetBSD-6.
|
|
|
|
What's new in pyatspi 2.15.90:
* Add new roles for fractions, roots, subscripts, and superscripts (bgo#743413).
What's new in pyatspi 2.15.4:
* Support ATSPI_STATE_READ_ONLY (bgo#690108).
What's new in pyatspi 2.15.3:
* Add ROLE_STATIC (bgo#740364).
|
|
|
|
=============================================
- Pango modules, engines, and config have been removed (#733882)
=> no need to deal with modules on INSTALL and DEINSTALL.
Thanks to wiz for the report.
|
|
|
|
|
|
python-hglib is a library with a fast, convenient interface to Mercurial.
It uses Mercurial's command server for communication with hg.
|
|
expat so that it understands Glade.
|
|
|
|
This module is used by Schmorp's modules during configuration stage
to test the installed perl for compatibility with his modules.
It's not, at this stage, meant as a tool for other module authors,
although in principle nothing prevents them from subscribing to
the same ideas.
|
|
2015-06-11 k <andk@cpan.org>
* release 2.15
* In the tests, always check for exists before checking for
definedness (Reini Urban)
|
|
0.413 2015-06-10
- Fix compiling in c++ mode (depreciated, but some people seem to
require it). Solves RT#104690
|
|
1.19 2015-06-12
- Fixed an uninitialized value warning from the pure Perl implementation under
5.8.8. Reported by Jim Bacon. RT #105198.
|
|
------
17.1.1
------
* Backed out unintended changes to pkg_resources, restoring removal of
deprecated imp module (`ref
<https://bitbucket.org/pypa/setuptools/commits/f572ec9563d647fa8d4ffc534f2af8070ea07a8b#comment-1881283>`_).
----
17.1
----
* Issue #380: Add support for range operators on environment
marker evaluation.
|
|
+ Version 2.14 (09.06.2015)
- Added CParser parameter to specify output directory for generated parsing
tables (#84).
- Removed lcc's cpp and its license from the distribution. Using lcc's cpp
is no longer recommended, now that Clang has binary builds available for
Windows.
|
|
1.1.2
=====
* ``ffi.gc()``: fixed a race condition in multithreaded programs
introduced in 1.1.1
1.1.1
=====
* Out-of-line mode: ``ffi.string()``, ``ffi.buffer()`` and
``ffi.getwinerror()`` didn't accept their arguments as keyword
arguments, unlike their in-line mode equivalent. (It worked in PyPy.)
* Out-of-line ABI mode: documented a restriction__ of ``ffi.dlopen()``
when compared to the in-line mode.
* ``ffi.gc()``: when called several times with equal pointers, it was
accidentally registering only the last destructor, or even none at
all depending on details. (It was correctly registering all of them
only in PyPy, and only with the out-of-line FFIs.)
.. __: cdef.html#dlopen-note
|
|
util-linux 2.26.2: Apr 30 2015
* see Documentation/releases/v2.26.2-ReleaseNotes or the complete changelog at
ftp://ftp.kernel.org/pub/linux/utils/util-linux/v2.26/v2.26.2-ChangeLog
util-linux 2.26.1: Mar 13 2015
* see Documentation/releases/v2.26.1-ReleaseNotes or the complete changelog at
ftp://ftp.kernel.org/pub/linux/utils/util-linux/v2.26/v2.26.1-ChangeLog
util-linux 2.26: Feb 19 2015
* see Documentation/releases/v2.26-ReleaseNotes or the complete changelog at
ftp://ftp.kernel.org/pub/linux/utils/util-linux/v2.26/v2.26-ChangeLog
util-linux 2.26-rc2: Feb 4 2015
* see Documentation/releases/v2.26-ReleaseNotes or the complete changelog at
ftp://ftp.kernel.org/pub/linux/utils/util-linux/v2.26/v2.26-rc2-ChangeLog
util-linux 2.26-rc1: Jan 14 2015
* see Documentation/releases/v2.26-ReleaseNotes or the complete changelog at
ftp://ftp.kernel.org/pub/linux/utils/util-linux/v2.26/v2.26-rc1-ChangeLog
util-linux 2.25: Jul 22 2014
* see Documentation/releases/v2.25-ReleaseNotes or the complete changelog at
ftp://ftp.kernel.org/pub/linux/utils/util-linux/v2.25/v2.25-ChangeLog
util-linux 2.25-rc2: Jul 2 2014
* see Documentation/releases/v2.25-ReleaseNotes or the complete changelog at
ftp://ftp.kernel.org/pub/linux/utils/util-linux/v2.25/v2.25-rc2-ChangeLog
util-linux 2.25-rc1: Jun 18 2014
* see Documentation/releases/v2.25-ReleaseNotes or the complete changelog at
ftp://ftp.kernel.org/pub/linux/utils/util-linux/v2.25/v2.25-rc1-ChangeLog
|
|
Version 6.5 - June 10 2015
[CHANGES]
New facilities:
o global: New -N (--nearness=[start]) option.
Nearness sort method is available for the output of tag search command.
The result of nearness sort is concatenation of the followings ([1]-[n])
in this order. The default of 'start' is the current directory.
[1] Output of local search in the 'start' directory.
[2] Output of local search in the parent directory except for [1].
[3] Output of local search in the grandparent directory except for [1]-[2].
(repeat until the project root directory)
[n] Output of local search in the project root directory except for [1]-[n-1].
In each directory, they are sorted by alphabetical order.
o global: Now the --literal option also works with the tag search command,
-P command and -I command as well as the -g command.
[FIXED BUGS]
o htags: The -c and -x option of htags(1) were still available in the help
message, even though they had actually been removed. Now, these options
are removed completely.
o gtags (PHP): Against the following source code, gtags(1) aborted with a message
'short of memory'. Now it works.
[nullstring.php]
+----------------
|<?php
|define('');
|?>
+----------------
o gtags (C++): Gtags(1) didn't recognize the shift operator. Now it works.
[a.hh]
+----------------
|class const_mod<uint64_t(1) << 48>
|{
|};
+----------------
$ gtags
gtags: failed to parse template [+1 ./a.hh].
o gtags (C, C++): Gtags(1) couldn't pick up 'E2' as a definition. Now it works.
[test.c]
+----------------
|enum my_enum2
|{
| E2
|};
+----------------
|
|
2.44.0 (stable):
Gio:
* Action, ActionGroup: Avoid memory leaks in funcs.
(Kjell Ahlstedt) Bug #705124
gmmproc:
* _WRAP_VFUNC(): Add keep_return parameter.
(Kjell Ahlstedt) Bug #705124
2.43.91 (unstable):
Glib:
* OptionContext: Add get/set_strict_posix().
(Murray Cumming)
Gio:
* Application:
- Add get/set/unset_resource_base_path() and property.
- Add get_is_busy() and property.
(Murray Cumming)
* File: Add replace_contents_bytes_aync().
(Murray Cumming)
* InputStream: Add read_all_async() and read_all_finish().
(Murray Cumming)
* MemoryInputStream: Add add_bytes().
(Murray Cumming)
* OutputStream: Add write_all_async() and write_all_finish().
(Murray Cumming)
Gio::DBus
* InterfaceInfo: Add cache_build() and cache_release().
(Murray Cumming)
2.43.90 (unstable):
Glib:
* Error::register_init(): Call Glib::wrap_register_init().
(Kjell Ahlstedt) Bug #743466 (Mike Fleetwood).
* OptionGroup:
- Fix enable/disable bool option pairs.
(Kjell Ahlstedt) Bug #744854 (Tom Schoonjans)
- Fix memory leaks
(Kjell Ahlstedt) Bug #745173.
- Don't use deprecate g_option_group_free().
(Kjell Ahlstedt)
* Value: Deprecate Value<char>, add Value<signed char>.
Because g_value_[get,set]_char() are deprecated in favour of
g_value_[get,set]_schar().
(Kjell Ahlstedt)
Gio:
* Application: Deprecate property_action_group().
(Kjell Ahlstedt)
* Notification: Add set_priority() and enum NotificationPriority.
(Kjell Ahlstedt)
* Add TcpWrapperConnection.
(Murray Cumming)
* UnixSocketAddress: Deprecate property_abstract().
(Kjell Ahlstedt)
gmmproc:
* Fix error messages in glib and gio by removing unnecessary _IGNORES().
(Kjell Ahlstedt)
* h2def.py: Remove *_DEPRECATED_IN_*_*_FOR(*) prefixes with white space.
(Kjell Ahlstedt)
* Put DOXYGEN_SHOULD_SKIP_THIS around *_Class prototypes.
To workaround a doxygen bug, to fix the genereated DevHelp search index.
(Murray Cumming) Bug #743918
2.43.3 (unstable):
Glib:
* Binding: Rename and change BindingTransformSlot to SlotTransform.
(Kjell Ahlstedt) Bug #738663.
* Add SlotSpawnChildSetup.
(Kjell Ahlstedt) Bug #528285.
Documentation:
* Resource: Suppress incorrect doxygen links.
(Kjell Ahlstedt)
2.43.2 (unstable):
Gio:
* Added NetworkMonitor.
(Murray Cumming)
* UnixFDList, UnixFDMessage: Correct array lengths in steal_fds()
(Kjell Ahlstedt) Bug #741365 (Matthew Balkam)
gmmproc:
* Don't make one very long line for the enum documentation.
(Kjell Ahlstedt)
* Improve the conversion of Since to @newin.
(Kjell Ahlstedt)
* Add an empty line after @newin where it's missing
(Kjell Ahlstedt)
2.43.1 (unstable):
gmmproc:
* Tidy up the generation of enum docs
(Kjell Ahlstedt)
* _WRAP_GERROR: Add documentation to the generated enum Code.
(Kjell Ahlstedt)
* Change messages that MS Visual Studio can misunderstand.
(Kjell Ahlstedt)
* Warn when an ignored method or signal doesn't exist.
(Marcin Kolny) Bug #737212.
Glib:
* Add Binding.
(Kjell Ahlstedt) Bug #738663.
* Checksum::ChecksumType: Remove erroneous documentation
(Kjell Ahlstedt)
* Property: Add some documentation.
(Kjell Ahlstedt) Bug #523043.
Gio:
Add Resource.
(Kjell Ahlstedt)
|
|
Version 1.82b:
--------------
- Fixed a harmless but annoying race condition in persistent mode - signal
delivery is a bit more finicky than I thought.
- Updated the documentation to explain persistent mode a bit better.
- Tweaked AFL_PERSISTENT to force AFL_NO_VAR_CHECK.
--------------
Version 1.81b:
--------------
- Added persistent mode for in-process fuzzing. See llvm_mode/README.llvm.
Inspired by Kostya Serebryany and Christian Holler.
- Changed the in-place resume code to preserve crashes/README.txt. Suggested
by Ben Nagy.
- Included a potential fix for LLVM mode issues on MacOS X, based on the
investigation done by teor2345.
|
|
### 3.3.1 / 2015-06-14
[Full Changelog](http://github.com/rspec/rspec-rails/compare/v3.3.0...v3.3.1)
Bug Fixes:
* Fix regression that caused stubbing ActiveRecord model classes to
trigger internal errors in rails. (Myron Marston, Aaron Kromer, #1395)
|
|
DEPENDS on devel/p5-Module-Build.
|
|
Addition of \csmeaning command.
|
|
## 1.3.2
### Changes
- now works and passes tests with Lua 5.3
- utils.import will NOT override global symbols (import 'math' caused global type() to be clobbered)
- Updated pl.dir.file_op to return true on success and false on failure...
- workaround for issues with pl.lapp with amalg.lua - will look at global LAPP_SCRIPT if arg[0] is nil
### Fixes
- func was broken: do NOT use ipairs to iterate if __index is overriden!
- issue #133 pretty.read (naively) confused by unbalanced brackets
- xml attribute underscore fix for simple parser
- Fix path.normpath
- lexer: fix parsing block comments/string. fix hang on empty string.
- Fixed utils.execute returning different values for Lua 5.1 and Lua 5.2
- Issue #97; fixed attempt to put a month into a day
- problem with tablex.count_map with custom comparison
### Features
- Add Python style url module for quote and unquote.
- stringx.quote_string, which scans for embedded long-string quote matches and escapes them by creating a long-string quote.
- issue #117: tablex.range now works with decreasing numbers, consistent with numerical for loop
- utils.import will NOT override global symbols (import 'math' caused global type() to be clobbered)
- issue #125: DOCTYPE ignored in xml documents as well
- Allow XML tostring() function to customize the default prefacing with <?xml...>
- More Robust Quoted Strings
- lapp: improved detection of unsupported short flags
## 1.3.0
### Changes
- class: RIP base method - not possible to implement correctly
- lapp: short flags can now always be followed directly by their value, for instance,
`-I/usr/include/lua/5.1`
- Date: new explicit `Date.Interval` class; `toUTC/toLocal` return new object; `Date.__tostring`
always returns ISO 8601 times for exact serialization. `+/-` explicit operators. Date objects
are explicitly flagged as being UTC or not.
### Fixes
- class: super method fixed.
- Date: DST is now accounted for properly.
- Date: weekday calculation borked.
### Features
- All tests pass with no-5.1-compatible Lua 5.2; now always uses `utils.load` and
`utils.unpack` is always available.
- types: new module containing `utils.is_xxx` methods plus new `to_bool`.
- class: can be passed methods in a table (see `test=klass.lua`). This is
particularly convenient for using from Moonscript.
- general documentation improvements, e.g `class`
## 1.2.1
### Changes
- utils.set(get)fenv always defined (_not_ set as globals for 5.2 anymore!).
These are defined in new module pl.compat, but still available through utils.
- class.Frodo now puts 'Frodo' in _current environment_
### Fixes
- lapp.add_type was broken (Pete Kazmier)
- class broke with classes that redefined __newindex
- Set.isdisjoint was broken because of misspelling; default ctor Set() now works as expected
- tablex.transform was broken; result now has same keys as original (CoolistheName007)
- xml match not handling empty matches (royalbee)
- pl.strict: assigning nil to global declares it, as God intended. (Pierre Chapuis)
- tests all work with pl.strict
- 5.2 compatible load now respects mode
- tablex.difference thought that a value of `false` meant 'not present' (Andrew Starke)
### Features
- tablex.sort(t) iterates over sorted keys, tablex.sortv(t) iterates over sorted values (Pete Kazmier)
- tablex.readonly(t) creates a read-only proxy for a table (John Schember)
- utils.is_empty(o) true if o==nil, o is an empty table, or o is an empty string (John Schember)
- utils.executeex(cmd,bin) returns true if successful, return code, plus stdout and stderr output as strings. (tieske)
- class method base for calling inherited methods (theypsilon)
- class supports pre-constructor _create for making a custom self (used in pl.List)
- xml HTML mode improvements - can parse non-trivial well-formed HTML documents.
xml.parsehtml is a parse function, no longer a flag
- if a LOM document has ordered attributes, use these when stringifying
- xml.tostring has yet another extra parm to force prefacing with <?xml...>
- lapp boolean flags may have `true` default
- lapp slack mode where 'short' flags can be multi-char
- test.asserteq etc take extra arg, which is extra level where error must be reported at
- path.currentdir,chdir,rmdir,mkdir and dir as alias to lfs are exported; no dependencies on luafilesystem outside pl.path, making it easier to plug in different implementations.
|
|
patch-gold_stringpool.cc
|
|
Changes: fix regressions in 0.5 (sigh) and a latent bug that was part
of PR 49969.
|
|
|
|
### 3.3.0 / 2015-06-12
[Full Changelog](http://github.com/rspec/rspec-rails/compare/v3.2.3...v3.3.0)
Enhancements:
* Add support for PATCH to route specs created via scaffold. (Igor Zubkov, #1336)
* Improve controller and routing spec calls to `routes` by using `yield`
instead of `call`. (Anton Davydov, #1308)
* Add support for `ActiveJob` specs as standard `RSpec::Rails::RailsExampleGoup`s
via both `:type => :job` and inferring type from spec directory `spec/jobs`.
(Gabe Martin-Dempesy, #1361)
* Include `RSpec::Rails::FixtureSupport` into example groups using metadata
`:use_fixtures => true`. (Aaron Kromer, #1372)
* Include `rspec:request` generator for generating request specs; this is an
alias of `rspec:integration` (Aaron Kromer, #1378)
* Update `rails_helper` generator with a default check to abort the spec run
when the Rails environment is production. (Aaron Kromer, #1383)
|
|
This is meta rubygems contains rspc-core, rspec-mocks,
ruby-rspec-expectations and rspec-support.
|
|
### 3.3.0 / 2015-06-12
[Full Changelog](http://github.com/rspec/rspec-expectations/compare/v3.2.1...v3.3.0)
Enhancements:
* Expose `RSpec::Matchers::EnglishPhrasing` to make it easier to write
nice failure messages in custom matchers. (Jared Beck, #736)
* Add `RSpec::Matchers::FailMatchers`, a mixin which provides
`fail`, `fail_with` and `fail_including` matchers for use in
specifying that an expectation fails for use by
extension/plugin authors. (Charlie Rudolph, #729)
* Avoid loading `tempfile` (and its dependencies) unless
it is absolutely needed. (Myron Marston, #735)
* Improve failure output when attempting to use `be_true` or `be_false`.
(Tim Wade, #744)
* Define `RSpec::Matchers#respond_to_missing?` so that
`RSpec::Matchers#respond_to?` and `RSpec::Matchers#method` handle
dynamic predicate matchers. (Andrei Botalov, #751)
* Use custom Time/DateTime/BigDecimal formatting for all matchers
so they are consistently represented in failure messages.
(Gavin Miller, #740)
* Add configuration to turn off warnings about matcher combinations that
may cause false positives. (Jon Rowe, #768)
* Warn when using a bare `raise_error` matcher that you may be subject to
false positives. (Jon Rowe, #768)
* Warn rather than raise when using the`raise_error` matcher in negative
expectations that may be subject to false positives. (Jon Rowe, #775)
* Improve failure message for `include(a, b, c)` so that if `a` and `b`
are included the failure message only mentions `c`. (Chris Arcand, #780)
* Allow `satisfy` matcher to take an optional description argument
that will be used in the `description`, `failure_message` and
`failure_message_when_negated` in place of the undescriptive
"sastify block". (Chris Arcand, #783)
* Add new `aggregate_failures` API that allows multiple independent
expectations to all fail and be listed in the failure output, rather
than the example aborting on the first failure. (Myron Marston, #776)
* Improve `raise_error` matcher so that it can accept a matcher as a single argument
that matches the message. (Time Wade, #782)
Bug Fixes:
* Make `contain_exactly` / `match_array` work with strict test doubles
that have not defined `<=>`. (Myron Marston, #758)
* Fix `include` matcher so that it omits the diff when it would
confusingly highlight items that are actually included but are not
an exact match in a line-by-line diff. (Tim Wade, #763)
* Fix `match` matcher so that it does not blow up when matching a string
or regex against another matcher (rather than a string or regex).
(Myron Marston, #772)
* Silence whitespace-only diffs. (Myron Marston, #801)
|
|
### 3.3.0 / 2015-06-12
[Full Changelog](http://github.com/rspec/rspec-mocks/compare/v3.2.1...v3.3.0)
Enhancements:
* When stubbing `new` on `MyClass` or `class_double(MyClass)`, use the
method signature from `MyClass#initialize` to verify arguments.
(Myron Marston, #886)
* Use matcher descriptions when generating description of received arguments
for mock expectation failures. (Tim Wade, #891)
* Avoid loading `stringio` unnecessarily. (Myron Marston, #894)
* Verifying doubles failure messages now distinguish between class and instance
level methods. (Tim Wade, #896, #908)
* Improve mock expectation failure messages so that it combines both
number of times and the received arguments in the output. (John Ceh, #918)
* Improve how test doubles are represented in failure messages.
(Siva Gollapalli, Myron Marston, #932)
* Rename `RSpec::Mocks::Configuration#when_declaring_verifying_double` to
`RSpec::Mocks::Configuration#before_verifying_doubles` and utilise when
verifying partial doubles. (Jon Rowe, #940)
* Use rspec-support's `ObjectFormatter` for improved formatting of
arguments in failure messages so that, for example, full time
precisions is displayed for time objects. (Gavin Miller, Myron Marston, #955)
Bug Fixes:
* Ensure expectations that raise eagerly also raise during RSpec verification.
This means that if exceptions are caught inside test execution the test will
still fail. (Sam Phippen, #884)
* Fix `have_received(msg).with(args).exactly(n).times` and
`receive(msg).with(args).exactly(n).times` failure messages
for when the message was received the wrong number of times with
the specified args, and also received additional times with other
arguments. Previously it confusingly listed the arguments as being
mis-matched (even when the double was allowed to receive with any
args) rather than listing the count. (John Ceh, #918)
* Fix `any_args`/`anything` support so that we avoid calling `obj == anything`
on user objects that may have improperly implemented `==` in a way that
raises errors. (Myron Marston, #924)
* Fix edge case involving stubbing the same method on a class and a subclass
which previously hit a `NoMethodError` internally in RSpec. (Myron Marston #954)
* Fix edge case where the message received count would be incremented multiple
times for one failure. (Myron Marston, #957)
* Fix failure messages for when spies received the expected message with
different arguments and also received another message. (Maurcio Linhares, #960)
* Silence whitespace-only diffs. (Myron Marston, #969)
|
|
### 3.3.0 / 2015-06-12
[Full Changelog](http://github.com/rspec/rspec-core/compare/v3.2.3...v3.3.0)
Enhancements:
* Expose the reporter used to run examples via `RSpec::Core::Example#reporter`.
(Jon Rowe, #1866)
* Make `RSpec::Core::Reporter#message` a public supported API. (Jon Rowe, #1866)
* Allow custom formatter events to be published via
`RSpec::Core::Reporter#publish(event_name, hash_of_attributes)`. (Jon Rowe, #1869)
* Remove dependency on the standard library `Set` and replace with `RSpec::Core::Set`.
(Jon Rowe, #1870)
* Assign a unique id to each example and group so that they can be
uniquely identified, even for shared examples (and similar situations)
where the location isn't unique. (Myron Marston, #1884)
* Use the example id in the rerun command printed for failed examples
when the location is not unique. (Myron Marston, #1884)
* Add `config.example_status_persistence_file_path` option, which is
used to persist the last run status of each example. (Myron Marston, #1888)
* Add `:last_run_status` metadata to each example, which indicates what
happened the last time an example ran. (Myron Marston, #1888)
* Add `--only-failures` CLI option which filters to only the examples
that failed the last time they ran. (Myron Marston, #1888)
* Add `--next-failure` CLI option which allows you to repeatedly focus
on just one of the currently failing examples, then move on to the
next failure, etc. (Myron Marston, #1888)
* Make `--order random` ordering stable, so that when you rerun a
subset with a given seed, the examples will be order consistently
relative to each other. (Myron Marston, #1908)
* Set example group constant earlier so errors when evaluating the context
include the example group name (Myron Marson, #1911)
* Make `let` and `subject` threadsafe. (Josh Cheek, #1858)
* Add version information into the JSON formatter. (Mark Swinson, #1883)
* Add `--bisect` CLI option, which will repeatedly run your suite in
order to isolate the failures to the smallest reproducible case.
(Myron Marston, #1917)
* For `config.include`, `config.extend` and `config.prepend`, apply the
module to previously defined matching example groups. (Eugene Kenny, #1935)
* When invalid options are parsed, notify users where they came from
(e.g. `.rspec` or `~/.rspec` or `ENV['SPEC_OPTS']`) so they can
easily find the source of the problem. (Myron Marston, #1940)
* Add pending message contents to the json formatter output. (Jon Rowe, #1949)
* Add shared group backtrace to the output displayed by the built-in
formatters for pending examples that have been fixed. (Myron Marston, #1946)
* Add support for `:aggregate_failures` metadata. Tag an example or
group with this metadata and it'll use rspec-expectations'
`aggregate_failures` feature to allow multiple failures in an example
and list them all, rather than aborting on the first failure. (Myron
Marston, #1946)
* When no formatter implements #message add a fallback to prevent those
messages being lost. (Jon Rowe, #1980)
* Profiling examples now takes into account time spent in `before(:context)`
hooks. (Denis Laliberté, Jon Rowe, #1971)
* Improve failure output when an example has multiple exceptions, such
as one from an `it` block and one from an `after` block. (Myron Marston, #1985)
Bug Fixes:
* Handle invalid UTF-8 strings within exception methods. (Benjamin Fleischer, #1760)
* Fix Rake Task quoting of file names with quotes to work properly on
Windows. (Myron Marston, #1887)
* Fix `RSpec::Core::RakeTask#failure_message` so that it gets printed
when the task failed. (Myron Marston, #1905)
* Make `let` work properly when defined in a shared context that is applied
to an individual example via metadata. (Myron Marston, #1912)
* Ensure `rspec/autorun` respects configuration defaults. (Jon Rowe, #1933)
* Prevent modules overriding example group defined methods when included,
prepended or extended by config defined after an example group. (Eugene Kenny, #1935)
* Fix regression which caused shared examples to be mistakenly run when specs
where filtered to a particular location. (Ben Axnick, #1963)
* Fix time formatting logic so that it displays 70 seconds as "1 minute,
10 seconds" rather than "1 minute, 1 second". (Paul Brennan, #1984)
* Fix regression where the formatter loader would allow duplicate formatters.
(Jon Rowe, #1990)
|
|
### 3.3.0 / 2015-06-12
[Full Changelog](http://github.com/rspec/rspec-support/compare/v3.2.2...v3.3.0)
Enhancements:
* Improve formatting of arrays and hashes in failure messages so they
use our custom formatting of matchers, time objects, etc.
(Myron Marston, Nicholas Chmielewski, #205)
* Use improved formatting for diffs as well. (Nicholas Chmielewski, #205)
Bug Fixes:
* Fix `FuzzyMatcher` so that it checks `expected == actual` rather than
`actual == expected`, which avoids errors in situations where the
`actual` object's `==` is improperly implemented to assume that only
objects of the same type will be given. This allows rspec-mocks'
`anything` to match against objects with buggy `==` definitions.
(Myron Marston, #193)
|
|
|
|
|
|
|
|
Revbump as well...
|
|
|
|
3.26: 2015-06-10
Output redirect with '>>' appends to the output file.
The strip blob command can take a selection set.
Repo source type is now kept inline in stream files and emitted on write.
Legacy IDs are now kept inline in stream files and emitted on write.
The selection syntax <#nnn> now names commit nnn, 1-origin numbering.
3.25: 2015-06-03
Reading hg is now supported through an extractor class, no plugin required.
Now 'blob' command allows creating new content from lift scripts.
3.24: 2015-05-31
Fix repository-type detection bug introduced in 3.23.
DVCS Migration HOWTO is now merged into this distribution.
Legacy-ID detection is now more discriminating, based on sourcetype.
3.23: 2015-05-29
The 'prefer' command no longer changes the repo type, but 'sourcetype' does.
The =N set is all commits and tags with text matching a legacy ID.
Legacy-ID recognition is more intelligent, depending on the source type.
The 'lint' command now checks for time and action-stamp collisions.
Exporters can now embed a repository type declaration in an import stream.
Fast-import syntax extensions are documented.
3.22: 2015-05-20
The path rename --relax option has been removed as too confusing.
In selection expressions, =Z is the set of all commits with no fileops.
repopuller has been replaced by the more general repotool.
For consistency, the 'changed' modifier of mailbox_in is now '--changed'.
New '--create' option of mailbox_in to create new tags and commits.
The command prompt can be queried or set with the new command 'prompt'.
Change in terminology: fossil IDs and maps are now legacy IDs and maps.
A front end can now set legacy IDs via the "legacy-id" property.
DMRCN suffix flags on restrict path regexp matches to specified op types.
The 'remove' command can now be told to look for specific fileop types.
3.21: 2015-04-02
In selection expressions, =U is the set of all commits with callouts.
3.20: 2015-02-22
Code is now fully functional, all regression test passing, on Mac OS X.
Warn in the docs about the consequences of case-smashing filesystems.
diff command no longer relies on external diff(1).
Bugfix for writing callouts in partial dumps.
|
|
Bump PKGREVISION.
Addresses PR 49663.
|
