GitLab

Analysis must have a resource class (because analysis_base.resource_class_id is NOT NULL and a foreign key)

This commit is the result of the pull-request #62 by @mira13
Summary:
* Tweak pipeline output changed to json
* Test for tweak_pipeline output
* Waste changes revert
* Waste hcanges revert
* Init pipeline fix
* Test fix, removed unused tests.
* Tweak pipeline output with tests
* Review changes
* Brandon review changas
* Undef added, minor review fixes
* Error msgs changed, action is guaranteed
* End of support for older Perl versions
  Enabled testing on 5.26, as per our announcement. Hopefully it will work !
* Text msgs spaces aligned.
* responceStructure renamed in tweak_pipeline
* Ids converted to number, unused varible removed.
* Error msgs defined outside tweak_pipeline
* Missing field value filled.
* Medow type msf bug fix.
* Action setup refactoring
* Test fix, field is guaranteed.
* Correct values of json
* += Action is added to shash as SET
* Warning added on value can't be dumped
* Minor sintax fix
  1. Typo fix
  2. Consistent aligment of code
* Syntax and intention fixes.
* Constants rename.
* Field value set refactoring.
* Addition NewValue field includes old value.
* Tweak object type converted to constatnt.
* Exceed tab removed.
* Indentation fix
* Lines order refactoring.
* New val stringify.
* Intentions align.
* Arrow alignned in constant
* Old value print fix
* New value obtained from collection request.
* Reverse rebase wrong changes
* Values for += analysis shown as a list
* Restored the original indentation
* Correctly indent this line
* Added the information message recently introduced on master

(cherry-picked from 2.5: 9112b933)

Bug identified when implementing a JSON output for tweak_pipeline.pl
(see #62) and backported to the earliest branch.

Finally added the copyright statement on all source code files and the test to ensure it is kept up to date

This brings a minor performance improvement.

init_pipeline is now more consistently using stdout and stderr.

init_pipeline should not die if the control-flow is not defined, since it could be made valid later (with a tweak or an analysis topup)

switching to having every HivePipeline object to be automatically registered in TheApiary by default

schema_change: new implementation of Semaphores as separate objects that can cross pipeline database boundaries

This can't be easily patched in sqlite

This way we don't need a read+write user when querying parameters

We noticed that beekeeper was sometimes very slow doing anything on the
database. This was tracked down to running a query over and over (from
RoleAdaptor::get_hive_current_load()).
This commit introduces two methods in HivePipeline: one to fetch + cache the
value, and one to invalidate the cache. Other modules have been adapted to
invalidate the cache only when needed, resulting in an optimal use of the
database.

dark_collection/shadow_object : tracking and subsequently deleting DB representations of objects removed from HivePipeline