GitLab

Revisited 'Drop error on null hash key in accu , PR #177'

Set take_time parameter to 0 - no need to wait here.

Use the /tmp folder for temporary files the pipeline creates. Makes it safer to
create and unlink them.

Remove the sleep time for this pipeline, makes it run a little faster.

Otherwise the test suite fails

Unit test for null hash key

Simplified code, nicer errors for users

Made the bracket finding logic simpler
Give user more information in case of error

Yearly copyright update - 2022 for version/2.5

* version/2.5:
  ignore failures in coverage testing

* version/2.5:
  update documentation related to JSON serialisation of events including dataflow
  Create example PipeConfig using SystemCmd to create dataflow from a JSON file

* version/2.5:
  convert libdbi from apt-get to manual download
  update requirements.txt for readthedocs compatibility

* version/2.5:
  When using a tweak to set a resource description, set the resource class too
  [ENSCORESW-3002] Forbid setting a resource-class that doesn't exist
  Analysis must have a resource class (because analysis_base.resource_class_id is NOT NULL and a foreign key)
  Clear error message if the tweak attemps to change the dbID

Added a test to check that we can create a resource class and use it in the same tweak_pipeline.pl invocation

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

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

* version/2.5:
  rework setting of ensembl_cvs_root_dir in response to user bug report

* version/2.5:
  Fix ensembl_csv_root_dir value setting priorities
  EnsemblGeneric_conf checks for ENSEMBL_ROOT_DIR and ENSEMBL_CSV_ROOT_DIR
  typo
  Don't put the schema definition there, as it may go out of sync (it already has !)