15 years agoApparent fix - simply delay the in_storage flagging of the main object until all...
Peter Rabbitson [Thu, 2 Jul 2009 22:14:09 +0000]
Apparent fix - simply delay the in_storage flagging of the main object until all prefetched objects are inflated. The rest of the changes are just cosmetics, preparing for the collapse_result rewrite

15 years agoMerge 'trunk' into 'unresolvable_prefetch'
Peter Rabbitson [Thu, 2 Jul 2009 20:55:02 +0000]
Merge 'trunk' into 'unresolvable_prefetch'

r6910@Thesaurus (orig r6909):  ribasushi | 2009-07-01 13:27:15 +0200
Optimize set_column on uninserted objects
r6921@Thesaurus (orig r6920):  caelum | 2009-07-01 17:40:32 +0200
 r5859@hlagh (orig r6912):  caelum | 2009-07-01 06:21:30 -0700
 new connected() for dbd::sybase users
 r5860@hlagh (orig r6913):  caelum | 2009-07-01 06:25:46 -0700
 add a couple of dbd::sybase reconnection tests
 r5861@hlagh (orig r6914):  caelum | 2009-07-01 06:35:07 -0700
 better connection test
 r5862@hlagh (orig r6915):  caelum | 2009-07-01 06:45:05 -0700
 use dbh->do for connected instead of prepare_cached
 r5863@hlagh (orig r6916):  ribasushi | 2009-07-01 06:55:21 -0700
 Segfault
 r5864@hlagh (orig r6917):  caelum | 2009-07-01 07:03:22 -0700
 use ->do instead of ->prepare_cached in oracle's connected() too
 r5865@hlagh (orig r6918):  caelum | 2009-07-01 08:20:52 -0700
 fix segfault with old DBD::Sybase
 r5866@hlagh (orig r6919):  caelum | 2009-07-01 08:39:18 -0700
 move connection tests into _ping()

r6924@Thesaurus (orig r6923):  ijw | 2009-07-01 19:34:32 +0200
Added a test for a resultset to related-resultset join for 0 related records
r6928@Thesaurus (orig r6927):  ijw | 2009-07-01 20:04:16 +0200
Additional tests on prefetch - illustrates the bug with left-join has_many (NULL row returned) and the one that results from the trivial fix (prefetch gives no artist)
r6932@Thesaurus (orig r6931):  ribasushi | 2009-07-02 08:08:33 +0200
Another candidate for somethingawful.com (fix left join-ed count)
r6934@Thesaurus (orig r6933):  ribasushi | 2009-07-02 09:04:13 +0200
Changelog
r6935@Thesaurus (orig r6934):  ribasushi | 2009-07-02 11:23:48 +0200
cleanup
r6936@Thesaurus (orig r6935):  ijw | 2009-07-02 12:41:01 +0200
Check fetched rows == count for related resultsets
r6937@Thesaurus (orig r6936):  ijw | 2009-07-02 12:43:47 +0200
Confirm prefetch doesn't affect main row fetch, and main row fetch works with and without counting
r6938@Thesaurus (orig r6937):  ribasushi | 2009-07-02 12:52:51 +0200
More fail (fix is known but needs work)
r6939@Thesaurus (orig r6938):  ribasushi | 2009-07-02 13:07:22 +0200
And more fail
r6940@Thesaurus (orig r6939):  ribasushi | 2009-07-02 13:16:46 +0200
These tests are in prefetch/count.t
r6941@Thesaurus (orig r6940):  ribasushi | 2009-07-02 13:38:31 +0200
cleanup
r6942@Thesaurus (orig r6941):  ribasushi | 2009-07-02 13:38:49 +0200
Solve more prefetch inflation crap
r6943@Thesaurus (orig r6942):  ribasushi | 2009-07-02 13:47:41 +0200
Make the code readable
r6944@Thesaurus (orig r6943):  ribasushi | 2009-07-02 15:52:35 +0200
Everything works, just need to fix join-path chaining over search_related (to guard against obscure db quirks)
r6946@Thesaurus (orig r6945):  caelum | 2009-07-02 21:06:32 +0200
add sybase reconnect test
r6948@Thesaurus (orig r6947):  ribasushi | 2009-07-02 22:20:21 +0200
Last part of the join handling puzzle

15 years agoLast part of the join handling puzzle
Peter Rabbitson [Thu, 2 Jul 2009 20:20:21 +0000]
Last part of the join handling puzzle

15 years agoMerge 'trunk' into 'sybase'
Rafael Kitover [Thu, 2 Jul 2009 17:06:32 +0000]
Merge 'trunk' into 'sybase'

15 years agoadd sybase reconnect test
Rafael Kitover [Thu, 2 Jul 2009 19:06:32 +0000]
add sybase reconnect test

15 years agoEverything works, just need to fix join-path chaining over search_related (to guard...
Peter Rabbitson [Thu, 2 Jul 2009 13:52:35 +0000]
Everything works, just need to fix join-path chaining over search_related (to guard against obscure db quirks)

15 years agoMake the code readable
Peter Rabbitson [Thu, 2 Jul 2009 11:47:41 +0000]
Make the code readable

15 years agoSolve more prefetch inflation crap
Peter Rabbitson [Thu, 2 Jul 2009 11:38:49 +0000]
Solve more prefetch inflation crap

15 years agocleanup
Peter Rabbitson [Thu, 2 Jul 2009 11:38:31 +0000]
cleanup

15 years agoThese tests are in prefetch/count.t
Peter Rabbitson [Thu, 2 Jul 2009 11:16:46 +0000]
These tests are in prefetch/count.t

15 years agoAnd more fail
Peter Rabbitson [Thu, 2 Jul 2009 11:07:22 +0000]
And more fail

15 years agoMore fail (fix is known but needs work)
Peter Rabbitson [Thu, 2 Jul 2009 10:52:51 +0000]
More fail (fix is known but needs work)

15 years agoConfirm prefetch doesn't affect main row fetch, and main row fetch works with and...
Ian Wells [Thu, 2 Jul 2009 10:43:47 +0000]
Confirm prefetch doesn't affect main row fetch, and main row fetch works with and without counting

15 years agoCheck fetched rows == count for related resultsets
Ian Wells [Thu, 2 Jul 2009 10:41:01 +0000]
Check fetched rows == count for related resultsets

15 years agocleanup
Peter Rabbitson [Thu, 2 Jul 2009 09:23:48 +0000]
cleanup

15 years agoChangelog
Peter Rabbitson [Thu, 2 Jul 2009 07:04:13 +0000]
Changelog

15 years agoAnother candidate for somethingawful.com (fix left join-ed count)
Peter Rabbitson [Thu, 2 Jul 2009 06:08:33 +0000]
Another candidate for somethingawful.com (fix left join-ed count)

15 years agoMerge 'trunk' into 'mssql_top_fixes'
Peter Rabbitson [Thu, 2 Jul 2009 05:43:30 +0000]
Merge 'trunk' into 'mssql_top_fixes'

r6875@Thesaurus (orig r6874):  ribasushi | 2009-06-30 12:39:06 +0200
Allow broken resultsource-class-derived objects to still work
r6876@Thesaurus (orig r6875):  ribasushi | 2009-06-30 12:40:46 +0200
clarify
r6878@Thesaurus (orig r6877):  ash | 2009-06-30 13:48:13 +0200
Update POD on Dynamic sub-classing

r6883@Thesaurus (orig r6882):  ribasushi | 2009-06-30 17:36:38 +0200
 r6815@Thesaurus (orig r6814):  ribasushi | 2009-06-28 10:32:42 +0200
 Branch to explore double joins on search_related
 r6816@Thesaurus (orig r6815):  ribasushi | 2009-06-28 10:34:16 +0200
 Thetest case that started it all
 r6817@Thesaurus (orig r6816):  ribasushi | 2009-06-28 10:35:11 +0200
 The proposed fix (do not add an extra join if it is already present in the topmost join)
 r6818@Thesaurus (orig r6817):  ribasushi | 2009-06-28 11:04:26 +0200
 Minor omission
 r6819@Thesaurus (orig r6818):  ribasushi | 2009-06-28 11:07:33 +0200
 Adjust a couple of tests for new behavior (thus all of this might be backwards incompatible to the point of being useless):
 The counts in t/90join_torture.t are now 5*3, not 5*3*3, as a second join is not induced by search_related
 The raw sql scan in t/prefetch/standard.t is just silly, won't even try to understand it
 Just to maintain the TreeLike folding, I add a 3rd children join which was inserted by search_related before the code changes

r6889@Thesaurus (orig r6888):  ribasushi | 2009-06-30 19:36:11 +0200
Todoify test for now
r6890@Thesaurus (orig r6889):  ribasushi | 2009-06-30 19:37:05 +0200
Todoify test for now (2)
r6892@Thesaurus (orig r6891):  ribasushi | 2009-06-30 19:52:31 +0200
Todoify test for now (3)
r6903@Thesaurus (orig r6902):  ribasushi | 2009-07-01 08:46:12 +0200
Fixed deadlock test
r6904@Thesaurus (orig r6903):  ribasushi | 2009-07-01 12:22:00 +0200
Clarify exception text
r6907@Thesaurus (orig r6906):  ribasushi | 2009-07-01 13:23:46 +0200
 r6821@Thesaurus (orig r6820):  ribasushi | 2009-06-28 13:09:11 +0200
 Branch for prefetch+group play
 r6823@Thesaurus (orig r6822):  ribasushi | 2009-06-28 14:38:36 +0200
 Normalize group_by
 r6824@Thesaurus (orig r6823):  ribasushi | 2009-06-28 14:39:54 +0200
 Proper prefetch+group test
 r6826@Thesaurus (orig r6825):  ribasushi | 2009-06-28 14:42:48 +0200
 Whoops
 r6828@Thesaurus (orig r6827):  ribasushi | 2009-06-28 15:06:57 +0200
 Lose the literal sql bits - castaway is right it's silly to support those
 r6833@Thesaurus (orig r6832):  ribasushi | 2009-06-28 22:38:43 +0200
 Rogue comments
 r6837@Thesaurus (orig r6836):  ribasushi | 2009-06-29 09:44:25 +0200
 A couple of test fixes
 r6838@Thesaurus (orig r6837):  ribasushi | 2009-06-29 09:46:13 +0200
 Support for -select/-as in SQLAHacks field selection
 r6839@Thesaurus (orig r6838):  ribasushi | 2009-06-29 09:49:53 +0200
 This is tested elsewhere
 r6840@Thesaurus (orig r6839):  ribasushi | 2009-06-29 09:50:43 +0200
 This is tested elsewhere (2)
 r6841@Thesaurus (orig r6840):  ribasushi | 2009-06-29 10:07:09 +0200
 Test cleanups
 r6842@Thesaurus (orig r6841):  ribasushi | 2009-06-29 10:11:13 +0200
 Most of the grouped prefetch solution
 r6843@Thesaurus (orig r6842):  ribasushi | 2009-06-29 10:14:45 +0200
 clearer
 r6845@Thesaurus (orig r6844):  ribasushi | 2009-06-29 12:05:37 +0200
 And score! (all works)
 r6882@Thesaurus (orig r6881):  ribasushi | 2009-06-30 16:23:06 +0200
 rs->get_column now properly recognizes prefetch and collapses if at all possible
 r6886@Thesaurus (orig r6885):  ribasushi | 2009-06-30 17:39:58 +0200
 Whoops

r6910@Thesaurus (orig r6909):  ribasushi | 2009-07-01 13:27:15 +0200
Optimize set_column on uninserted objects
r6921@Thesaurus (orig r6920):  caelum | 2009-07-01 17:40:32 +0200
 r5859@hlagh (orig r6912):  caelum | 2009-07-01 06:21:30 -0700
 new connected() for dbd::sybase users
 r5860@hlagh (orig r6913):  caelum | 2009-07-01 06:25:46 -0700
 add a couple of dbd::sybase reconnection tests
 r5861@hlagh (orig r6914):  caelum | 2009-07-01 06:35:07 -0700
 better connection test
 r5862@hlagh (orig r6915):  caelum | 2009-07-01 06:45:05 -0700
 use dbh->do for connected instead of prepare_cached
 r5863@hlagh (orig r6916):  ribasushi | 2009-07-01 06:55:21 -0700
 Segfault
 r5864@hlagh (orig r6917):  caelum | 2009-07-01 07:03:22 -0700
 use ->do instead of ->prepare_cached in oracle's connected() too
 r5865@hlagh (orig r6918):  caelum | 2009-07-01 08:20:52 -0700
 fix segfault with old DBD::Sybase
 r5866@hlagh (orig r6919):  caelum | 2009-07-01 08:39:18 -0700
 move connection tests into _ping()

r6924@Thesaurus (orig r6923):  ijw | 2009-07-01 19:34:32 +0200
Added a test for a resultset to related-resultset join for 0 related records
r6928@Thesaurus (orig r6927):  ijw | 2009-07-01 20:04:16 +0200
Additional tests on prefetch - illustrates the bug with left-join has_many (NULL row returned) and the one that results from the trivial fix (prefetch gives no artist)

15 years agoAdditional tests on prefetch - illustrates the bug with left-join has_many (NULL...
Ian Wells [Wed, 1 Jul 2009 18:04:16 +0000]
Additional tests on prefetch - illustrates the bug with left-join has_many (NULL row returned) and the one that results from the trivial fix (prefetch gives no artist)

15 years agoAdded a test for a resultset to related-resultset join for 0 related records
Ian Wells [Wed, 1 Jul 2009 17:34:32 +0000]
Added a test for a resultset to related-resultset join for 0 related records

15 years agoMerge 'syb_connected' into 'trunk'
Rafael Kitover [Wed, 1 Jul 2009 15:40:32 +0000]
Merge 'syb_connected' into 'trunk'

r5859@hlagh (orig r6912):  caelum | 2009-07-01 06:21:30 -0700
new connected() for dbd::sybase users
r5860@hlagh (orig r6913):  caelum | 2009-07-01 06:25:46 -0700
add a couple of dbd::sybase reconnection tests
r5861@hlagh (orig r6914):  caelum | 2009-07-01 06:35:07 -0700
better connection test
r5862@hlagh (orig r6915):  caelum | 2009-07-01 06:45:05 -0700
use dbh->do for connected instead of prepare_cached
r5863@hlagh (orig r6916):  ribasushi | 2009-07-01 06:55:21 -0700
Segfault
r5864@hlagh (orig r6917):  caelum | 2009-07-01 07:03:22 -0700
use ->do instead of ->prepare_cached in oracle's connected() too
r5865@hlagh (orig r6918):  caelum | 2009-07-01 08:20:52 -0700
fix segfault with old DBD::Sybase
r5866@hlagh (orig r6919):  caelum | 2009-07-01 08:39:18 -0700
move connection tests into _ping()

15 years agomove connection tests into _ping()
Rafael Kitover [Wed, 1 Jul 2009 15:39:18 +0000]
move connection tests into _ping()

15 years agofix segfault with old DBD::Sybase
Rafael Kitover [Wed, 1 Jul 2009 15:20:52 +0000]
fix segfault with old DBD::Sybase

15 years agouse ->do instead of ->prepare_cached in oracle's connected() too
Rafael Kitover [Wed, 1 Jul 2009 14:03:22 +0000]
use ->do instead of ->prepare_cached in oracle's connected() too

15 years agoSegfault
Peter Rabbitson [Wed, 1 Jul 2009 13:55:21 +0000]
Segfault

15 years agouse dbh->do for connected instead of prepare_cached
Rafael Kitover [Wed, 1 Jul 2009 13:45:05 +0000]
use dbh->do for connected instead of prepare_cached

15 years agobetter connection test
Rafael Kitover [Wed, 1 Jul 2009 13:35:07 +0000]
better connection test

15 years agoadd a couple of dbd::sybase reconnection tests
Rafael Kitover [Wed, 1 Jul 2009 13:25:46 +0000]
add a couple of dbd::sybase reconnection tests

15 years agonew connected() for dbd::sybase users
Rafael Kitover [Wed, 1 Jul 2009 13:21:30 +0000]
new connected() for dbd::sybase users

15 years agoOptimize set_column on uninserted objects
Peter Rabbitson [Wed, 1 Jul 2009 11:27:15 +0000]
Optimize set_column on uninserted objects

15 years agoMerge 'trunk' into 'unresolvable_prefetch'
Peter Rabbitson [Wed, 1 Jul 2009 11:25:42 +0000]
Merge 'trunk' into 'unresolvable_prefetch'

15 years agoMerge 'grouped_prefetch' into 'trunk'
Peter Rabbitson [Wed, 1 Jul 2009 11:23:46 +0000]
Merge 'grouped_prefetch' into 'trunk'

r6821@Thesaurus (orig r6820):  ribasushi | 2009-06-28 13:09:11 +0200
Branch for prefetch+group play
r6823@Thesaurus (orig r6822):  ribasushi | 2009-06-28 14:38:36 +0200
Normalize group_by
r6824@Thesaurus (orig r6823):  ribasushi | 2009-06-28 14:39:54 +0200
Proper prefetch+group test
r6826@Thesaurus (orig r6825):  ribasushi | 2009-06-28 14:42:48 +0200
Whoops
r6828@Thesaurus (orig r6827):  ribasushi | 2009-06-28 15:06:57 +0200
Lose the literal sql bits - castaway is right it's silly to support those
r6833@Thesaurus (orig r6832):  ribasushi | 2009-06-28 22:38:43 +0200
Rogue comments
r6837@Thesaurus (orig r6836):  ribasushi | 2009-06-29 09:44:25 +0200
A couple of test fixes
r6838@Thesaurus (orig r6837):  ribasushi | 2009-06-29 09:46:13 +0200
Support for -select/-as in SQLAHacks field selection
r6839@Thesaurus (orig r6838):  ribasushi | 2009-06-29 09:49:53 +0200
This is tested elsewhere
r6840@Thesaurus (orig r6839):  ribasushi | 2009-06-29 09:50:43 +0200
This is tested elsewhere (2)
r6841@Thesaurus (orig r6840):  ribasushi | 2009-06-29 10:07:09 +0200
Test cleanups
r6842@Thesaurus (orig r6841):  ribasushi | 2009-06-29 10:11:13 +0200
Most of the grouped prefetch solution
r6843@Thesaurus (orig r6842):  ribasushi | 2009-06-29 10:14:45 +0200
clearer
r6845@Thesaurus (orig r6844):  ribasushi | 2009-06-29 12:05:37 +0200
And score! (all works)
r6882@Thesaurus (orig r6881):  ribasushi | 2009-06-30 16:23:06 +0200
rs->get_column now properly recognizes prefetch and collapses if at all possible
r6886@Thesaurus (orig r6885):  ribasushi | 2009-06-30 17:39:58 +0200
Whoops

15 years agoExtend test
Peter Rabbitson [Wed, 1 Jul 2009 10:54:03 +0000]
Extend test

15 years agoClarify exception text
Peter Rabbitson [Wed, 1 Jul 2009 10:22:00 +0000]
Clarify exception text

15 years agoFixed deadlock test
Peter Rabbitson [Wed, 1 Jul 2009 06:46:12 +0000]
Fixed deadlock test

15 years agosilly misspells and trailing whitespace
Arthur Axel "fREW" Schmidt [Tue, 30 Jun 2009 19:38:26 +0000]
silly misspells and trailing whitespace

15 years agoTodoify test for now (3)
Peter Rabbitson [Tue, 30 Jun 2009 17:52:31 +0000]
Todoify test for now (3)

15 years agoTodoify test for now (2)
Peter Rabbitson [Tue, 30 Jun 2009 17:37:05 +0000]
Todoify test for now (2)

15 years agoTodoify test for now
Peter Rabbitson [Tue, 30 Jun 2009 17:36:11 +0000]
Todoify test for now

15 years agoWhoops
Peter Rabbitson [Tue, 30 Jun 2009 15:39:58 +0000]
Whoops

15 years agoMerge 'trunk' into 'grouped_prefetch'
Peter Rabbitson [Tue, 30 Jun 2009 15:39:11 +0000]
Merge 'trunk' into 'grouped_prefetch'

r6844@Thesaurus (orig r6843):  abraxxa | 2009-06-29 11:02:17 +0200
fixed typo in test

r6848@Thesaurus (orig r6847):  ribasushi | 2009-06-29 19:09:00 +0200
Minor Ordered optimization (don't use count)
r6856@Thesaurus (orig r6855):  caelum | 2009-06-29 23:42:11 +0200
 r5451@hlagh (orig r6605):  caelum | 2009-06-10 09:23:44 -0700
 new branch to implement on_connect_call
 r5484@hlagh (orig r6633):  caelum | 2009-06-11 11:03:10 -0700
 on_connect_call implementation and set_datetime_format support for Oracle
 r5492@hlagh (orig r6641):  caelum | 2009-06-11 16:39:28 -0700
 connect_call_set_datetime_format for Oracle, I have no idea why this didn't get committed before...
 r5504@hlagh (orig r6655):  caelum | 2009-06-12 17:28:06 -0700
 finished up on_connect_call stuff
 r5507@hlagh (orig r6658):  caelum | 2009-06-13 04:03:36 -0700
 fixup _setup_connect_do, other minor cleanups
 r5508@hlagh (orig r6659):  caelum | 2009-06-13 04:35:33 -0700
 make the on_(dis)?connect_do accessors returnn the original structure
 r5509@hlagh (orig r6660):  caelum | 2009-06-13 08:31:52 -0700
 allow undef for _setup_connect_do
 r5522@hlagh (orig r6679):  caelum | 2009-06-14 09:56:40 -0700
 rename connect_do store
 r5621@hlagh (orig r6769):  caelum | 2009-06-23 07:38:33 -0700
 minor doc update
 r5628@hlagh (orig r6777):  caelum | 2009-06-23 16:36:12 -0700
 properly test nanosecond precision with oracle and datetime_setup
 r5669@hlagh (orig r6784):  caelum | 2009-06-24 10:49:25 -0700
 IC::DT does support timestamp with timezone
 r5768@hlagh (orig r6846):  caelum | 2009-06-29 08:20:32 -0700
 remove DateTime from 73oracle.t
 r5781@hlagh (orig r6849):  caelum | 2009-06-29 13:07:43 -0700
 remove the _store stuff for on_connect_do
 r5785@hlagh (orig r6853):  ribasushi | 2009-06-29 14:38:30 -0700
 Some beautification

r6871@Thesaurus (orig r6870):  ribasushi | 2009-06-30 10:09:03 +0200
Cleanup dependency handling a bit
r6875@Thesaurus (orig r6874):  ribasushi | 2009-06-30 12:39:06 +0200
Allow broken resultsource-class-derived objects to still work
r6876@Thesaurus (orig r6875):  ribasushi | 2009-06-30 12:40:46 +0200
clarify
r6878@Thesaurus (orig r6877):  ash | 2009-06-30 13:48:13 +0200
Update POD on Dynamic sub-classing

r6883@Thesaurus (orig r6882):  ribasushi | 2009-06-30 17:36:38 +0200
 r6815@Thesaurus (orig r6814):  ribasushi | 2009-06-28 10:32:42 +0200
 Branch to explore double joins on search_related
 r6816@Thesaurus (orig r6815):  ribasushi | 2009-06-28 10:34:16 +0200
 Thetest case that started it all
 r6817@Thesaurus (orig r6816):  ribasushi | 2009-06-28 10:35:11 +0200
 The proposed fix (do not add an extra join if it is already present in the topmost join)
 r6818@Thesaurus (orig r6817):  ribasushi | 2009-06-28 11:04:26 +0200
 Minor omission
 r6819@Thesaurus (orig r6818):  ribasushi | 2009-06-28 11:07:33 +0200
 Adjust a couple of tests for new behavior (thus all of this might be backwards incompatible to the point of being useless):
 The counts in t/90join_torture.t are now 5*3, not 5*3*3, as a second join is not induced by search_related
 The raw sql scan in t/prefetch/standard.t is just silly, won't even try to understand it
 Just to maintain the TreeLike folding, I add a 3rd children join which was inserted by search_related before the code changes

15 years agoMerge 'search_related_prefetch' into 'trunk'
Peter Rabbitson [Tue, 30 Jun 2009 15:36:38 +0000]
Merge 'search_related_prefetch' into 'trunk'

r6815@Thesaurus (orig r6814):  ribasushi | 2009-06-28 10:32:42 +0200
Branch to explore double joins on search_related
r6816@Thesaurus (orig r6815):  ribasushi | 2009-06-28 10:34:16 +0200
Thetest case that started it all
r6817@Thesaurus (orig r6816):  ribasushi | 2009-06-28 10:35:11 +0200
The proposed fix (do not add an extra join if it is already present in the topmost join)
r6818@Thesaurus (orig r6817):  ribasushi | 2009-06-28 11:04:26 +0200
Minor omission
r6819@Thesaurus (orig r6818):  ribasushi | 2009-06-28 11:07:33 +0200
Adjust a couple of tests for new behavior (thus all of this might be backwards incompatible to the point of being useless):
The counts in t/90join_torture.t are now 5*3, not 5*3*3, as a second join is not induced by search_related
The raw sql scan in t/prefetch/standard.t is just silly, won't even try to understand it
Just to maintain the TreeLike folding, I add a 3rd children join which was inserted by search_related before the code changes

15 years agors->get_column now properly recognizes prefetch and collapses if at all possible
Peter Rabbitson [Tue, 30 Jun 2009 14:23:06 +0000]
rs->get_column now properly recognizes prefetch and collapses if at all possible

15 years agoUpdate POD on Dynamic sub-classing
Ash Berlin [Tue, 30 Jun 2009 11:48:13 +0000]
Update POD on Dynamic sub-classing

15 years agocode cleanups
Rafael Kitover [Tue, 30 Jun 2009 10:46:43 +0000]
code cleanups

15 years agoclarify
Peter Rabbitson [Tue, 30 Jun 2009 10:40:46 +0000]
clarify

15 years agoAllow broken resultsource-class-derived objects to still work
Peter Rabbitson [Tue, 30 Jun 2009 10:39:06 +0000]
Allow broken resultsource-class-derived objects to still work

15 years agoFix borked next invocation
Peter Rabbitson [Tue, 30 Jun 2009 09:54:34 +0000]
Fix borked next invocation

15 years agoMerge 'trunk' into 'sybase'
Peter Rabbitson [Tue, 30 Jun 2009 15:09:03 +0000]
Merge 'trunk' into 'sybase'

r6871@Thesaurus (orig r6870):  ribasushi | 2009-06-30 10:09:03 +0200
Cleanup dependency handling a bit

15 years agoMerge 'trunk' into 'mssql_top_fixes'
Peter Rabbitson [Tue, 30 Jun 2009 15:09:03 +0000]
Merge 'trunk' into 'mssql_top_fixes'

15 years agoCleanup dependency handling a bit
Peter Rabbitson [Tue, 30 Jun 2009 08:09:03 +0000]
Cleanup dependency handling a bit

15 years agoSome fixes after review
Peter Rabbitson [Tue, 30 Jun 2009 07:53:27 +0000]
Some fixes after review

15 years agomerge on_connect_call updates
Rafael Kitover [Tue, 30 Jun 2009 01:39:51 +0000]
merge on_connect_call updates

15 years agoMerge 'trunk' into 'sybase'
Rafael Kitover [Tue, 30 Jun 2009 01:31:02 +0000]
Merge 'trunk' into 'sybase'

r5735@hlagh (orig r6812):  castaway | 2009-06-27 21:11:08 -0700
Tests for grouping with prefetch

r5742@hlagh (orig r6819):  ribasushi | 2009-06-28 04:00:03 -0700
The prefetch+group_by is a complex problem - branch
r5766@hlagh (orig r6843):  abraxxa | 2009-06-29 02:02:17 -0700
fixed typo in test

r5779@hlagh (orig r6847):  ribasushi | 2009-06-29 10:09:00 -0700
Minor Ordered optimization (don't use count)
r5787@hlagh (orig r6855):  caelum | 2009-06-29 14:42:11 -0700
 r5451@hlagh (orig r6605):  caelum | 2009-06-10 09:23:44 -0700
 new branch to implement on_connect_call
 r5484@hlagh (orig r6633):  caelum | 2009-06-11 11:03:10 -0700
 on_connect_call implementation and set_datetime_format support for Oracle
 r5492@hlagh (orig r6641):  caelum | 2009-06-11 16:39:28 -0700
 connect_call_set_datetime_format for Oracle, I have no idea why this didn't get committed before...
 r5504@hlagh (orig r6655):  caelum | 2009-06-12 17:28:06 -0700
 finished up on_connect_call stuff
 r5507@hlagh (orig r6658):  caelum | 2009-06-13 04:03:36 -0700
 fixup _setup_connect_do, other minor cleanups
 r5508@hlagh (orig r6659):  caelum | 2009-06-13 04:35:33 -0700
 make the on_(dis)?connect_do accessors returnn the original structure
 r5509@hlagh (orig r6660):  caelum | 2009-06-13 08:31:52 -0700
 allow undef for _setup_connect_do
 r5522@hlagh (orig r6679):  caelum | 2009-06-14 09:56:40 -0700
 rename connect_do store
 r5621@hlagh (orig r6769):  caelum | 2009-06-23 07:38:33 -0700
 minor doc update
 r5628@hlagh (orig r6777):  caelum | 2009-06-23 16:36:12 -0700
 properly test nanosecond precision with oracle and datetime_setup
 r5669@hlagh (orig r6784):  caelum | 2009-06-24 10:49:25 -0700
 IC::DT does support timestamp with timezone
 r5768@hlagh (orig r6846):  caelum | 2009-06-29 08:20:32 -0700
 remove DateTime from 73oracle.t
 r5781@hlagh (orig r6849):  caelum | 2009-06-29 13:07:43 -0700
 remove the _store stuff for on_connect_do
 r5785@hlagh (orig r6853):  ribasushi | 2009-06-29 14:38:30 -0700
 Some beautification

15 years agoMake sure storage classes use c3, just like the rest of dbic (tested on 5.8 as well)
Peter Rabbitson [Mon, 29 Jun 2009 23:13:49 +0000]
Make sure storage classes use c3, just like the rest of dbic (tested on 5.8 as well)

15 years agobranch pushed, removing
Rafael Kitover [Mon, 29 Jun 2009 21:45:19 +0000]
branch pushed, removing

15 years agoMerge 'on_connect_call' into 'trunk'
Rafael Kitover [Mon, 29 Jun 2009 21:42:11 +0000]
Merge 'on_connect_call' into 'trunk'

15 years agoSome beautification
Peter Rabbitson [Mon, 29 Jun 2009 21:38:30 +0000]
Some beautification

15 years agoMerge 'trunk' into 'mssql_top_fixes'
Peter Rabbitson [Mon, 29 Jun 2009 20:21:24 +0000]
Merge 'trunk' into 'mssql_top_fixes'

r6753@Thesaurus (orig r6752):  ribasushi | 2009-06-21 09:00:21 +0200
Clenaup text
r6754@Thesaurus (orig r6753):  ribasushi | 2009-06-21 14:37:56 +0200
make_column_dirty fix
r6756@Thesaurus (orig r6755):  ribasushi | 2009-06-21 23:12:40 +0200
Fix borked test
r6764@Thesaurus (orig r6763):  ribasushi | 2009-06-23 10:33:59 +0200
Real inheritance ordering for load_namespaces
r6772@Thesaurus (orig r6771):  ribasushi | 2009-06-23 16:46:18 +0200
Move tests around, add extra has_one relationship
r6773@Thesaurus (orig r6772):  caelum | 2009-06-23 18:36:22 +0200
add missing ' to doc
r6781@Thesaurus (orig r6780):  ribasushi | 2009-06-24 11:08:02 +0200
Properly name the relinfo variable
r6782@Thesaurus (orig r6781):  ribasushi | 2009-06-24 12:12:49 +0200
find_related fix for single-type relationships
r6783@Thesaurus (orig r6782):  nigel | 2009-06-24 17:28:33 +0200
 r11786@hex:  nigel | 2009-06-24 16:27:58 +0100
 Fixed set_$rel with where restriction deleting rows outside the restriction

r6784@Thesaurus (orig r6783):  nigel | 2009-06-24 17:47:31 +0200
 r11788@hex:  nigel | 2009-06-24 16:47:04 +0100
 Rework of set_$rel patch with less obfuscation

r6789@Thesaurus (orig r6788):  ribasushi | 2009-06-25 09:19:10 +0200
Commit test inspired by joel - it seemingly fails on Mac?
r6790@Thesaurus (orig r6789):  ribasushi | 2009-06-25 11:04:26 +0200
Minor cleanups
r6793@Thesaurus (orig r6792):  teejay | 2009-06-26 14:43:05 +0200
normalised artist_id, and plural relationships to plural names making use of alias/relname less ambiguous than relname/tablename being the same, also added a little more info on joining/relationships
r6794@Thesaurus (orig r6793):  tomboh | 2009-06-26 15:25:19 +0200
Documentation fix:
- timezone is no longer an extra setting
- fix a typo of 'subsequently'

r6795@Thesaurus (orig r6794):  gphat | 2009-06-26 16:33:35 +0200
Fix typo in ResultSet docs

r6803@Thesaurus (orig r6802):  ribasushi | 2009-06-27 12:39:03 +0200
Todoified (unsolvable) test from RT#42466
r6804@Thesaurus (orig r6803):  ribasushi | 2009-06-27 12:52:26 +0200
POD patch from RT#46808
r6805@Thesaurus (orig r6804):  ribasushi | 2009-06-27 13:59:03 +0200
Adjust sqlt schema parser to add tables in FK dependency order
r6806@Thesaurus (orig r6805):  ribasushi | 2009-06-27 14:08:35 +0200
Bump author SQLT dependency for early developer testing
Regenerate SQLite schema with new parser/sqlt
Use throw_exception in lieu of plain die when possible
r6813@Thesaurus (orig r6812):  castaway | 2009-06-28 06:11:08 +0200
Tests for grouping with prefetch

r6820@Thesaurus (orig r6819):  ribasushi | 2009-06-28 13:00:03 +0200
The prefetch+group_by is a complex problem - branch
r6844@Thesaurus (orig r6843):  abraxxa | 2009-06-29 11:02:17 +0200
fixed typo in test

r6848@Thesaurus (orig r6847):  ribasushi | 2009-06-29 19:09:00 +0200
Minor Ordered optimization (don't use count)

15 years agoremove the _store stuff for on_connect_do
Rafael Kitover [Mon, 29 Jun 2009 20:07:43 +0000]
remove the _store stuff for on_connect_do

15 years agoMinor Ordered optimization (don't use count)
Peter Rabbitson [Mon, 29 Jun 2009 17:09:00 +0000]
Minor Ordered optimization (don't use count)

15 years agoremove DateTime from 73oracle.t
Rafael Kitover [Mon, 29 Jun 2009 15:20:32 +0000]
remove DateTime from 73oracle.t

15 years agoAnd score! (all works)
Peter Rabbitson [Mon, 29 Jun 2009 10:05:37 +0000]
And score! (all works)

15 years agofixed typo in test
Alexander Hartmaier [Mon, 29 Jun 2009 09:02:17 +0000]
fixed typo in test

15 years agoclearer
Peter Rabbitson [Mon, 29 Jun 2009 08:14:45 +0000]
clearer

15 years agoMost of the grouped prefetch solution
Peter Rabbitson [Mon, 29 Jun 2009 08:11:13 +0000]
Most of the grouped prefetch solution

15 years agoTest cleanups
Peter Rabbitson [Mon, 29 Jun 2009 08:07:09 +0000]
Test cleanups

15 years agoThis is tested elsewhere (2)
Peter Rabbitson [Mon, 29 Jun 2009 07:50:43 +0000]
This is tested elsewhere (2)

15 years agoThis is tested elsewhere
Peter Rabbitson [Mon, 29 Jun 2009 07:49:53 +0000]
This is tested elsewhere

15 years agoSupport for -select/-as in SQLAHacks field selection
Peter Rabbitson [Mon, 29 Jun 2009 07:46:13 +0000]
Support for -select/-as in SQLAHacks field selection

15 years agoA couple of test fixes
Peter Rabbitson [Mon, 29 Jun 2009 07:44:25 +0000]
A couple of test fixes

15 years agoRogue comments
Peter Rabbitson [Sun, 28 Jun 2009 20:38:43 +0000]
Rogue comments

15 years agoLose the literal sql bits - castaway is right it's silly to support those
Peter Rabbitson [Sun, 28 Jun 2009 13:06:57 +0000]
Lose the literal sql bits - castaway is right it's silly to support those

15 years agoWhoops
Peter Rabbitson [Sun, 28 Jun 2009 12:42:48 +0000]
Whoops

15 years agoProper prefetch+group test
Peter Rabbitson [Sun, 28 Jun 2009 12:39:54 +0000]
Proper prefetch+group test

15 years agoNormalize group_by
Peter Rabbitson [Sun, 28 Jun 2009 12:38:36 +0000]
Normalize group_by

15 years agoThe prefetch+group_by is a complex problem - branch
Peter Rabbitson [Sun, 28 Jun 2009 11:00:03 +0000]
The prefetch+group_by is a complex problem - branch

15 years agoAdjust a couple of tests for new behavior (thus all of this might be backwards incomp...
Peter Rabbitson [Sun, 28 Jun 2009 09:07:33 +0000]
Adjust a couple of tests for new behavior (thus all of this might be backwards incompatible to the point of being useless):
The counts in t/90join_torture.t are now 5*3, not 5*3*3, as a second join is not induced by search_related
The raw sql scan in t/prefetch/standard.t is just silly, won't even try to understand it
Just to maintain the TreeLike folding, I add a 3rd children join which was inserted by search_related before the code changes

15 years agoMinor omission
Peter Rabbitson [Sun, 28 Jun 2009 09:04:26 +0000]
Minor omission

15 years agoThe proposed fix (do not add an extra join if it is already present in the topmost...
Peter Rabbitson [Sun, 28 Jun 2009 08:35:11 +0000]
The proposed fix (do not add an extra join if it is already present in the topmost join)

15 years agoThetest case that started it all
Peter Rabbitson [Sun, 28 Jun 2009 08:34:16 +0000]
Thetest case that started it all

15 years agoTests for grouping with prefetch
Jess Robinson [Sun, 28 Jun 2009 04:11:08 +0000]
Tests for grouping with prefetch

15 years agoprototype blob implementation
Rafael Kitover [Sun, 28 Jun 2009 00:14:56 +0000]
prototype blob implementation

15 years agoMerge 'trunk' into 'sybase'
Rafael Kitover [Sat, 27 Jun 2009 22:21:09 +0000]
Merge 'trunk' into 'sybase'

r5665@hlagh (orig r6780):  ribasushi | 2009-06-24 02:08:02 -0700
Properly name the relinfo variable
r5666@hlagh (orig r6781):  ribasushi | 2009-06-24 03:12:49 -0700
find_related fix for single-type relationships
r5667@hlagh (orig r6782):  nigel | 2009-06-24 08:28:33 -0700
 r11786@hex:  nigel | 2009-06-24 16:27:58 +0100
 Fixed set_$rel with where restriction deleting rows outside the restriction

r5668@hlagh (orig r6783):  nigel | 2009-06-24 08:47:31 -0700
 r11788@hex:  nigel | 2009-06-24 16:47:04 +0100
 Rework of set_$rel patch with less obfuscation

r5691@hlagh (orig r6788):  ribasushi | 2009-06-25 00:19:10 -0700
Commit test inspired by joel - it seemingly fails on Mac?
r5692@hlagh (orig r6789):  ribasushi | 2009-06-25 02:04:26 -0700
Minor cleanups
r5721@hlagh (orig r6792):  teejay | 2009-06-26 05:43:05 -0700
normalised artist_id, and plural relationships to plural names making use of alias/relname less ambiguous than relname/tablename being the same, also added a little more info on joining/relationships
r5722@hlagh (orig r6793):  tomboh | 2009-06-26 06:25:19 -0700
Documentation fix:
- timezone is no longer an extra setting
- fix a typo of 'subsequently'

r5723@hlagh (orig r6794):  gphat | 2009-06-26 07:33:35 -0700
Fix typo in ResultSet docs

r5724@hlagh (orig r6802):  ribasushi | 2009-06-27 03:39:03 -0700
Todoified (unsolvable) test from RT#42466
r5725@hlagh (orig r6803):  ribasushi | 2009-06-27 03:52:26 -0700
POD patch from RT#46808
r5726@hlagh (orig r6804):  ribasushi | 2009-06-27 04:59:03 -0700
Adjust sqlt schema parser to add tables in FK dependency order
r5727@hlagh (orig r6805):  ribasushi | 2009-06-27 05:08:35 -0700
Bump author SQLT dependency for early developer testing
Regenerate SQLite schema with new parser/sqlt
Use throw_exception in lieu of plain die when possible

15 years agoMerge 'trunk' into 'on_connect_call'
Rafael Kitover [Sat, 27 Jun 2009 22:20:25 +0000]
Merge 'trunk' into 'on_connect_call'

r5665@hlagh (orig r6780):  ribasushi | 2009-06-24 02:08:02 -0700
Properly name the relinfo variable
r5666@hlagh (orig r6781):  ribasushi | 2009-06-24 03:12:49 -0700
find_related fix for single-type relationships
r5667@hlagh (orig r6782):  nigel | 2009-06-24 08:28:33 -0700
 r11786@hex:  nigel | 2009-06-24 16:27:58 +0100
 Fixed set_$rel with where restriction deleting rows outside the restriction

r5668@hlagh (orig r6783):  nigel | 2009-06-24 08:47:31 -0700
 r11788@hex:  nigel | 2009-06-24 16:47:04 +0100
 Rework of set_$rel patch with less obfuscation

r5691@hlagh (orig r6788):  ribasushi | 2009-06-25 00:19:10 -0700
Commit test inspired by joel - it seemingly fails on Mac?
r5692@hlagh (orig r6789):  ribasushi | 2009-06-25 02:04:26 -0700
Minor cleanups
r5721@hlagh (orig r6792):  teejay | 2009-06-26 05:43:05 -0700
normalised artist_id, and plural relationships to plural names making use of alias/relname less ambiguous than relname/tablename being the same, also added a little more info on joining/relationships
r5722@hlagh (orig r6793):  tomboh | 2009-06-26 06:25:19 -0700
Documentation fix:
- timezone is no longer an extra setting
- fix a typo of 'subsequently'

r5723@hlagh (orig r6794):  gphat | 2009-06-26 07:33:35 -0700
Fix typo in ResultSet docs

r5724@hlagh (orig r6802):  ribasushi | 2009-06-27 03:39:03 -0700
Todoified (unsolvable) test from RT#42466
r5725@hlagh (orig r6803):  ribasushi | 2009-06-27 03:52:26 -0700
POD patch from RT#46808
r5726@hlagh (orig r6804):  ribasushi | 2009-06-27 04:59:03 -0700
Adjust sqlt schema parser to add tables in FK dependency order
r5727@hlagh (orig r6805):  ribasushi | 2009-06-27 05:08:35 -0700
Bump author SQLT dependency for early developer testing
Regenerate SQLite schema with new parser/sqlt
Use throw_exception in lieu of plain die when possible

15 years agoBump author SQLT dependency for early developer testing
Peter Rabbitson [Sat, 27 Jun 2009 12:08:35 +0000]
Bump author SQLT dependency for early developer testing
Regenerate SQLite schema with new parser/sqlt
Use throw_exception in lieu of plain die when possible

15 years agoAdjust sqlt schema parser to add tables in FK dependency order
Peter Rabbitson [Sat, 27 Jun 2009 11:59:03 +0000]
Adjust sqlt schema parser to add tables in FK dependency order

15 years agoPOD patch from RT#46808
Peter Rabbitson [Sat, 27 Jun 2009 10:52:26 +0000]
POD patch from RT#46808

15 years agoTodoified (unsolvable) test from RT#42466
Peter Rabbitson [Sat, 27 Jun 2009 10:39:03 +0000]
Todoified (unsolvable) test from RT#42466

15 years agoFix typo in ResultSet docs
Cory G Watson [Fri, 26 Jun 2009 14:33:35 +0000]
Fix typo in ResultSet docs

15 years agoDocumentation fix:
Tom Hukins [Fri, 26 Jun 2009 13:25:19 +0000]
Documentation fix:
- timezone is no longer an extra setting
- fix a typo of 'subsequently'

15 years agonormalised artist_id, and plural relationships to plural names making use of alias...
Aaron Trevena [Fri, 26 Jun 2009 12:43:05 +0000]
normalised artist_id, and plural relationships to plural names making use of alias/relname less ambiguous than relname/tablename being the same, also added a little more info on joining/relationships

15 years agoMinor cleanups
Peter Rabbitson [Thu, 25 Jun 2009 09:04:26 +0000]
Minor cleanups

15 years agoCommit test inspired by joel - it seemingly fails on Mac?
Peter Rabbitson [Thu, 25 Jun 2009 07:19:10 +0000]
Commit test inspired by joel - it seemingly fails on Mac?

15 years agofixup POD, comment out count
Rafael Kitover [Thu, 25 Jun 2009 03:31:06 +0000]
fixup POD, comment out count

15 years agoIC::DT does support timestamp with timezone
Rafael Kitover [Wed, 24 Jun 2009 17:49:25 +0000]
IC::DT does support timestamp with timezone

15 years agoRework of set_$rel patch with less obfuscation
Nigel Metheringham [Wed, 24 Jun 2009 21:47:04 +0000]
Rework of set_$rel patch with less obfuscation

15 years agoFixed set_$rel with where restriction deleting rows outside the restriction
Nigel Metheringham [Wed, 24 Jun 2009 21:27:58 +0000]
Fixed set_$rel with where restriction deleting rows outside the restriction

15 years agofind_related fix for single-type relationships
Peter Rabbitson [Wed, 24 Jun 2009 10:12:49 +0000]
find_related fix for single-type relationships

15 years agoProperly name the relinfo variable
Peter Rabbitson [Wed, 24 Jun 2009 09:08:02 +0000]
Properly name the relinfo variable