Peter Rabbitson [Fri, 18 Jul 2014 16:26:57 +0000]
Replace many closure-based proxy methods with static qsubs
Not only does this buy a quantum of performance, but it greatly enhances
readability of these methods on deparse
Deliberately not converting the ManyToMany helper - this code needs a
thorough regut :(
Peter Rabbitson [Sun, 20 Jul 2014 11:47:09 +0000]
Reformat add_relationship_accessor for future edit
No functional changes, read diff under -w
Peter Rabbitson [Fri, 18 Jul 2014 16:21:20 +0000]
Standardize the Moo import block, move quote_sub/qsub into ::_Util
This way we will have less boilerplate in subsequent commits where we
will use quote_sub standalone
Add yet another test to make absolutely sure quote_sub won't leak any of the
strictures insanity
Peter Rabbitson [Mon, 21 Jul 2014 07:57:24 +0000]
Replace B::perlstring with our own implmentation thereof
Haarg++ for pointing this out
Peter Rabbitson [Fri, 18 Jul 2014 16:11:15 +0000]
Add internal assertion guard for some indirect calls (for now only create/new)
Modeled on the idea of
a9da9b6a, this one blows up when a public proxy
is called insted of the equivalent public chan of methods
This particular set of changes to create() and new() is solely an
optimisation (unlike a subsequent commit to ::ResultSourceProxy)
Peter Rabbitson [Sun, 20 Jul 2014 12:07:28 +0000]
Greatly improve diagnostic messages of _resolve_relationship_condition
Peter Rabbitson [Fri, 18 Jul 2014 09:49:49 +0000]
Remove duplicate arg-check in create(), adjust exception text
Peter Rabbitson [Fri, 18 Jul 2014 08:42:58 +0000]
Simplify no wantarray assert (
a9da9b6a) - fish out args from caller
Peter Rabbitson [Thu, 17 Jul 2014 09:20:30 +0000]
is_X_value functions introduced in
3705e3b28 migrated to SQLA with fixups
Peter Rabbitson [Thu, 17 Jul 2014 07:05:19 +0000]
Better, consistent handling of -literal/-value in the cond collapser
Peter Rabbitson [Thu, 17 Jul 2014 12:53:42 +0000]
(travis) Temporary hard-spec of the SQLA dep, lockstep release
Peter Rabbitson [Thu, 17 Jul 2014 12:22:24 +0000]
Move the "special loading" check to DBICTest::Util
No functional changes
Peter Rabbitson [Thu, 17 Jul 2014 12:37:12 +0000]
Skip the namespace tests on plain install
Protect users from future spurious changes in uncleaned imports
Peter Rabbitson [Thu, 17 Jul 2014 13:12:52 +0000]
(travis) T::B 1.005 is superseded by 1.300 - remove handholding
Peter Rabbitson [Tue, 15 Jul 2014 01:48:40 +0000]
Add one extra is_exception check (missed a spot during
841efcb3f)
In addition tweak the message so that carp_unique can in fact catch it
properly, and test that the proper amount of warnings is in fact emitted
Peter Rabbitson [Tue, 15 Jul 2014 00:36:44 +0000]
Consolidate bits and pieces under ::_Util::refdesc
No functional changes
Peter Rabbitson [Tue, 15 Jul 2014 05:14:31 +0000]
Remove double has_column check in new=>store_column
Standardize exception texts across ::Row
Peter Rabbitson [Sat, 12 Jul 2014 15:32:52 +0000]
Another heavy refactor of the rel resolver prototype (sequel to
03f6d1f7)
Change the return value of _resolve_relationship_condition to a 3-value
hash, with a lot of the safety logic consolidated within that method. The
best way to gauge the significance of the changes is to look at the diff
of lib/DBIx/Class/Relationship/Base.pm
Also stop returning the "noncondition-columns" in the _resolve_condition
compa-shim. The information was only used by the code removed from
::Relationship::Base, and is rather new with no evidence of use within
CPAN/DarkPAN. It can be easily added back if necessary.
Peter Rabbitson [Sat, 12 Jul 2014 11:52:45 +0000]
Move a CDBI escape to the CDBI class hierarchy
Peter Rabbitson [Sat, 12 Jul 2014 10:32:20 +0000]
Refactor ::DBIHacks::_extract_fixed_condition_columns (sequel to
8d005ad9)
Instead of just returning an array of column names, switch to a hashref that
can double as an inferred-value bag. As a bonus deduplicates and folds-in
another codepath from ::ResultSet::_merge_with_rscond
In the process fixup _collapse_cond to be even more robust in some arrayref
corner-cases. Now it is guaranteed to return a hashref at all times
Peter Rabbitson [Sat, 12 Jul 2014 09:51:00 +0000]
Switch UNRESOLVABLE_CONDITION to an _Util constant
Peter Rabbitson [Tue, 8 Jul 2014 05:42:06 +0000]
Older perls get confused by this construct - rewrap
Peter Rabbitson [Tue, 8 Jul 2014 04:58:03 +0000]
Revert optimistic and sloppy changes from
3705e3b2 - DBI does not always DTRT
In addition I rebroke the issue from RT#79576, ffs man pay attention
<ribasushi> general question - does DBI guarantee that objects with stringification overload will get their stringification called
<ribasushi> or is this up to the DBD (and thus will vary)
<mje> ribasushi, I had an issue in DBD::ODBC ages ago with this - looking for it now
<mje> rt 78838 - bind_param does not correctly stringify blessed objects when connected to MS SQL Server, magic was not being applied in DBD::ODBC case
<mje> so I think the answer to your question is DBI does not but DBDs should if they are written correctly
<ribasushi> I wonder why DBI does not
<timbunce_> ribasushi: no explicit ‘guarantee’ but I think any place it doesn’t is probably a bug.
<ribasushi> timbunce_: given mje had to do magic in his DBD, I suspect DBI has a "hole" somewhere then no?
<ribasushi> basically I was looking into removing the explicit stringifications in DBIC, hence the question
<ribasushi> if there is consensus that DBI ought to do it all on its own, we could test for it, fix it up, and then I disable the checks when I detect a sufficiently advanced DBI.pm
<timbunce> yes, ribasushi: “… we could test for it, fix it up, and then I disable the checks when I detect a sufficiently advanced DBI.pm or something”
<timbunce> :)
<ribasushi> nod ;)
Brendan Byrd [Wed, 12 Feb 2014 14:32:45 +0000]
Initial (now passing) tests prompting the order_by analyzer revamp
302d35f8
Peter Rabbitson [Tue, 8 Jul 2014 04:09:40 +0000]
Teach order_by stability analyzer about search_related
There is more stuff we can do here as noted in the FIXME added to the very
end of _extract_colinfo_of_stable_main_source_order_by_portion. But for the
time being this will do (also see the subsequent tests for extra insanity we
could and should add)
Tests come in the next commit
Peter Rabbitson [Tue, 8 Jul 2014 04:18:15 +0000]
Pass the main rsrc in attrs, instead of trying to fish it out later
Peter Rabbitson [Tue, 8 Jul 2014 00:30:52 +0000]
Clearer name of method/variables before refactoring
Peter Rabbitson [Tue, 8 Jul 2014 00:05:04 +0000]
Switch DBIC::Storage::Statistics to Moo (for trial purposes)
This is a component which has some subclassing in the wild, use it as a
canary to highlight any remaining issues Moo might have within DBIC
Peter Rabbitson [Tue, 8 Jul 2014 00:46:14 +0000]
Break out the test tracer into a standalone file
Peter Rabbitson [Mon, 7 Jul 2014 22:57:20 +0000]
IFF debug goes to STDERR by default - silence possible wide-char warnings
Peter Rabbitson [Mon, 7 Jul 2014 22:45:32 +0000]
Stop permanently enabling autoflush on the debug filehandle
Peter Rabbitson [Mon, 7 Jul 2014 22:34:24 +0000]
Expand closed STDERR test, check actual exception
skaufman [Sun, 29 Jun 2014 02:33:27 +0000]
Better (and much more precise) explanation of Moose/Moo subclassing
skaufman [Sun, 29 Jun 2014 02:33:27 +0000]
Better example/documentation of custom resultset classes
skaufman [Sun, 29 Jun 2014 02:33:27 +0000]
Change tab space (4 => 2), no further changes
Peter Rabbitson [Tue, 1 Jul 2014 09:00:57 +0000]
Move the Subclassing POD under Examples
Peter Rabbitson [Tue, 1 Jul 2014 08:45:02 +0000]
namespace::autoclean is a really stupid idea - purge with fire
Peter Rabbitson [Sat, 28 Jun 2014 10:23:52 +0000]
Forgotten reenable of syntax test to guard for
37af2968e
Peter Rabbitson [Tue, 1 Jul 2014 09:33:14 +0000]
Add an explicit webchat link, change IRC to x_IRC
This should work around the issue of 'preferred company/startup' raised in
https://github.com/CPAN-API/metacpan-web/issues/1235#issuecomment-
47610925
Alexander Hartmaier [Thu, 8 Aug 2013 08:07:51 +0000]
Improved docs for ResultSet columns attribute on how to retrieve related columns
Alexander Hartmaier [Thu, 18 Jul 2013 16:31:46 +0000]
More coherent +(columns/select/as) quoting notices and descriptions
Dagfinn Ilmari Mannsåker [Fri, 28 Mar 2014 16:48:17 +0000]
Populate caches for related result sets even if they're empty
This avoids unnecessary database hits when accessing prefetched related
resultsets with no rows.
Peter Rabbitson [Tue, 17 Jun 2014 10:41:06 +0000]
Add a self-explanatory *compile-time* $ENV{DBIC_SHUFFLE_UNORDERED_RESULTSETS}
The idea is to be able to catch tests that rely on a particular order of
the results of ->all. ->next is untouched as injecting extra \'RANDOM'
order clauses is too heavy-handed for a test environment (and besides, having
->all and ->next return stuff in differing order is just as good of a
monkey-wrench)
Peter Rabbitson [Mon, 16 Jun 2014 14:37:27 +0000]
Minor optimization of codepath (no func changes)
Peter Rabbitson [Mon, 16 Jun 2014 14:15:44 +0000]
Fix ::Ordered in combination with delete_all
See pod-changes for description of the solution. I cringe but... oh well
Peter Rabbitson [Mon, 16 Jun 2014 12:28:25 +0000]
(travis) Announce what we set during POISON_ENV
Peter Rabbitson [Mon, 16 Jun 2014 12:21:59 +0000]
Remove vestigial test envvar
Peter Rabbitson [Fri, 13 Jun 2014 14:37:31 +0000]
More fallout of the _resolve_cond refactor
03f6d1f7
I had in the back of my mind "one can't just create random objects" but
figured it may work anyway - WROOOOOOOONG. Instead pass through the hash
as before, will need to rethink that part before settling on an API for
resolve_relationship_condition. FML
Peter Rabbitson [Thu, 12 Jun 2014 04:53:38 +0000]
Fix test tripping on external warnings
Peter Rabbitson [Thu, 12 Jun 2014 04:54:12 +0000]
Even more robust _resolve_condition shim - guess rel_name when needed
Karen Etheridge [Wed, 11 Jun 2014 19:46:25 +0000]
replace all remaining uses of self_rowobj with self_resultobj in pod, test schemas
Gerda Shank [Mon, 10 Feb 2014 19:22:06 +0000]
Add (now passing) tests for set_from_related() via 'foreign_resultobj'
Peter Rabbitson [Wed, 11 Jun 2014 14:32:59 +0000]
Add foreign_resultobj to the customrel signature
Tests in next commit
Peter Rabbitson [Wed, 11 Jun 2014 14:09:22 +0000]
Pedantry - settle on rel_name vs relname, consistency throughout
Only outlier is relname_to_table_alias, which should have never existed
in the first place. Punting deprecation/etc cycle until future Oracle
work... sigh
Peter Rabbitson [Wed, 11 Jun 2014 13:22:31 +0000]
Rename (with a silent compat shim) couple of badly named customcond args
foreign_relname makes absolutely no sense - it *is* a relationship after all,
of course it has a name. Now renamed to rel_name to be consistent with
the relationship condition resolver
self_rowobj uses the old rowobj nomenclature - switch to self_resultobj
Peter Rabbitson [Wed, 11 Jun 2014 13:51:30 +0000]
Centralize custom rel args check, be more thorough
Peter Rabbitson [Wed, 11 Jun 2014 11:15:02 +0000]
Adjust error messages - joinfrees can be returned from noncode rels
Peter Rabbitson [Wed, 11 Jun 2014 10:05:13 +0000]
First step to add some sanity to _resolve_condition
The original problem starts here:
<mst> oh, yeah, I just used to check ref and randomly swap things
<mst> it worked for enough years :)
<ribasushi> I will be forwarding you my psychiatrist bill
READ DIFF AT YOUR OWN RISK
Over the years _resolve_condition has accumulated 3 (or 4, or 5, depends
how you look at it) distinct call-modes. None having anything to do with
another. Also it is a hot method, holding crucial functionality, which
of course means that currently at least 3 projects on CPAN are using it,
despite the private attribute. Which in turn means couple more orders of
magnitude of users on the DarkPAN. Thus just killing this method
outright *without a replacement* is not an option.
A from-scratch replacement in the face of only one person currently
*barely* understanding this codepath is a scary proposition.
Instead create an elaborate (and scarily complete) shim to proxy to a
new method holding all the logic (with the idea of making it an official
API in the coming commits).
There are no changes to any other codepaths, as this is how we ensure that
the shim is sane, and works. Next step is to erradicate all cases of the old
call in the current codebase (while leaving the sub/shim intact)/
Now let's see if we can fix CampusExplorer's bugs first...
Peter Rabbitson [Tue, 10 Jun 2014 11:54:32 +0000]
Fix incorrect handling of custom relationship conditions containing literals
Peter Rabbitson [Fri, 6 Jun 2014 16:38:52 +0000]
Revert heading change from
06752a03d, centralize the URL spec
There were several broken links referring to the old way of doing
things, and help/support sounds better anyway. Reverting with added
comments and whatnot
Peter Rabbitson [Wed, 4 Jun 2014 23:44:37 +0000]
Fix grave resultset-level delete/update bug
In case of a read-modify cycle (usually due to multicolumn PK, without
multicolumn IN support) DBIC would end up with a null condition and nuke
the entire underlying table from orbit
I have no words towards my younger self...
Peter Rabbitson [Wed, 4 Jun 2014 23:10:50 +0000]
Do not skip running from_storage filter when a NULL is returned
Peter Rabbitson [Wed, 4 Jun 2014 23:05:31 +0000]
Use a nullable column for the ::FC test
This is just a retarget, no test was actually modified
Peter Rabbitson [Wed, 4 Jun 2014 12:53:48 +0000]
Teach FC about literals
Peter Rabbitson [Wed, 4 Jun 2014 11:55:30 +0000]
A number of equivalent-logic ::FC refactors
Stop invoking a comparison on the unfiltered values - delegate it properly
to set_column()
Peter Rabbitson [Wed, 4 Jun 2014 11:28:42 +0000]
Reorganize IC column setters
Should again result in no functional changes
Peter Rabbitson [Mon, 2 Jun 2014 12:05:45 +0000]
Rename some variables and reformat the FC/IC codepaths for clarity
Zero functional changes
Peter Rabbitson [Mon, 2 Jun 2014 10:20:45 +0000]
Clarify FC/IC conflict exception
Peter Rabbitson [Sat, 31 May 2014 09:34:29 +0000]
Call mro reinit in tests only on 5.8
Peter Rabbitson [Wed, 4 Jun 2014 00:17:20 +0000]
Fix incorrect handling of stringifiable objects in cond collapser
Thinko introduced in the recent
8d005ad9
Peter Rabbitson [Wed, 4 Jun 2014 03:30:26 +0000]
Consolidate handling of "is this a literal" and "is this a value"
In the process fix inability of IC to deal with \[], and simplify
the overal codepath bind value passing codepath
Aside from the bugfixes there should be no functional changes
Work inspired by a report and preliminary patch from dim0xff++
Peter Rabbitson [Tue, 27 May 2014 23:21:13 +0000]
Use the correct $PROGRAM_NAME for the initial dbicadmin POD gneration
Karen Etheridge [Tue, 27 May 2014 23:09:20 +0000]
spelling fix
Peter Rabbitson [Mon, 26 May 2014 11:47:52 +0000]
Retire DBIC/SqlMakerTest.pm now that SQLA::Test provides the same function
Peter Rabbitson [Wed, 28 May 2014 03:04:27 +0000]
Switch most remaining debug-hooks to $dbictest_schema->is_executed_querycount()
Peter Rabbitson [Tue, 27 May 2014 17:17:26 +0000]
Retire DBIC::DebugObj, replace with $dbictest_schema->is_executed_sql_bind()
This cuts down on a lot of the silliness with debugcb/debugobj, and makes
for more precise tests as a whole. Went through great pains to not disturb
any existing tests, fingercross this is indeed the case
Read under -w for sanity
Peter Rabbitson [Tue, 27 May 2014 16:58:12 +0000]
Excise live test left over after
ac0c082542
No access to a live test system currently, and likely the test is incorrect
anyway (no mention of significant parenthesis)
Peter Rabbitson [Tue, 27 May 2014 06:29:46 +0000]
Consolidate pg/mysql and standalone SQLite savepoint tests to run on all 3
A larger consolidation will take place later with T::WV, but all in due time
Peter Rabbitson [Tue, 27 May 2014 11:15:45 +0000]
Fix multiple savepointing transactions on DBD::SQLite
Problem was missed during review of
86a51471ce (how the fuck did I let this
abomination through anyway), and not encountered due to insufficient testing.
The naive statement parser in DBD::SQLite when running against older libsqlite
mistakenly treats ROLLBACK TRANSACTION TO... as an actual TXN rollback, and
as a result desyncs the internal AutoCommit flag state [1]
Fix by simply using the shorter (still valid) syntax [2], and by removing the
sloppy workaround hiding the actual problem.
[1] https://github.com/DBD-SQLite/DBD-SQLite/blob/1.42/dbdimp.c#L824:L852
[2] http://www.sqlite.org/lang_savepoint.html
Peter Rabbitson [Tue, 27 May 2014 11:34:31 +0000]
Fix erroneous todoification in sqlite test
Peter Rabbitson [Tue, 27 May 2014 08:22:29 +0000]
Less warning noise from sqlite test
Peter Rabbitson [Tue, 27 May 2014 08:09:43 +0000]
Move the DSN-lock machinery from
8d6b1478d into DBICTest::BaseSchema
This is just a c/p job, expecting zero functional changes
The t/admin/02ddl.t are cosmetic changes fixing a nested &cref leak never
before encountered on <= 5.8.7
Peter Rabbitson [Tue, 27 May 2014 08:05:08 +0000]
Consolidate lib-wide frameskip, adjust the ::RunMode loading order
Peter Rabbitson [Tue, 27 May 2014 05:53:38 +0000]
Reaching for IO::File here makes no sense
Peter Rabbitson [Tue, 27 May 2014 05:50:48 +0000]
This test was essentially c/p-ed in
2cc3a7be3, consolidate
The overriding difference is the new-connection == new-storage == new-sqla
Add an explicit codepath to test that, but overall there is plenty of
other places that verify this behavior to death
Examine under -w -C -M to make sense out of the consolidation (minimal
changes were attempted)
Peter Rabbitson [Sun, 25 May 2014 09:52:13 +0000]
Improve complex order+prefetch exception message
Naveed Massjouni [Fri, 9 May 2014 18:51:26 +0000]
Updated discard_changes docs regarding { force_pool => 'master' }
Alexander Hartmaier [Wed, 5 Mar 2014 17:43:48 +0000]
The initial (now passing) test for RT#93244 (fixed in
8d005ad9)
Peter Rabbitson [Sat, 17 May 2014 09:39:49 +0000]
Massively refactor and sanify condition collapsing
Wow... what a ride. This commit adds a faithful reimplementation of the
SQLA descend algorithm, but instead of SQL produces a reduced HASH that
nevertheless corresponds to the original query 1:1 (or so I *really* hope)
This is another one of these "I can try it with DQ but I don't even have
tests" stories, so once again opted to implement things the "dumb" way.
The benefits are quite substantial:
- Better deduplication of WHERE condition
- Consolidated functions - the vaguely similar logic in the current version
of _collapse_cond replaces almost the entirety of:
::DBIHacks::_extract_fixed_condition_columns
::ResultSet::_collapse_cond
::ResultSet::_stack_cond
- Extra fixes for create/populate inheritance corner cases
- More predictable SQL condition generation order (which incidentally may
prove problematic down the road with broken tests, but oh well, we'll
burn when we get there)
- Ton of extra tests and corner cases
We even managed to fulfill a longstanding TODO, even though it is a *lucky*
side-effect and not a real fix. Making a note to address this later... sigh.
This work started from
d8b7d9f58..337f3ee80, which while taking the right
direction had too many loose ends. ilmari++
Peter Rabbitson [Thu, 22 May 2014 12:34:11 +0000]
Remove dead code missed by cleanup
b72339859
Peter Rabbitson [Fri, 16 May 2014 09:31:21 +0000]
Add test ensuring we do not lose binds on esoteric RSC+distinct
Test inspired by auditing
cc2b92553 (a precursor of ad1d374e60)
Dagfinn Ilmari Mannsåker [Sat, 12 Apr 2014 17:29:47 +0000]
Refactor _recurse_fields to return the bind values
Only ->select actually wants it in $self->{select_bind}, the others
either don't care or want them somewhere else.
Peter Rabbitson [Thu, 22 May 2014 11:37:08 +0000]
Fix multi-value literal populate not working with simplified bind spec
This arcane use-case got missed when
1b5ddf23 was integrated
Peter Rabbitson [Wed, 21 May 2014 07:57:50 +0000]
Fix bogus reference to related class (introduced way back in
97c96475)
Peter Rabbitson [Tue, 20 May 2014 14:16:40 +0000]
Remove forgotten part of
a48693f4 (incompletely reverted by e327f126)
Mintywalker [Sat, 12 Apr 2014 17:13:10 +0000]
$pedantry =~ s/columns/column/
Peter Rabbitson [Thu, 10 Apr 2014 22:43:35 +0000]
Properly detect and test mysql v3 default JOIN behavior
While investigating how to rewrite the rather useless
39712b481, it became
clear that not only can the test be performed offline, but that there is
already a pseudo-test due to sqlite's version of 3.x.y being mistaken for
MySQL 3
Properly fix the test to check behavior on both versions 3 and 4, and in the
process fix a potential bug of a stale sqlmaker when a reconnect cycle takes
place against a *different* physical rdbms (which btw makes me realize another
problem with trying to do mixed-environment replication... sigh)
Peter Rabbitson [Thu, 10 Apr 2014 22:38:10 +0000]
Clarify that sql_maker is just a getter at this point
Peter Rabbitson [Tue, 11 Mar 2014 06:39:56 +0000]
Minimal optimization of the new+prefetch codepath (no func. changes)
Peter Rabbitson [Sat, 12 Apr 2014 04:36:01 +0000]
Fix missing true return lost during
e5a62c46f
This never resulted in a serious problem because it was papered over by a
bug in Class::C3::Componentised. Addressing it at the source requires a bit
extra coding, hence no new CC3C release yet.