11 years agoMore extensive travis testing plan
Peter Rabbitson [Thu, 29 Nov 2012 13:57:30 +0000]
More extensive travis testing plan

11 years agoFix the leaktracer rewrite in 218b7c12 which broke under 5.10
Peter Rabbitson [Fri, 30 Nov 2012 14:44:26 +0000]
Fix the leaktracer rewrite in 218b7c12 which broke under 5.10

Reduce the amount of shuffling happening in CLONE - use a fresh hashref
instead

11 years agoSome podlator stacks do not like multi-word =begin
Peter Rabbitson [Thu, 29 Nov 2012 13:50:36 +0000]
Some podlator stacks do not like multi-word =begin

11 years agoUpdate .travis.yml with faster deps and DBICTESTs
Brendan Byrd [Mon, 12 Nov 2012 21:06:28 +0000]
Update .travis.yml with faster deps and DBICTESTs

12 years agoStop stripping table name qualifier on complex $rs->update/delete
Peter Rabbitson [Mon, 19 Nov 2012 07:49:45 +0000]
Stop stripping table name qualifier on complex $rs->update/delete

RT#80015, RT#78844

12 years agoBack out frivolous loading of threads.pm
Peter Rabbitson [Sun, 18 Nov 2012 20:17:09 +0000]
Back out frivolous loading of threads.pm

Not only was this an unsuccesful attempt to mask out a bug introduced
in Moo's inlining of Devel::GlobalDestruction (since fixed), but it
broke a test under older win32 perls.
Revert 8bab2062, bump Moo dep

12 years agoRewrite 461c7168 solving Oracle silent connection failures
Peter Rabbitson [Sun, 18 Nov 2012 22:32:57 +0000]
Rewrite 461c7168 solving Oracle silent connection failures

The patch incorrectly masks the underlying issue - during the refactor
of 584ea6e4 a bunch of exception traps were introduced which were never
there in the first place. _dbh_get_info() should always be throwing an
error if one takes place (the only place before 584ea6e4 was in the
FreeTDS version determination, and that one made no sense anyway).
_server_info() OTOH is indeed an opportunistic routine - leave the
"rethrow if determining driver" escape as-is

12 years agoRename and expand false-value prefetch test
Peter Rabbitson [Sun, 18 Nov 2012 20:08:58 +0000]
Rename and expand false-value prefetch test

12 years agoWhile the TODO was solved in b11221ff the source of confusion was not
Peter Rabbitson [Sun, 18 Nov 2012 19:59:44 +0000]
While the TODO was solved in b11221ff the source of confusion was not

12 years agoList of what needs undoing since 0.08196
Peter Rabbitson [Sat, 25 Aug 2012 22:45:30 +0000]
List of what needs undoing since 0.08196

12 years agoExtended mailmap (name match only)
Peter Rabbitson [Sun, 18 Nov 2012 18:35:37 +0000]
Extended mailmap (name match only)

12 years agoFix test skips just like was done in 49b3a264
Peter Rabbitson [Wed, 14 Nov 2012 08:54:53 +0000]
Fix test skips just like was done in 49b3a264

12 years agoMore robust behavior under heavily threaded environments
Peter Rabbitson [Wed, 14 Nov 2012 08:54:18 +0000]
More robust behavior under heavily threaded environments

12 years agoReorganize metadata, add github addresses
Peter Rabbitson [Wed, 14 Nov 2012 07:16:50 +0000]
Reorganize metadata, add github addresses
Make the meta repository read github, so that metacpan shows cool
buttons for the cool kids. Note - the official repo still does and will
continue to reside on sc

12 years agodoc fix for Resultset find_or_new and update_or_new methods
Michael Jemmeson [Thu, 8 Nov 2012 14:01:27 +0000]
doc fix for Resultset find_or_new and update_or_new methods

12 years agoRelease 0.08204 v0.08204
Torsten Raudssus [Thu, 8 Nov 2012 19:56:08 +0000]
Release 0.08204

12 years agoBump CAG to a bugfixed version
Peter Rabbitson [Thu, 8 Nov 2012 12:00:10 +0000]
Bump CAG to a bugfixed version

12 years agoAsk for newer CAG which asks for bugfixed CXSA
Peter Rabbitson [Mon, 5 Nov 2012 08:31:44 +0000]
Ask for newer CAG which asks for bugfixed CXSA

12 years agoSquash NaN warnings on older perls
Peter Rabbitson [Mon, 5 Nov 2012 08:15:28 +0000]
Squash NaN warnings on older perls

12 years agoSwitch vestigial warn() calls to carp()
Peter Rabbitson [Mon, 5 Nov 2012 08:11:19 +0000]
Switch vestigial warn() calls to carp()

12 years agoBetter diagnostics in the case of missing drivers, slight ADO/ODBC refactor
Peter Rabbitson [Mon, 5 Nov 2012 02:35:45 +0000]
Better diagnostics in the case of missing drivers, slight ADO/ODBC refactor

12 years agoMore 5.8 nested-cref test leaks
Peter Rabbitson [Mon, 5 Nov 2012 05:18:16 +0000]
More 5.8 nested-cref test leaks

12 years agoExtra tests from an old debugging session
Peter Rabbitson [Thu, 5 Jan 2012 02:46:01 +0000]
Extra tests from an old debugging session

12 years agoRemove code which migrated into the core of SQLA
Peter Rabbitson [Wed, 11 Apr 2012 05:20:10 +0000]
Remove code which migrated into the core of SQLA

12 years agoTest that ->new() on a ResultSet object returns a Row
Fitz Elliott [Thu, 12 Jul 2012 21:26:59 +0000]
Test that ->new() on a ResultSet object returns a Row

12 years agoFix API mismatch between new_related() and new_result() (RT#78336)
Peter Rabbitson [Sat, 3 Nov 2012 17:12:06 +0000]
Fix API mismatch between new_related() and new_result() (RT#78336)

Almost 7 years ago a refactor in fea3d045 fed an (undocumented) $attrs
parameter from new_related() to new_result(), while new_result() never
expected (and ignored) said parameter. Since this is an undocumented
feature, which nobody complained about for all this time - document
extensively and kill it with fire.

12 years agoClarify CookBook example (RT#78283)
Peter Rabbitson [Sat, 3 Nov 2012 16:32:13 +0000]
Clarify CookBook example (RT#78283)

12 years agoFix CookBook example until the rsrc refactor (RT#77910)
Peter Rabbitson [Sat, 3 Nov 2012 16:18:54 +0000]
Fix CookBook example until the rsrc refactor (RT#77910)

12 years agoImprove IC documentation (RT#76525)
Peter Rabbitson [Sat, 3 Nov 2012 16:12:37 +0000]
Improve IC documentation (RT#76525)

12 years agoMention the internal join pruner in POD
Alessandro Ranellucci [Sat, 7 Apr 2012 17:37:33 +0000]
Mention the internal join pruner in POD

12 years agoFix unique constraint violations in Ordered.pm blanket movement (RT#79773)
Peter Rabbitson [Sat, 3 Nov 2012 12:44:57 +0000]
Fix unique constraint violations in Ordered.pm blanket movement (RT#79773)

This essentially reverts 5e6fde33e5 which was applied without much thinking.
Added a test to ensure this doesn't re-break again.

12 years agoProperly signify internal cursor methods, trash their accessors as well
Peter Rabbitson [Fri, 14 Sep 2012 05:39:06 +0000]
Properly signify internal cursor methods, trash their accessors as well

(the accessors were not used throughout anyway)

12 years agofix Cursor SYNOPSIS
Rafael Kitover [Sun, 26 Aug 2012 16:21:37 +0000]
fix Cursor SYNOPSIS

12 years agoReorganize constants handling, add escapes for fork-less OSes
Peter Rabbitson [Mon, 17 Sep 2012 13:55:30 +0000]
Reorganize constants handling, add escapes for fork-less OSes

12 years agoLet SQLMaker rs_attr 'for' support string literals
Matt Phillips [Fri, 19 Oct 2012 18:49:46 +0000]
Let SQLMaker rs_attr 'for' support string literals

SQLMaker previously only allowed hardcoded values with the 'for' attr,
overriding in storage specific subclasses. Rather than attempt to provide an
exhaustive list of possible options, the base class can now take \$scalaras
an override that is embedded directly in the returned $sql

12 years agoCompress changelog a bit (excise devrel fluff)
Peter Rabbitson [Sat, 3 Nov 2012 11:08:06 +0000]
Compress changelog a bit (excise devrel fluff)

12 years agoFix invalid pod example syntax
Felix Antonius Wilhelm Ostmann [Sat, 20 Oct 2012 13:54:34 +0000]
Fix invalid pod example syntax

12 years agorelease 0.08203_01
Peter Rabbitson [Sat, 20 Oct 2012 14:03:02 +0000]
release 0.08203_01

12 years agoFix embarassing test-leak introduced in 0.08203
Peter Rabbitson [Sat, 20 Oct 2012 13:50:42 +0000]
Fix embarassing test-leak introduced in 0.08203

12 years agorelease 0.08203 v0.08203
Arthur Axel 'fREW' Schmidt [Thu, 18 Oct 2012 18:24:46 +0000]
release 0.08203

12 years agorelease 0.08202_01
Arthur Axel 'fREW' Schmidt [Mon, 15 Oct 2012 12:49:55 +0000]
release 0.08202_01

12 years agoReally fix the DBD::SQLite ping() issue
Peter Rabbitson [Tue, 9 Oct 2012 11:35:09 +0000]
Really fix the DBD::SQLite ping() issue

Apparently DBD::SQLite is notoriously bad at synchronizing its internal
transaction state with {AutoCommit}
https://metacpan.org/source/ADAMK/DBD-SQLite-1.37/lib/DBD/SQLite.pm#L921
There is a function http://www.sqlite.org/c3ref/get_autocommit.html
but DBD::SQLite does not expose it (nor does it seem to properly use it)

Furthermore the detection is rather broken as described in RT#80087.
Bend over backwards to attempt to preserve as much sanity as possible.
While at it issue a non-trappable warning so that folks fix the offending
codepath (which arguably is still broken)

It is possible to have a proper "connection", and have "ping" return
false anyway (e.g. corrupted file). In such cases DBD::SQLite still
keeps the actual file handle open. We don't really want this to happen,
so force-close the handle via DBI itself (this solves a bunch of Win32
test failures)

12 years agoRemove TODO on fixed recent DBD::SQLite
Peter Rabbitson [Fri, 12 Oct 2012 14:20:11 +0000]
Remove TODO on fixed recent DBD::SQLite

12 years agorelease commit for 0.08202 v0.08202
Matt S Trout [Sat, 6 Oct 2012 17:35:24 +0000]
release commit for 0.08202

12 years agorelease 0.08201 v0.08201
Arthur Axel 'fREW' Schmidt [Fri, 5 Oct 2012 21:18:28 +0000]
release 0.08201

12 years agoFix test using legacy order_by spec
Peter Rabbitson [Thu, 30 Aug 2012 17:09:09 +0000]
Fix test using legacy order_by spec

12 years agoReplace inadequate $dbh->ping SQLite implementation (RT#78420)
Peter Rabbitson [Thu, 30 Aug 2012 17:06:13 +0000]
Replace inadequate $dbh->ping SQLite implementation (RT#78420)

When SQLite attempts to connect to a file that is not a database,
it nevertheless maintains a true $dbh->{Active} and $dbh->ping.
Replace with a schema listing SELECT, and fix test erroneously
assuming it can portably do chmod 000

12 years agoDon't ignore README, it's supposed to be cleaned automatically.
Dagfinn Ilmari Mannsåker [Sat, 1 Sep 2012 16:13:20 +0000]
Don't ignore README, it's supposed to be cleaned automatically.

12 years agoRevert "Fix t/storage/reconnect.t when run as root"
Dagfinn Ilmari Mannsåker [Sat, 1 Sep 2012 13:39:19 +0000]
Revert "Fix t/storage/reconnect.t when run as root"

The test change just papered over the problem, a proper fix is pending
on the for_current/riba/fix_sqlite_ping_RT78420 branch.

This reverts commit 64819b0e69a844bb790701e13931afcc34f3d981.

12 years agoFix t/storage/reconnect.t when run as root
Dagfinn Ilmari Mannsåker [Sat, 1 Sep 2012 01:22:27 +0000]
Fix t/storage/reconnect.t when run as root

As root, the chmod 000 doesn't prevent SQLite form opening the fake db
file, so we need to disconnect to make it reopen the restored real one.

12 years agoFix filename interpolation in DBIC::Optional::Dependecies
Dagfinn Ilmari Mannsåker [Sat, 25 Aug 2012 00:40:09 +0000]
Fix filename interpolation in DBIC::Optional::Dependecies

12 years agoIgnore Emacs droppings and generated README
Dagfinn Ilmari Mannsåker [Fri, 24 Aug 2012 23:55:02 +0000]
Ignore Emacs droppings and generated README

12 years agorelease 0.08200 v0.08200
Arthur Axel 'fREW' Schmidt [Fri, 24 Aug 2012 02:37:13 +0000]
release 0.08200

12 years agoput off deprecation and clarify test output
Arthur Axel 'fREW' Schmidt [Fri, 24 Aug 2012 13:33:30 +0000]
put off deprecation and clarify test output

12 years agorewrite test to use existing schema
Arthur Axel 'fREW' Schmidt [Fri, 24 Aug 2012 02:34:58 +0000]
rewrite test to use existing schema

12 years agorelease 0.08199 v0.08199
Arthur Axel 'fREW' Schmidt [Thu, 23 Aug 2012 01:18:26 +0000]
release 0.08199

12 years agoDev release changes
Jess Robinson [Wed, 1 Aug 2012 18:26:13 +0000]
Dev release changes

12 years agoBack out constructor/prefetch rewrite introduced mainly by 43245ada4a
Peter Rabbitson [Tue, 24 Jul 2012 10:12:31 +0000]
Back out constructor/prefetch rewrite introduced mainly by 43245ada4a

It was shipped against the authors advice, while containing multiple known
bugs. After the expected bugreports went warnocked for over two weeks by the
new DBIC release team, it seems that the only way to partially restore the
release quality DBIC users have come to expect, is to currently throw this
code away until better times.

Should resolve RT#78456 and the issues reported in these threads:
http://lists.scsys.co.uk/pipermail/dbix-class/2012-July/010681.html
http://lists.scsys.co.uk/pipermail/dbix-class/2012-July/010682.html

12 years agofailing test for prefetch bug where prefetched values of 0 might break inflation
Eden Cardim [Wed, 22 Aug 2012 19:22:53 +0000]
failing test for prefetch bug where prefetched values of 0 might break inflation

12 years agoremove ->deploy from castaway's test
Rafael Kitover [Thu, 2 Aug 2012 14:22:18 +0000]
remove ->deploy from castaway's test

12 years agoNew test for recent prefetch/undef bug
Jess Robinson [Wed, 1 Aug 2012 22:16:07 +0000]
New test for recent prefetch/undef bug

12 years ago Fix Class::AccessorGroup to allow instances in get_component_class.
Gareth Kirwan [Tue, 17 Jul 2012 15:49:50 +0000]
 Fix Class::AccessorGroup to allow instances in get_component_class.
 Fixes result_class($obj) exploding instead of just using the object.

12 years agorelease v0.08198 v0.08198
Arthur Axel 'fREW' Schmidt [Wed, 11 Jul 2012 02:55:05 +0000]
release v0.08198

12 years agofix silent Oracle connection failures
Rafael Kitover [Sun, 24 Jun 2012 17:42:46 +0000]
fix silent Oracle connection failures

::Storage::DBI::_server_info was eating connection exceptions by
ignoring exceptions from $dbh->get_info, but this information is
important for _rebless when e.g. determining which class to use based on
the DB server version, so rethrow the exceptions when
$self->{_in_determine_driver} is true.

12 years agobetter type check for mssql+ado binary null strip
Rafael Kitover [Sun, 24 Jun 2012 06:59:11 +0000]
better type check for mssql+ado binary null strip

The data_type in the t/lib result class for the blob for the tests
changed to 'blob', while the binary type null stripper only expected
'binary' or 'image' data types, which is wrong for possibly
cross-deployed schemas.

Change _strip_trailing_binary_nulls in ::ADO::CursorUtils to use the
newly added and more generic ::Storage::DBI::_is_binary_type instead of
just checking for 'binary' or 'image'.

12 years agomark some tests in admin/10script.t on Win32 TODO
Rafael Kitover [Sun, 24 Jun 2012 05:32:03 +0000]
mark some tests in admin/10script.t on Win32 TODO

The test_dbicadmin() tests are failing on Win32 due to some sort of
quoting issue with the temp directory. I'm not sure how the temp
directory is being used here, so it will have to wait for someone else,
in the meanwhile I marked the tests TODO.

Also changed the " backwhacking to use Win32::ShellQuote instead and
added it to optdeps, unfortunately that didn't fix the problem, since
the problem seems to be some sort of DBICTest::RunMode interaction with
dbicadmin.

12 years agopreload threads.pm in storage/txn.t for fork error
Rafael Kitover [Mon, 18 Jun 2012 15:49:37 +0000]
preload threads.pm in storage/txn.t for fork error

SineSwiper reported this error in t/storage/txn.t on Strawberry Perl
5.12 32bit:

You must load the 'threads' module before Moo::_Utils at t/storage/txn.t
line 130.

Preload threads in a BEGIN block at the beginning of the .t file for
Win32. This fixes the error.

12 years agoskip 51threadnodb.t on Win32 on < 5.14
Rafael Kitover [Mon, 18 Jun 2012 15:11:36 +0000]
skip 51threadnodb.t on Win32 on < 5.14

SineSwiper reported a crash with a dialog from "make test" in
t/51threadnodb.t on Strawberry Perl 5.12 32bit.

The test passes on my ActivePerl 64bit 5.14.2 and there is a passing
test report for 5.14 here:
http://www.cpantesters.org/cpan/report/239d2ae4-87c3-1015-ad08-a09405085451

Disable test on Win32 for perls below 5.14 unless the environment
variable TEST_AUTHOR is set to be safe.

12 years agorelease 0.08197 v0.08197
Arthur Axel 'fREW' Schmidt [Tue, 10 Jul 2012 21:35:54 +0000]
release 0.08197

12 years agono longer dep on dev rel since this is not a dev rel
Arthur Axel 'fREW' Schmidt [Tue, 10 Jul 2012 21:33:52 +0000]
no longer dep on dev rel since this is not a dev rel

12 years agofix pod for unicode name
Arthur Axel 'fREW' Schmidt [Tue, 10 Jul 2012 21:28:49 +0000]
fix pod for unicode name

12 years agoAdd Travis YML for Travis-CI support
Brendan Byrd [Wed, 20 Jun 2012 15:27:04 +0000]
Add Travis YML for Travis-CI support
   (See http://travis-ci.org/#!/SineSwiper/dbix-class/jobs/1665376)

12 years agofix broken syntax
Justin Hunter [Fri, 15 Jun 2012 19:34:19 +0000]
fix broken syntax

12 years agoAdd authority for new moules
Peter Rabbitson [Fri, 15 Jun 2012 15:28:05 +0000]
Add authority for new moules

12 years agofix tests on solaris
Arthur Axel 'fREW' Schmidt [Fri, 15 Jun 2012 01:20:00 +0000]
fix tests on solaris

12 years agofix debian mssql/odbc setup docs for multiarch
Rafael Kitover [Tue, 5 Jun 2012 01:40:03 +0000]
fix debian mssql/odbc setup docs for multiarch

12 years agorelease 0.08197_01
Arthur Axel 'fREW' Schmidt [Wed, 30 May 2012 02:12:09 +0000]
release 0.08197_01

12 years agofix for RT#77241
Justin Hunter [Wed, 23 May 2012 13:31:03 +0000]
fix for RT#77241

12 years agofix typo
Justin Hunter [Tue, 15 May 2012 22:59:13 +0000]
fix typo

12 years agoadd more docs on software_limit resultset attr
Rafael Kitover [Tue, 15 May 2012 12:02:48 +0000]
add more docs on software_limit resultset attr

Add a paragraph to Storage/DBI/Sybase/ASE and a section under SEARCHING
to Cookbook.

12 years agoupdate proxy/cascade_update doc in Relationship::Base
Gerda Shank [Tue, 8 May 2012 16:22:42 +0000]
update proxy/cascade_update doc in Relationship::Base

12 years agoRemove mention of obsolete DBIx::Class::WebForm.
Dagfinn Ilmari Mannsåker [Sun, 6 May 2012 16:22:44 +0000]
Remove mention of obsolete DBIx::Class::WebForm.

Mention HTML::FormHandler::Model::DBIC instead, which is actively
maintained.

12 years agoRemove/fix bogus defaults values in example
Dagfinn Ilmari Mannsåker [Sun, 6 May 2012 16:19:13 +0000]
Remove/fix bogus defaults values in example

is_auto_increment already provides a default, and for the other columns
we either don't want a default value or it needs to be an integer.
Explicit is_auto_increment => 0 is also pointless and distracting.

12 years agoAdjust tests for new Moo
Peter Rabbitson [Fri, 27 Apr 2012 07:39:44 +0000]
Adjust tests for new Moo

12 years agoSimplify untainting - Module::Find can't return invalid names
Peter Rabbitson [Fri, 27 Apr 2012 07:09:15 +0000]
Simplify untainting - Module::Find can't return invalid names

12 years agoUse the now pure-perlized Devel::GlobalDestroy module
Peter Rabbitson [Fri, 27 Apr 2012 07:00:42 +0000]
Use the now pure-perlized Devel::GlobalDestroy module

12 years agoFix more fallout from 9b871b00, centralize the ugly in a base Schema class
Peter Rabbitson [Mon, 23 Apr 2012 16:41:28 +0000]
Fix more fallout from 9b871b00, centralize the ugly in a base Schema class

12 years agoRemove realiasing overengineering introduced in 86bb5a27da
Peter Rabbitson [Sat, 21 Apr 2012 04:57:41 +0000]
Remove realiasing overengineering introduced in 86bb5a27da

mst is right - there is no viable use case for this, cleanse with fire
Also I managed to misspell subUery... twice

12 years agoMore limit torture
Peter Rabbitson [Sat, 21 Apr 2012 04:33:11 +0000]
More limit torture

12 years agoFix lapse on limited incomplete has_many prefetches
Peter Rabbitson [Sat, 21 Apr 2012 04:15:33 +0000]
Fix lapse on limited incomplete has_many prefetches

12 years agoDo not artificially order the internals of a has_many prefetch subquery
Peter Rabbitson [Sat, 21 Apr 2012 04:03:34 +0000]
Do not artificially order the internals of a has_many prefetch subquery

Thre is no point using the artificial order which is only needed for efficient
collapse. The user didn't request an order - this means any rows will do

12 years agoSaner naming for helper test method
Peter Rabbitson [Sat, 21 Apr 2012 01:32:04 +0000]
Saner naming for helper test method

12 years agoFix test failure on MSWin32
Peter Rabbitson [Fri, 20 Apr 2012 07:41:28 +0000]
Fix test failure on MSWin32

12 years agoMake sure order realiasing remains in proper sequence on sorting
Peter Rabbitson [Fri, 20 Apr 2012 07:33:22 +0000]
Make sure order realiasing remains in proper sequence on sorting

We now support 999 realiased order criteria, this is beyond sufficient

12 years agoFix test borkage after 9b871b00c - ALWAYS RUN TESTS RETARDO!
Peter Rabbitson [Fri, 20 Apr 2012 07:29:34 +0000]
Fix test borkage after 9b871b00c - ALWAYS RUN TESTS RETARDO!

12 years agoBetter handling of bogus tmpdir returned by File::Spec on MSWin32
Peter Rabbitson [Wed, 18 Apr 2012 16:16:22 +0000]
Better handling of bogus tmpdir returned by File::Spec on MSWin32

12 years agoadd missing catfile so tests can run on win32
Arthur Axel 'fREW' Schmidt [Wed, 18 Apr 2012 13:46:20 +0000]
add missing catfile so tests can run on win32

12 years agoVersion/dist Makefile.PL safety checks
Peter Rabbitson [Mon, 16 Apr 2012 04:47:47 +0000]
Version/dist Makefile.PL safety checks

12 years agoAdd strictures to the Makefile.PL inc-snippets
Peter Rabbitson [Mon, 16 Apr 2012 04:42:01 +0000]
Add strictures to the Makefile.PL inc-snippets

12 years agoThis is stupid - no need to build the SQL manually, the local flag is enough
Peter Rabbitson [Mon, 16 Apr 2012 03:16:47 +0000]
This is stupid - no need to build the SQL manually, the local flag is enough