9 years ago(travis) Much lighter CI runs when smoking a PR ghpr/closed/test_no_ci_on_pr
Peter Rabbitson [Fri, 3 Apr 2015 06:52:22 +0000]
(travis) Much lighter CI runs when smoking a PR

9 years ago(travis) Enhance post-success to run a full CPAN install cycle
Peter Rabbitson [Thu, 2 Apr 2015 00:52:56 +0000]
(travis) Enhance post-success to run a full CPAN install cycle

This is as much a check for our installability, as it is a smoke of CPAN.pm

9 years ago(travis) Refactor the NUMTHREADS codepaths, rename variable (see next commit)
Peter Rabbitson [Thu, 2 Apr 2015 08:13:40 +0000]
(travis) Refactor the NUMTHREADS codepaths, rename variable (see next commit)

9 years ago(travis) Factor out sitelib cleaner, we will call it more than once soon
Peter Rabbitson [Thu, 2 Apr 2015 00:41:11 +0000]
(travis) Factor out sitelib cleaner, we will call it more than once soon

9 years ago(travis) Restore proper post-success testing
Peter Rabbitson [Thu, 2 Apr 2015 00:37:19 +0000]
(travis) Restore proper post-success testing

The functionality was silently lost due to TravisCI issue #3533

9 years agoFix incorrect optdep output introduced in 34d2deae / 31c31b8d
Peter Rabbitson [Wed, 1 Apr 2015 20:52:41 +0000]
Fix incorrect optdep output introduced in 34d2deae / 31c31b8d

Too many thinkos for a simple feature like this. When given "foo" cpanm
actually tries to find '"foo"' on CPAN, and nothing works. The original
thinking was that the > in ~>= needs escaping when copy/pasted by a user, so
the ""-enclosing seemed like a valid approach. Of course this does not work
at all when the output is piped into cpanm directly:

 perl -Ilib -MDBIx::Class::Optional::Dependencies=-list_missing,deploy | cpanm

Moreover, all of the above is *utterly unnecessary*, as Foo~>=2 is entirely
equivalent to Foo~2, which requires no escaping (thus quoting) whatsoever.

Hopefully this is the end of the embarrassing flailing on this feature :(

9 years ago(travis) Revert backout of ODBC firebird testing aa0e82a2
Peter Rabbitson [Thu, 2 Apr 2015 01:04:24 +0000]
(travis) Revert backout of ODBC firebird testing aa0e82a2

It seems that sourceforge got their act together (for now)

9 years agoWork around Firebird/InterBase/ODBC crash in tests
Dagfinn Ilmari Mannsåker [Wed, 1 Apr 2015 17:52:22 +0000]
Work around Firebird/InterBase/ODBC crash in tests

If DBD::Firebird and DBD::InterBase are used in the same process, the
last one loaded leaves an active statement handle around and later
segfaults at some point in the tests.  Also, if DBD::ODBC is used
before either of them, they leak the handle but don't segfault.

Thus, to avoid breaking installing DBIC if more than one of
DBICTEST_FIREBIRD(_INTERBASE|ODBC)?_DSN is set, skip Interbase if any
of the others are set, and make sure to run the ODBC test last.

However, keep the leak-triggering order for author, CI and smoker
runs, so we don't forget that it needs fixing.

9 years ago(travis) Cut down the amount of APT repos during updates
Peter Rabbitson [Sat, 28 Mar 2015 07:22:48 +0000]
(travis) Cut down the amount of APT repos during updates

Sometimes things outside of ubuntu's control go down, and the entire smoke
will crash and burn for reasons completely unrelated to what we are doing

9 years agoExamples and documentation update and cleanup
Alex Beamish [Wed, 4 Sep 2013 17:45:39 +0000]
Examples and documentation update and cleanup

This version updates the original POD with slightly more information
so that more DBIx::Class functionality is revealed.  The source code is
removed from POD, and instead the reader is directed to the examples
directory.

It also changes the example schema from silly auto-inferred column
names to proper *id-suffixed ones.

9 years agoInflators are not respected by find() (being a ResultSet method)
Alastair McGowan-Douglas [Fri, 7 Nov 2014 17:00:21 +0000]
Inflators are not respected by find() (being a ResultSet method)

Adjust docs and add an extra IC::DT-specific FAQ section

9 years agoCross-reference alias with current_source_alias
Karen Etheridge [Wed, 8 Oct 2014 19:42:49 +0000]
Cross-reference alias with current_source_alias

(slightly modified 4e4b848d)

9 years agoClarify load_namespaces call -- more than one call can be made
Mike Wisener [Tue, 30 Sep 2014 03:58:59 +0000]
Clarify load_namespaces call -- more than one call can be made

9 years agoAdded add_columns documentation about controlling float precision
Robert Stone [Sat, 27 Dec 2014 19:19:10 +0000]
Added add_columns documentation about controlling float precision

9 years agoLinked "prefetch" in columns documentation
C.J. Adams-Collier [Mon, 2 Mar 2015 19:41:31 +0000]
Linked "prefetch" in columns documentation

9 years agoExample doesn't work without 'year' column
Ingy döt Net [Tue, 3 Mar 2015 19:29:39 +0000]
Example doesn't work without 'year' column

9 years agoFix uninitialized warnings on empty hashes passed to join/prefetch
Henry Van Styn [Wed, 29 Oct 2014 13:25:33 +0000]
Fix uninitialized warnings on empty hashes passed to join/prefetch

9 years ago(travis) Better handle ENV poisoning, get jobs under 4mb limit again
Peter Rabbitson [Wed, 18 Mar 2015 12:15:45 +0000]
(travis) Better handle ENV poisoning, get jobs under 4mb limit again

9 years ago(travis) Revert LMU pin df604e83d
Peter Rabbitson [Wed, 18 Mar 2015 12:41:22 +0000]
(travis) Revert LMU pin df604e83d

9 years agoRun the entire test suite under replicated SQLite on DBICTEST_VIA_REPLICATED
Peter Rabbitson [Tue, 17 Mar 2015 15:31:56 +0000]
Run the entire test suite under replicated SQLite on DBICTEST_VIA_REPLICATED

There isn't real replication, the reader and writer in fact talk to the same
on-disk file. Still this is effective enough to put into CI.

9 years agoFix exception on complex update/delete under a replicated setup
Peter Rabbitson [Tue, 17 Mar 2015 15:21:47 +0000]
Fix exception on complex update/delete under a replicated setup

THIS COMMIT IS A HACK!!! The actual fix is migration of the logic to the
SQLMaker subclass and deal with it there. However the entire class hierarchy
is currently in flux, so go with the least invasive change until the storage
rewrite settles.

Testing for this will show up in the next commit

9 years agoSome stylistic test changes in preparation for next commits
Peter Rabbitson [Tue, 17 Mar 2015 15:09:44 +0000]
Some stylistic test changes in preparation for next commits

There are zero functional changes - everything has to do with making sure
we are not blindly poking into storage objects which may be changed to
delegates AND to make sure we do not leave any open cursors, which would show
up as deadlocks on multiple connections to the same SQLite file

Read under -w

9 years ago(travis) Temporary pin LMU, work around RT#102853
Peter Rabbitson [Wed, 18 Mar 2015 08:25:15 +0000]
(travis) Temporary pin LMU, work around RT#102853

9 years agoFix test suite to work again with DBICTEST_SQLITE_USE_FILE
Peter Rabbitson [Tue, 17 Mar 2015 12:06:13 +0000]
Fix test suite to work again with DBICTEST_SQLITE_USE_FILE

It was never included as part of POISON_ENV and subsequently lapsed. Sigh.

9 years agoClarify name of guard function from 3d56e026, add it globally
Peter Rabbitson [Tue, 17 Mar 2015 11:29:18 +0000]
Clarify name of guard function from 3d56e026, add it globally

Instead of just marking all my DESTROYs with the function, just add it to a
base class, thus capturing *everything* there is to find. Yes it will be a
tad slower. And yes it will have a massive benefit to any DBIC user happening
to be caught in an unfriendly app-space

9 years ago'undef =>' isn't what one would want
Peter Rabbitson [Mon, 16 Mar 2015 14:56:53 +0000]
'undef =>' isn't what one would want

9 years agoRelax sanity check in _resolve_relationship_condition
Peter Rabbitson [Fri, 13 Mar 2015 12:43:08 +0000]
Relax sanity check in _resolve_relationship_condition

It doesn't buy much and breaks legitimate class redirection techniques
(some of them even documented in the cookbook)

9 years agoForgotten test adjustment in bd510251
Peter Rabbitson [Thu, 12 Mar 2015 10:06:09 +0000]
Forgotten test adjustment in bd510251

9 years agoThe real workaround for txn_scope_guard being called twice
Peter Rabbitson [Thu, 12 Mar 2015 09:27:36 +0000]
The real workaround for txn_scope_guard being called twice

Silently fixing this up is nothing short of irresponsible, hence the
elaborate detection and alert mechanism

9 years agoRemove erroneous undefer introduced in 8d73fcd4 (confuses old perls in END)
Peter Rabbitson [Mon, 9 Mar 2015 10:25:42 +0000]
Remove erroneous undefer introduced in 8d73fcd4 (confuses old perls in END)

While it is important to check this during namespace cleansing checks, it
actually makes no difference whatsoever during leaktests - things will either
be closed over and thus seen, or they won't be

9 years agoGet the leaktest to work with Test::More-(hopefully-not)-to-be
Peter Rabbitson [Thu, 26 Feb 2015 22:25:39 +0000]
Get the leaktest to work with Test::More-(hopefully-not)-to-be

9 years agoFix brainfart from cb551b07 - 'if' is still a module, don't need it
Peter Rabbitson [Thu, 12 Mar 2015 09:36:47 +0000]
Fix brainfart from cb551b07 - 'if' is still a module, don't need it

Also rewrite the load pattern test to be saner

9 years agoMerge branch 'current/for_cpan_index' into current/blead
Peter Rabbitson [Wed, 11 Mar 2015 04:54:34 +0000]
Merge branch 'current/for_cpan_index' into current/blead

9 years agoMore desc_env fixes: Win32 lib matching, and better formatting of final diag
Peter Rabbitson [Thu, 5 Mar 2015 23:39:00 +0000]
More desc_env fixes: Win32 lib matching, and better formatting of final diag

This is what the slash-situation is on a typical strawberry:

          'inc' => [
                     'C:/Strawberry/perl/site/lib/MSWin32-x64-multi-thread',
                     'C:/Strawberry/perl/site/lib',
                     'C:/Strawberry/perl/vendor/lib',
                     'C:/Strawberry/perl/lib',
                     '.'
                   ],
          'archlib' => 'C:\\STRAWB~1\\perl\\lib',
          'archlibexp' => 'C:\\STRAWB~1\\perl\\lib',
          'privlib' => 'C:\\STRAWB~1\\perl\\lib',
          'privlibexp' => 'C:\\STRAWB~1\\perl\\lib',
          'sitearch' => 'C:\\STRAWB~1\\perl\\site\\lib',
          'sitearchexp' => 'C:\\STRAWB~1\\perl\\site\\lib',
          'sitelib' => 'C:\\STRAWB~1\\perl\\site\\lib',
          'sitelibexp' => 'C:\\STRAWB~1\\perl\\site\\lib',
          'vendorarch' => 'C:\\STRAWB~1\\perl\\vendor\\lib',
          'vendorarchexp' => 'C:\\STRAWB~1\\perl\\vendor\\lib',
          'vendorlib' => 'C:\\STRAWB~1\\perl\\vendor\\lib',
          'vendorlibexp' => 'C:\\STRAWB~1\\perl\\vendor\\lib'

Instead of trying to wade through it with File::Spec, just turn short names
into longnames and every \ into / and hope it works.

Also make it so that the MD5 line will align correctly on an 80-col terminal
(on wider terminals it won't matter either way), and show *all* the hashes
at all times (omitting the information makes little sense).

In addition some clarification to the undef/empty ->VERSION return error
and the diag contents themselves (Mithaldu++)

9 years agoRewire OptDeps to not attempt any module loads under missing envvars
Peter Rabbitson [Fri, 6 Mar 2015 16:08:17 +0000]
Rewire OptDeps to not attempt any module loads under missing envvars

The thinking is that if some envvars are missing, it is of secondary concern
to provide the exact list of missing module reqs to the user, while the
downside can be rather severe in some cases (broken DBD::Oracle on windows is
particularly obnoxious). Thus the logic for req_missing_for switched to only
look for "definitely missing" modules via a simple @INC-walk whenever it is
already clear that the req group failed

9 years agoFix t/54taint.t failures on perl-in-hell-in-space
Peter Rabbitson [Thu, 5 Mar 2015 19:14:12 +0000]
Fix t/54taint.t failures on perl-in-hell-in-space

Given its a core bug, and the test is not critical, just skip the whole thing

9 years agoAdd comprehensive 'report-prereqs'-like tooling
Peter Rabbitson [Wed, 4 Mar 2015 01:30:48 +0000]
Add comprehensive 'report-prereqs'-like tooling

Unlike what is currently available in various Dzil plugins, this is more of
a descendent of Test::DiagINC - it tries to load as much as possible and
subsequently reports on whatever can be found in the symtable.

There is no (and never will be any) attempt made to validate if all modules
listed in the various METAs are in fact available (this is what the rest of
the test suite is for).

Module versions are broken up into groups depending on their place in @INC,
without leaking the contents of @INC itself.

9 years agoExtract LeakTracer symtable visitor into DBICTest::Util
Peter Rabbitson [Tue, 3 Mar 2015 17:07:48 +0000]
Extract LeakTracer symtable visitor into DBICTest::Util

Minor rewrapping and renaming of variables as we copy it

No functional changes at all, verified to produce same number of visits

9 years ago(travis) Make a separate is_ci runmode, separated from is_smoker
Peter Rabbitson [Thu, 5 Mar 2015 17:59:48 +0000]
(travis) Make a separate is_ci runmode, separated from is_smoker

9 years ago(travis) shunt intermittent failures stopping postgres
Peter Rabbitson [Thu, 5 Mar 2015 18:11:31 +0000]
(travis) shunt intermittent failures stopping postgres

9 years agoStop noise from a particular combination of DateTime and DateTime::TimeZone
Peter Rabbitson [Thu, 5 Mar 2015 15:07:13 +0000]
Stop noise from a particular combination of DateTime and DateTime::TimeZone

While things went quiet for a while after 556c4fe6, it seems that another
refactor made things go awry again. I can't actually reproduce this on any
of my machines, including Travis. Besides this isn't anything to worry much
about anyway - just kill the test

9 years agoAt last a safer Moo (lots of kludges undone)
Peter Rabbitson [Thu, 26 Feb 2015 16:47:29 +0000]
At last a safer Moo (lots of kludges undone)

9 years agoRemove warning introduced in 75a1d824d
Peter Rabbitson [Mon, 2 Mar 2015 06:08:00 +0000]
Remove warning introduced in 75a1d824d

9 years ago(travis) Only consider the runmode a 'Smoker' when we CI ourselves
Peter Rabbitson [Mon, 2 Mar 2015 06:02:24 +0000]
(travis) Only consider the runmode a 'Smoker' when we CI ourselves

9 years ago(travis) Update apt cache when doing a large set of installs
Peter Rabbitson [Sat, 28 Feb 2015 15:57:39 +0000]
(travis) Update apt cache when doing a large set of installs

Started failing without it:
https://s3.amazonaws.com/archive.travis-ci.org/jobs/52536716/log.txt

9 years agoMove a number of tests to xt, restructure extra lists
Peter Rabbitson [Thu, 26 Feb 2015 18:46:47 +0000]
Move a number of tests to xt, restructure extra lists

Instead of the rigid t / xt separation, add more nuance and move some heavy
tests around, in order to improve regular user test times, without
sacrificing cpantesters coverage

Additionally makes it more obvious which tests are decidedly not being run
on user-side installs

Minimal changes: read diff under -C -C -M

9 years agoFactor out author count anouncement, making it seen on plain installs
Peter Rabbitson [Thu, 26 Feb 2015 18:22:22 +0000]
Factor out author count anouncement, making it seen on plain installs

9 years agoThis hasn't failed in over a year, enable for plain installs
Peter Rabbitson [Thu, 26 Feb 2015 16:18:30 +0000]
This hasn't failed in over a year, enable for plain installs

9 years agoMerge branch 'current/for_cpan_index' into current/blead
Peter Rabbitson [Sun, 15 Feb 2015 10:55:36 +0000]
Merge branch 'current/for_cpan_index' into current/blead

9 years agoRewrite hot-ish _dbi_attrs_for_bind codepath in a leaner way
Peter Rabbitson [Sun, 15 Feb 2015 10:08:01 +0000]
Rewrite hot-ish _dbi_attrs_for_bind codepath in a leaner way

Document various caveats better

9 years agoRevert 2131aa2, cf52a9a and 2aa3f4c
Peter Rabbitson [Sun, 15 Feb 2015 10:29:46 +0000]
Revert 2131aa2, cf52a9a and 2aa3f4c

The warning will be fixed by roper integration with the blockrunner in further
commits

9 years agoRoll back massive unrelated test changeset introduced in 74b5397c
Peter Rabbitson [Wed, 11 Feb 2015 12:54:36 +0000]
Roll back massive unrelated test changeset introduced in 74b5397c

The actual fix/test are ok (and pushed to main as b00d3209), they just
touched too many other unrelated pieces

9 years agoRevert c8dc7d33 - entirely superseded by 83eef5621
Peter Rabbitson [Wed, 11 Feb 2015 16:31:38 +0000]
Revert c8dc7d33 - entirely superseded by 83eef5621

9 years agoRevert 2c2bc4e5 - it is entirely superseded by cb551b07, 2baba3d9 and 83eef562
Peter Rabbitson [Wed, 11 Feb 2015 13:13:28 +0000]
Revert 2c2bc4e5 - it is entirely superseded by cb551b07, 2baba3d9 and 83eef562

<ilmari> if all the skipping already happens before DBICTest(::Schema) is
         loaded, I have no objection to reverting that

9 years agoFix updating multiple CLOB/BLOB columns on Oracle
Kevin L. Kane [Tue, 18 Nov 2014 13:02:33 +0000]
Fix updating multiple CLOB/BLOB columns on Oracle

The genric _dbi_attrs_for_bind caches the attribute hashrefs by data
type, so we can't modify them directly with column-specific data.

Instead, copy it and add the ora_field attribute to the copy.

9 years ago(travis) Run a couple tests in single-threaded mode
Peter Rabbitson [Tue, 10 Feb 2015 18:16:03 +0000]
(travis) Run a couple tests in single-threaded mode

Semi-dogfooding - the massive paralellism hides the pain of how long exactly
does it take for a common user to install DBIC

(FWIW *dependency* installer timeout had to be bumped from 9 to 15 *MINUTES*)

9 years ago(travis) Temporarily kill Firebird ODBC testing
Peter Rabbitson [Tue, 10 Feb 2015 18:12:15 +0000]
(travis) Temporarily kill Firebird ODBC testing

Sourceforge no longer facilitates downloads without a javascript enabled
browser (WAT?!), so the Odbc driver download fails. Disable the testing for
the time being, to be reinstated with the DB2 testing work

9 years agoSwitch CDBICompat and its tests to OptDeps
Peter Rabbitson [Tue, 10 Feb 2015 15:17:55 +0000]
Switch CDBICompat and its tests to OptDeps

This work is a direct descendant of ilmari's c8dc7d33, even though it does
not contain almost any of the original code. While the original approach was
a big improvement, this commit is taking advantage of all improvements that
took place in OptDep handling in earlier commits. The result (at the expense
of a much much larger changeset) is a skip time of < 0.5s for the entire cdbi
set

ilmari++ # this revamp would not have taken place without your first nudge

9 years agoRemove redundant skip checks in CDBICompat tests
Peter Rabbitson [Tue, 10 Feb 2015 13:24:27 +0000]
Remove redundant skip checks in CDBICompat tests

f56e59cd added the extra module, but nothing in the tests actually loads it
under Film (there likely was a switch during dev, but it was never committed).
Just rip it out entirely.

9 years agoRemove useless INIT blocks from CDBI tests - no changes
Peter Rabbitson [Sun, 25 Jan 2015 10:35:23 +0000]
Remove useless INIT blocks from CDBI tests - no changes

A use() is compile-time even when it is in an INIT block, hence removing the
INITs is safe. Do not do it for require() invocations, as it is not clear if
the tests rely on the specific order or not (more sleeping dogs left to lie)

Read under -w

9 years agoIntroduce ad hoc requirements and add skip_without method to optdeps
Peter Rabbitson [Tue, 10 Feb 2015 12:45:32 +0000]
Introduce ad hoc requirements and add skip_without method to optdeps

On rare occasions tests need a specific module version, but nothing else in
the dist needs this. Additionally only parts of a test may rely on this extra
requirement, making the "all or nothing" approach of -skip_all_without not
very attractive.

Hence introducing ability to supply arbitrary module specifications in lieu of
group names. Each specification is checked against a special group 'test_adhoc'
which ensures that optional deps are not "forgotten" within the suite (causing
a test to never run in practice).

9 years agoFinally implement compound OptDep group augmentation
Peter Rabbitson [Tue, 10 Feb 2015 08:38:51 +0000]
Finally implement compound OptDep group augmentation

A group can specify not only what it requires (with includes), but can also
request a specific set of modules IFF another group is required within the
same call. This allows to do e.g. IC::DT+RDBMS dependencies in a sane manner.

Delete test_dt* groups from our optdeps (they were never documented, and a
cpangrep does not show any use) - instead we replace them with an elaborate
network of augments.

9 years agoRewrite internals of the optdep include mechanism (introduced in e163f484b)
Peter Rabbitson [Fri, 6 Feb 2015 11:37:29 +0000]
Rewrite internals of the optdep include mechanism (introduced in e163f484b)

No functional changes, "simply" moves the include-parsing logic and the env
checks to an earlier point, in order to get a complete list of all groups
mentioned (facilitating the augment functionality in the next commit)

9 years agoIntroduce (empty for now) intermediate optdep groups related to IC::DT
Peter Rabbitson [Sun, 8 Feb 2015 11:33:28 +0000]
Introduce (empty for now) intermediate optdep groups related to IC::DT

They will be filled in during the next commits, done separately to reduce
diff-noise

9 years agoRelax the optdep naming criteria, introduce _-prefixed intermediate groups
Peter Rabbitson [Fri, 6 Feb 2015 18:37:54 +0000]
Relax the optdep naming criteria, introduce _-prefixed intermediate groups

9 years agoMove the SQLite test dependency specification to optdeps
Peter Rabbitson [Mon, 9 Feb 2015 23:05:01 +0000]
Move the SQLite test dependency specification to optdeps

Even though this is a hard test_requires (for now), move things to a proper
specification within OptDeps like all other RDBMS (will make icdt specs
much easier).

The OptDep module is very lightweight anyway, and loading it on all installs
will be a good exercise before spinning it off.

9 years agoStabilize and further test complex optdep include scenarious
Peter Rabbitson [Fri, 6 Feb 2015 10:39:44 +0000]
Stabilize and further test complex optdep include scenarious

9 years agoClarify optdep API signatures to be explicitly single-argument
Peter Rabbitson [Fri, 6 Feb 2015 10:06:01 +0000]
Clarify optdep API signatures to be explicitly single-argument

9 years agoRemove DateTime requirement from tests that do not rely on it
Peter Rabbitson [Sat, 7 Feb 2015 16:02:02 +0000]
Remove DateTime requirement from tests that do not rely on it

The deterministic_value test checks how stringification behaves wrt inherited
values from a resultset. Rename test while we are at it.

The plus_select.t checks behavior of InflateColumn combined with get_columns.

In both cases the column in question (year) is declared as VarChar and is not
connected with IC::DT in any way

Read under -C -C -M -w

9 years agoMove some ICDT dependent (sub)tests around (no functional changes)
Peter Rabbitson [Sat, 7 Feb 2015 15:02:26 +0000]
Move some ICDT dependent (sub)tests around (no functional changes)

Read diff/blame with -C -C -M -w

9 years agoFix overspecified msaccess test optdep (should not include icdt)
Peter Rabbitson [Sun, 8 Feb 2015 10:11:00 +0000]
Fix overspecified msaccess test optdep (should not include icdt)

The test_rdbms_msaccess_* groups are only referenced in two tests. The basic
one does not exercise (doesn't even load) any of the ICDT handling components

The ICDT test already explicitly requires test_dt, so no changes are needed
within the actual tests

The erroneous requirement was introduced in 726c8f65 and never revised since

9 years agoRemove obsolete startup check (introduced in 5e724964, superseded by a5a7bb73)
Peter Rabbitson [Tue, 10 Feb 2015 13:05:40 +0000]
Remove obsolete startup check (introduced in 5e724964, superseded by a5a7bb73)

9 years agoAdd import-time-skip support to OptDeps, switch most tests over to that
Peter Rabbitson [Fri, 17 Oct 2014 09:35:46 +0000]
Add import-time-skip support to OptDeps, switch most tests over to that

This is the meat of this refactor, inspired by multiple parties complaining
(with good reason) about the sluggishness of our skip_all()s, and by groundwork
in both c8dc7d333 and 2c2bc4e5 (both never applied - the former is superseded
by a more direct approach and the latter being too intrusive with too little
benefit after the refactor)

Only tests with unambiguous skips were switched over - later commits will
address multi-rdbms loops in a different manner.

The loss of skip-time warnings "this test creates tables blah and blah" was
deliberate, and due to several reasons

*) They are displayed at the wrong time
*) They are not well sycnrhonized with what the actual test does
*) Some time this year all table names will switch to dbictest_-prefixing

All in all the thinking is that nothing of value is lost, and the changeset
is greatly simplified as a result.

The loading before strict/warning in tests is actually significant. It appears
that in "all deps fail" cases the speedup of not loading strictures is quite
significant (demonstrated below difference on WiP CDBICompat streamlining):

strictures first:
Files=43, Tests=0,  1 wallclock secs ( 0.14 usr  0.08 sys +  0.35 cusr  0.15 csys =  0.72 CPU)

optdeps first:
Files=43, Tests=0,  1 wallclock secs ( 0.13 usr  0.06 sys +  0.14 cusr  0.11 csys =  0.44 CPU)

9 years agoAdd import-time action stub to OptDeps, switch distbuild checks to it
Peter Rabbitson [Sun, 18 Jan 2015 13:33:28 +0000]
Add import-time action stub to OptDeps, switch distbuild checks to it

More in the following commit

9 years agoIntroduce basic optdeps group inclusion mechanism, simplify many declarations
Peter Rabbitson [Tue, 20 Jan 2015 16:09:44 +0000]
Introduce basic optdeps group inclusion mechanism, simplify many declarations

No logical changes: verified before and after the commit via:

perl -MData::Dumper::Concise -Ilib -MDBIx::Class::Optional::Dependencies -e '
  my $d = DBIx::Class::Optional::Dependencies->req_group_list;
  delete $d->{$_} for grep /rdbms_generic/, keys %$d;
  delete $_->{effective_modreqs_differ} for values %$d;
  print Dumper $d
' | sha1sum

Read under -w

9 years agoMake OptDeps aware of envvars as first-class requirements (test_* groups only)
Peter Rabbitson [Sun, 26 Oct 2014 12:23:28 +0000]
Make OptDeps aware of envvars as first-class requirements (test_* groups only)

9 years agoDeprecate req_errorlist_for, proxy to to modreq_errorlist_for
Peter Rabbitson [Thu, 8 Jan 2015 18:33:57 +0000]
Deprecate req_errorlist_for, proxy to to modreq_errorlist_for

Leaving the old method in place indefinitely - may be used in the wild

9 years agoRewrite optdeps to accept a list of groups
Peter Rabbitson [Fri, 17 Oct 2014 09:27:27 +0000]
Rewrite optdeps to accept a list of groups

Requires a one-time break of req_group_list()
Also tighten up the optdep spec and handle conflicting versions saner

As a side effect fix 34d2deae to be truly copy-paste-able (no , separator)

Changes to ::Opt::Deps and xt/optional_deps.t best viewed under -w

9 years agoOptDeps doc-mangling, no functional changes
Peter Rabbitson [Sat, 17 Jan 2015 13:23:44 +0000]
OptDeps doc-mangling, no functional changes

Switch synopsis away from M::I (it still scares people, boggle)
Document volatility of req_group_list
Better language in places
Coarser granularity of heredocs (easier to read)

Read changes under -w

9 years agoMinor streamline of optdeps usage (no functional changes)
Peter Rabbitson [Sun, 18 Jan 2015 22:22:39 +0000]
Minor streamline of optdeps usage (no functional changes)

9 years agoStandardize the struct-cloning interface throughout the codebase
Peter Rabbitson [Sun, 25 Jan 2015 09:59:55 +0000]
Standardize the struct-cloning interface throughout the codebase

9 years agoNeither of these can fail now that we dep on 5.8.1 (missed in 55087b99)
Peter Rabbitson [Sat, 24 Jan 2015 19:53:22 +0000]
Neither of these can fail now that we dep on 5.8.1 (missed in 55087b99)

9 years agoDelete useless test
Peter Rabbitson [Sat, 24 Jan 2015 17:23:23 +0000]
Delete useless test

9 years agoFix intermittent failures in the LeakTracer on 5.18+, remove all workarounds
Peter Rabbitson [Thu, 22 Jan 2015 22:27:22 +0000]
Fix intermittent failures in the LeakTracer on 5.18+, remove all workarounds

ARGH! In the end the issue turned out to be stupid-simple: when PadWalker is
called it always returns us a hashref. This hashref is traversed, with its
own address recorded as seen. Once processed, the hashref itself is GCed,
leaving a stale entry in $seen_refs.

Up until 5.18 this was not a problem, likely due to no address reuse. However
after 5.18 PadWalker is very happy to return different hashrefs under the same
address over and over again. The $seen_refs entry leads to a skip, and the
actual contents of the pad are never examined - we've seen this hash already.

ARGH!

The decisive change is the single line around PadWalker - everything else is
cleanup of workarounds for this problem.

9 years agoKeep T::B IO open in t/52leaks.t END - other ENDs pending (esp. since 961d79db)
Peter Rabbitson [Thu, 22 Jan 2015 15:19:57 +0000]
Keep T::B IO open in t/52leaks.t END - other ENDs pending (esp. since 961d79db)

Read under -w

9 years agoFix Postgres test hang with DBD::Pg 3.5.0+ (work around RT#100648)
Peter Rabbitson [Thu, 8 Jan 2015 00:42:52 +0000]
Fix Postgres test hang with DBD::Pg 3.5.0+ (work around RT#100648)

It is not 100% clear to me why this is happening, as the change discussed[1]
was not supposed to change behaviour outside of a transaction.

The test in question essentially does a blocking request with a timeout. The
way the default DBIC machinery works is to issue a ping() after the timeout
induced exception, in order to opossibly retry the operation IFF the link
went stale ($dbh thinks everything is in order, the RDBMS believes otherwise).

After the changes in 3.5.0 this ping() now hangs, even though the access
pattern did not change (there are no transactions involved on the handle in
question). Either way, the fix is simple: just shim-away the ping for the
duration of the test.

In general this should be resolved on a higher level by disabling retries for
the duration of the test (or maybe even disabling retries by default in the
distant feature - see RT#47005)

[1] https://rt.cpan.org/Ticket/Display.html?id=100648#txn-1438093

9 years ago(travis) Remove trial SQLA workaround (no longer on CPAN) - retry another time
Peter Rabbitson [Sun, 11 Jan 2015 09:39:48 +0000]
(travis) Remove trial SQLA workaround (no longer on CPAN) - retry another time

Read diff under -w

9 years ago(travis) Revert 17b09b77 - Test::Deep 0.115 now available
Peter Rabbitson [Sun, 11 Jan 2015 09:45:08 +0000]
(travis) Revert 17b09b77 - Test::Deep 0.115 now available

https://github.com/rjbs/Test-Deep/issues/22

9 years agoPreserve space after <nick>: in the AUTHORS section of DBIx/Class.pod
Dagfinn Ilmari Mannsåker [Tue, 6 Jan 2015 19:24:05 +0000]
Preserve space after <nick>: in the AUTHORS section of DBIx/Class.pod

9 years agoPreserve space after <nick>: in AUTHORS POD
Dagfinn Ilmari Mannsåker [Tue, 6 Jan 2015 19:24:05 +0000]
Preserve space after <nick>: in AUTHORS POD

9 years agoEmit Optional::Dependencies error messages in a way simplifying c/p for cpanm
Peter Rabbitson [Sat, 3 Jan 2015 16:30:14 +0000]
Emit Optional::Dependencies error messages in a way simplifying c/p for cpanm

The contrived no-space-containing output is sadly necessary - cpanm only
considers an optspec if the module name is directly followed by ~

rabbit@Ahasver:~$ cpanm "SQL::Translator~ >= 0.11018"
SQL::Translator is up to date. (0.11020)
rabbit@Ahasver:~$ cpanm "SQL::Translator~ >=0.11018"
SQL::Translator is up to date. (0.11020)
rabbit@Ahasver:~$ cpanm "SQL::Translator ~ >= 0.11018"
--> Working on SQL::Translator
^C

9 years agoMore robust 'nothing loaded' test for optdeps (read under -w)
Peter Rabbitson [Sat, 3 Jan 2015 12:16:25 +0000]
More robust 'nothing loaded' test for optdeps (read under -w)

9 years agoReshuffle optdeps code making next commits more readable (no func. changes)
Peter Rabbitson [Fri, 2 Jan 2015 13:43:35 +0000]
Reshuffle optdeps code making next commits more readable (no func. changes)

Need to switch the entire POD generator to something else before going solo

9 years ago(travis) randomize the dependency install order to catch even more fish
Peter Rabbitson [Fri, 2 Jan 2015 15:56:26 +0000]
(travis) randomize the dependency install order to catch even more fish

9 years ago(travis) remove two more workarounds for issues that have since been resolved
Peter Rabbitson [Fri, 2 Jan 2015 12:23:00 +0000]
(travis) remove two more workarounds for issues that have since been resolved

https://rt.cpan.org/Ticket/Display.html?id=99747
https://github.com/travis-ci/travis-ci/issues/1477

9 years ago(travis) stop trying to dynamically determine the CPAN mirror
Peter Rabbitson [Mon, 29 Dec 2014 14:59:01 +0000]
(travis) stop trying to dynamically determine the CPAN mirror

Even if Travis brings their mirror back, we are not going to be able to
use it - we need backpan at this point

9 years ago(travis) minimize mysqld's memory footprint
Peter Rabbitson [Mon, 29 Dec 2014 11:49:49 +0000]
(travis) minimize mysqld's memory footprint

This implements a part of "solution 2)" described in ac4e80df. Historically
the OOM issues so far have always been somehow related to mysqld

Unfortunately this alone does not seem sufficient to raise the thread count
back up, that will need more minimizations

9 years ago(travis) Invoke apt_install only once and do not change the source list
Peter Rabbitson [Sun, 28 Dec 2014 12:46:27 +0000]
(travis) Invoke apt_install only once and do not change the source list

This shaves off a surprising amount of time

9 years ago(travis) Document app-nopaste workaround better
Peter Rabbitson [Sun, 28 Dec 2014 12:42:01 +0000]
(travis) Document app-nopaste workaround better

9 years ago(travis) Stop trying to get the Oracle apt source - use the debs directly
Peter Rabbitson [Sun, 28 Dec 2014 08:52:24 +0000]
(travis) Stop trying to get the Oracle apt source - use the debs directly