AnalysisData entries can be shared between jobs, so we can't delete them...

AnalysisData entries can be shared between jobs, so we can't delete them without checking if they're still used

At the moment, I can't think of a way of checking that without a
transaction / a locking mechanism, which could be slow in large databases
(cf hoover_pipeline.pl which doesn not hoover the analysis_data table for
the same reason).
Since I don't expect many input_ids to be changed this way, I think we can
accept keeping the old entries.

scripts/lib/hive_extended.pm

@@ -151,45 +151,6 @@ use Bio::EnsEMBL::Hive::DBSQL::BaseAdaptor;
   $self->db->get_ResourceDescriptionAdaptor->store($rd);
 };
 
-## _input_id_is_extended determines if the input_id of a job is too large in the
-## database. If it is, returns the extended_data_id
-*Bio::EnsEMBL::Hive::DBSQL::AnalysisJobAdaptor::_input_id_is_extended = sub {
-  my ($self, $job) = @_;
-
-  my $sql = "SELECT input_id FROM job WHERE job_id = ?";
-  my $sth = $self->prepare($sql);
-  $sth->execute($job->dbID);
-  my ($real_input_id) = $sth->fetchrow_array();
-  $sth->finish();
-  if ($real_input_id =~ /^_ext(?:\w+)_data_id (\d+)$/) {
-    return $1;
-  }
-  return 0; # 0 is not used as an extended_data_id
-};
-
-## If the input_id of a job has grown over its size limit (255 characters)
-## _move_input_id_to_analysis_data puts it into analysis_data and update the job's input_id
-## to link to that entry in analysis_data
-## The method returns the newly created dbID in the analysis_data table
-*Bio::EnsEMBL::Hive::DBSQL::AnalysisJobAdaptor::_move_input_id_to_analysis_data = sub {
-  my ($self, $job) = @_;
-  my $extended_data_id = $self->db->get_AnalysisDataAdaptor->store($job->input_id);
-  my $extended_input_id = "_extended_data_id $extended_data_id";
-  $self->_update_input_id_in_job_table($job, $extended_input_id);
-  return $extended_data_id;
-};
-
-
-*Bio::EnsEMBL::Hive::DBSQL::AnalysisJobAdaptor::_move_input_id_from_analysis_data = sub {
-  my ($self, $job, $extended_data_id) = @_;
-  my $input_id = $job->input_id;
-  $self->throw("input_id is too large to be moved into the job table (it has to remain in analysis_data)")
-    if (length($input_id)>=255);
-  $self->db->get_AnalysisDataAdaptor->remove($extended_data_id);
-  $self->_update_input_id_in_job_table($job, $input_id);
-  return;
-};
-
 *Bio::EnsEMBL::Hive::DBSQL::AnalysisJobAdaptor::_update_input_id_in_job_table = sub {
   my ($self, $job, $input_id) = @_;
 
@@ -200,27 +161,6 @@ use Bio::EnsEMBL::Hive::DBSQL::BaseAdaptor;
   return;
 };
 
-## update updates the corresponding entry in analysis_data with the values
-## provided as arguments
-*Bio::EnsEMBL::Hive::DBSQL::AnalysisDataAdaptor::update = sub {
-  my ($self, $analysis_data_id, $data) = @_;
-  my $sql = "UPDATE analysis_data SET data = ? WHERE analysis_data_id = ?";
-  my $sth = $self->prepare($sql);
-  $sth->execute($data, $analysis_data_id);
-  $sth->finish();
-  return;
-};
-
-## remove deletes the corresponding entry fort the given analysis_data_id in the
-## analysis_data table
-*Bio::EnsEMBL::Hive::DBSQL::AnalysisDataAdaptor::remove = sub {
-  my ($self, $analysis_data_id) = @_;
-  my $sql = "DELETE FROM analysis_data WHERE analysis_data_id = ?";
-  my $sth = $self->prepare($sql);
-  $sth->execute($analysis_data_id);
-  return;
-};
-
 ## AnalysisJobAdaptor doesn't have a generic update method
 ## and it doesn't inherits from Hive::DBSQL::BaseAdaptor either
 ## I guess that this will change in the future,
@@ -230,21 +170,12 @@ use Bio::EnsEMBL::Hive::DBSQL::BaseAdaptor;
 
   return Bio::EnsEMBL::Hive::DBSQL::BaseAdaptor::update($self, $job, $_[2]) if $_[2];
 
-  my $curr_data_id = $self->_input_id_is_extended($job);
-
   if (length($job->input_id) >= 255) {
-    if ($curr_data_id) {
-      $self->db->get_AnalysisDataAdaptor->update($curr_data_id, $job->input_id);
-    } else {
-      $self->_move_input_id_to_analysis_data($job)
-    }
-  } else {
-    if ($curr_data_id) {
-      $self->_move_input_id_from_analysis_data($job, $curr_data_id);
-    } else {
-      $self->_update_input_id_in_job_table($job, $job->input_id);
-    }
+    my $extended_input_id = $self->db->get_AnalysisDataAdaptor->store_if_needed($job->input_id);
+    $job->input_id($extended_input_id);
   }
+  $self->_update_input_id_in_job_table($job, $job->input_id);
+
   my $sql = "UPDATE job SET ";
   $sql .= "status='" . $job->status . "'";
   $sql .= ",retry_count=" . $job->retry_count;