Merge 'trunk' into 'diamond_relationships'

Merge 'trunk' into 'diamond_relationships'

r6164@mullet (orig r6163):  ribasushi | 2009-05-07 10:09:01 -0700
 r6115@Thesaurus (orig r6114):  plu | 2009-05-03 10:39:16 +0200
 new branch to fix $rs->update and $rs->delete using the new as_query method

 r6116@Thesaurus (orig r6115):  plu | 2009-05-03 10:52:07 +0200
 Methods update/delete on resultset use now new as_query method to updated/delete properly on joined/prefetched resultset using a subquery. Therefore some tests have been added and some have been changed as well as the warnings around $rs->update/delete have been removed. Cheers!
 r6117@Thesaurus (orig r6116):  plu | 2009-05-03 11:13:48 +0200
 Using "is" instead of "cmp_ok"
 r6160@Thesaurus (orig r6159):  ribasushi | 2009-05-07 11:58:14 +0200
 Back out skip_parens support in as_query
 r6161@Thesaurus (orig r6160):  ribasushi | 2009-05-07 19:00:48 +0200
 This test is completely borked, needs a rewrite
 r6162@Thesaurus (orig r6161):  ribasushi | 2009-05-07 19:07:19 +0200
 Temporary fix or the IN ( ( ... ) ) problem until we get proper SQLA AST (needs SQLA released with commit 6158 to work)

r6165@mullet (orig r6164):  ribasushi | 2009-05-07 10:11:46 -0700
Changes, remove merged branch
r6169@mullet (orig r6168):  ribasushi | 2009-05-07 10:24:54 -0700
Bump SQLA dependency so -in/-between workarounds overload properly
r6172@mullet (orig r6171):  ribasushi | 2009-05-07 11:49:26 -0700
Cookbook cleanup
r6174@mullet (orig r6173):  ribasushi | 2009-05-08 01:13:30 -0700
Throw away some debugging code
r6175@mullet (orig r6174):  ribasushi | 2009-05-08 01:21:53 -0700
Documentation patch by nniuq
r6176@mullet (orig r6175):  plu | 2009-05-08 01:30:20 -0700
Set NLS_LANG so we  have a predictable date format when using MON
r6177@mullet (orig r6176):  ribasushi | 2009-05-08 03:15:15 -0700
Fix POD
r6179@mullet (orig r6178):  jgoulah | 2009-05-08 07:27:49 -0700
renaming rh performance test so it will show up at the end of test output
r6195@mullet (orig r6194):  caelum | 2009-05-09 06:46:55 -0700
added postgres default port stuff to FAQ
r6199@mullet (orig r6198):  mo | 2009-05-10 08:37:16 -0700
set_$rel accepts now a $link_vals hashref like add_to_$rel does
r6202@mullet (orig r6201):  ribasushi | 2009-05-10 10:57:31 -0700
Require DBICTest.pm in all tests even if it is not needed at all
r6203@mullet (orig r6202):  ribasushi | 2009-05-10 11:17:09 -0700
No more 'I forgot to run perl Makefile.PL'
r6204@mullet (orig r6203):  ribasushi | 2009-05-10 12:35:03 -0700
Switch the m2m method warnings from warnings::register to $ENV{DBIC_METHOD_CLOBBER_OK} = 1
r6206@mullet (orig r6205):  ribasushi | 2009-05-11 03:50:54 -0700
Change the makefile to test r6202
r6218@mullet (orig r6217):  ribasushi | 2009-05-11 22:02:14 -0700
Trying to untangle failing merge (whitespace change)
r6220@mullet (orig r6219):  ribasushi | 2009-05-11 22:02:56 -0700
 r5757@Thesaurus (orig r5756):  arcanez | 2009-03-14 14:55:55 +0100
 created count_distinct branch
 r5758@Thesaurus (orig r5757):  arcanez | 2009-03-14 15:33:03 +0100
  * change count with group_by (distinct) to use a subquery
  * rewrite of _bind_to_sql (uses placeholders and bindvars)
  * tests for count distinct
  * fixed tests for from subquery

 r5760@Thesaurus (orig r5759):  arcanez | 2009-03-16 16:48:28 +0100
 don't remove the where clause unless we're doing distinct, it needs to be there
 r5850@Thesaurus (orig r5849):  arcanez | 2009-03-30 21:40:05 +0200
  * add more tests
  * remove old cruft
  * remove old note

 r6035@Thesaurus (orig r6034):  ribasushi | 2009-04-30 09:10:36 +0200
 Add joined count test
 r6079@Thesaurus (orig r6078):  arcanez | 2009-05-01 08:41:34 +0200
 cleanup/fix some broken tests
 r6097@Thesaurus (orig r6096):  arcanez | 2009-05-01 19:37:04 +0200
 make sure merge bind
 test for aformentioned
 TODO count_joined test for a little while

 r6105@Thesaurus (orig r6104):  arcanez | 2009-05-02 02:33:49 +0200
 remove hackish ways

 r6106@Thesaurus (orig r6105):  arcanez | 2009-05-02 03:20:04 +0200
 more fixes to tests
 r6107@Thesaurus (orig r6106):  arcanez | 2009-05-02 03:33:47 +0200
 remove DBIx::Class::Storage::DBI::MultiDistinctEmulation
 r6111@Thesaurus (orig r6110):  ribasushi | 2009-05-03 02:00:19 +0200
 Failing test without immediate fixes go to branches, not to trunk
 r6114@Thesaurus (orig r6113):  ribasushi | 2009-05-03 10:23:28 +0200
 Bump SQLA ependencies so parenthesis_significant is guaranteed to be there
 r6148@Thesaurus (orig r6147):  ribasushi | 2009-05-06 17:40:31 +0200
 Add subquery/from test by michaelr (copied from subquery branch r5742)
 r6151@Thesaurus (orig r6150):  ribasushi | 2009-05-06 17:56:07 +0200
 TODOify sqla-dependent tests
 r6152@Thesaurus (orig r6151):  arcanez | 2009-05-06 18:33:58 +0200
 fix old test using new bind vars (no more interpolating)
 r6154@Thesaurus (orig r6153):  ribasushi | 2009-05-07 00:56:40 +0200
 Rename internal function to clarify what it does
 r6171@Thesaurus (orig r6170):  ribasushi | 2009-05-07 19:53:30 +0200
 Adjust tests for the IN fixes
 r6187@Thesaurus (orig r6186):  ribasushi | 2009-05-08 23:11:43 +0200
 Final count tests
 r6188@Thesaurus (orig r6187):  arcanez | 2009-05-09 03:50:12 +0200
 rewrite DISTINCT/COUNT(DISTINCT) Cookbook entries
 r6213@Thesaurus (orig r6212):  arcanez | 2009-05-11 22:41:21 +0200
 warn/die based on { select => { distinct => { } } }
 r6214@Thesaurus (orig r6213):  arcanez | 2009-05-11 23:21:11 +0200
 use carp instead of warn
 r6216@Thesaurus (orig r6215):  arcanez | 2009-05-11 23:45:05 +0200
 make sure we get just a string
 r6217@Thesaurus (orig r6216):  arcanez | 2009-05-11 23:52:11 +0200
 oops

r6222@mullet (orig r6221):  ribasushi | 2009-05-11 23:11:27 -0700
 r6112@Thesaurus (orig r6111):  nniuq | 2009-05-03 02:36:33 +0200
 Initially, fixes to enable saving of LOB types in Oracle.  Possibly timestamp tweaks.

 r6113@Thesaurus (orig r6112):  nniuq | 2009-05-03 03:52:29 +0200
 Support for saving CLOB and BLOB types in Oracle.

 r6118@Thesaurus (orig r6117):  nniuq | 2009-05-04 03:58:03 +0200
 Proper support for timestamp inflation.  Added last_updated_at to DBICTest::Schema::Track as a date by default, initialized in sqlite loader, redefined to timestamp for Oracle tests.

 r6119@Thesaurus (orig r6118):  nniuq | 2009-05-04 04:03:28 +0200
 Re-added last_updated_at to create table statement.

 r6134@Thesaurus (orig r6133):  nniuq | 2009-05-05 15:11:49 +0200
 Added self to contributors; clarified comment on :ora_types imports.

 r6182@Thesaurus (orig r6181):  nniuq | 2009-05-08 17:10:58 +0200
 Refactored to call _{inflate_to,deflate_from}_datetime through a _flate_or_fallback wrapper handling a parser's lack of support for the requested type.

 r6183@Thesaurus (orig r6182):  ribasushi | 2009-05-08 17:24:54 +0200
 Remove redundant var
 r6189@Thesaurus (orig r6188):  nniuq | 2009-05-09 03:59:20 +0200
 Changed test of lob values from is to ok on an eq expr, to avoid a huge got/expected diagnosis.  In doing so, discovered it was testing undef vs undef!  Whoops.  Fixed.

 r6210@Thesaurus (orig r6209):  nniuq | 2009-05-11 16:24:27 +0200
 Moved DBIC::SQL::Abstract inner classes to DBIx::Class::SQLAHacks namespace to decouple 41orrible.t tests from use of DBD::Oracle in Oracle Generic driver.

r6224@mullet (orig r6223):  ribasushi | 2009-05-11 23:13:31 -0700
 r6223@Thesaurus (orig r6222):  ribasushi | 2009-05-12 08:12:50 +0200
 Whops, forgotten post-merge tweaks

r6226@mullet (orig r6225):  ribasushi | 2009-05-11 23:43:58 -0700
Cleanup
r6227@mullet (orig r6226):  ribasushi | 2009-05-12 00:42:45 -0700
Shuffle delete tests, and sanify the delete related ones
r6228@mullet (orig r6227):  ribasushi | 2009-05-12 00:54:37 -0700
Rename m2m warn disable envvar
r6229@mullet (orig r6228):  ribasushi | 2009-05-12 01:20:08 -0700
Switch warn to carp and die to throw_exception where possible
r6230@mullet (orig r6229):  bricas | 2009-05-12 04:31:35 -0700
added info to Changes from my DateTime::InflateColumn modifications
r6231@mullet (orig r6230):  nniuq | 2009-05-12 06:08:14 -0700
Added synopsis of merged oracle-tweaks branch to Changes
r6233@mullet (orig r6232):  debolaz | 2009-05-12 08:17:09 -0700
Test for failing code related to many to many in MySQL

r6234@mullet (orig r6233):  arcanez | 2009-05-12 08:33:56 -0700
Changes for count_distinct branch merge
r6235@mullet (orig r6234):  ribasushi | 2009-05-12 09:09:14 -0700
Some more die/croak conversions
r6236@mullet (orig r6235):  ribasushi | 2009-05-12 09:51:38 -0700
Fix shot test
r6244@mullet (orig r6243):  ribasushi | 2009-05-12 18:02:11 -0700
fix bogus test
r6245@mullet (orig r6244):  ribasushi | 2009-05-12 18:06:42 -0700
Fix _select_for_update/delete - bring back old code, use subqueries only when resultset attributes call for it
r6252@mullet (orig r6251):  ribasushi | 2009-05-13 16:41:31 -0700
Better testing and heuristics for subqueried _cond_for_update_delete
r6253@mullet (orig r6252):  ribasushi | 2009-05-13 16:54:41 -0700
join can be found without seen_join