GitLab

In all the feature adaptors, use either _create_feature() or
_create_feature_fast() from BaseFeatureAdaptor in _objs_from_sth() to
create features.

In get_available_adaptors(), add GeneCollection, TranscriptCollection,
ExonCollection, and RepeatFeatureCollection so that one may say, e.g.

  my $rfc = $registry->get_adaptor( 'Human', 'Core',
                                    'RepeatFeatureCollection' );

Add the methods _create_feature() and _create_feature_fast() that will
be called from _objs_from_sth() in the feature adaptors to create
features.

Turn _remap() into a proper method so that we're able to specialize it.

Beautify.

a string that can be 0 (zero).

Similarly use "if (exists( some hash compartment ))" for the same reason.

A couple of changes from "@{ function_returning_array_ref() }($index)"
to "function_returning_array_ref()->[$index]";

General cleanup in comments and code.

Passes the unit tests.

Proper new() for most classes:
  * $class = ref($proto) || $proto;
  * Setting attributes the same way.

Format the getter and setters the same way, especially in
Mapper/Coordinate.pm, getting rid of unneccesary shifts.

sprintf() formatting including '%1$s'.

Minor formatting.

and later into Ensembl release 50.

Added defaults for common usage mode.

#Changed transcript_probeset_count cache to use dbID instead of complete name
#Renamed check_names_match to validate_arrays and extended to perform probeset warning, external_db name guess and returns cache, user array param validation and association check
#Merged delete_unampped objects into delete_existing_xrefs due to dependency. Now throws if arrays param set until force_delete specified, deletes using IN list
#cache_arrays_per_probeset now uses arrays param to reduce size of cache
#removed unmapped_object dbID = 2000>
#Change all probe specific unmapped object to reeflect the individual probe rather than the probeset
#Updated logs
#Updated docs
#Added control of promiscuous probesets and unmapped objects