AnalysisStats.get_analysis/analysis_id are substituted by an autoloaded analysis/analysis_id pair

modules/Bio/EnsEMBL/Hive/AnalysisStats.pm

@@ -49,12 +49,13 @@ sub min_batch_time {
     return 2*60*1000;
 }
 
-## pre-settable storable object's getters/setters:
 
-sub analysis_id {   # an alias
-    my $self = shift;
-    return $self->dbID(@_);
-}
+=head1 AUTOLOADED
+
+    analysis_id / analysis
+
+=cut
+
 
 sub batch_size {
     my $self = shift;
@@ -220,16 +221,6 @@ sub update_status {
     $self->status($status);
 }
 
-sub get_analysis {
-    my $self = shift;
-    unless($self->{'_analysis'}) {
-        unless($self->analysis_id) {
-            throw("self->analysis_id undefined, please investigate");
-        }
-        $self->{'_analysis'} = $self->adaptor->db->get_AnalysisAdaptor->fetch_by_dbID($self->analysis_id);
-    }
-    return $self->{'_analysis'};
-}
 
 sub get_or_estimate_batch_size {
     my $self = shift;
@@ -281,7 +272,7 @@ sub job_count_breakout {
 sub toString {
     my $self = shift @_;
 
-    my $analysis = $self->get_analysis;
+    my $analysis = $self->analysis;
 
     my $output .= sprintf("%-27s(%2d) %11s jobs(Sem:%d, Rdy:%d, InProg:%d, Done+Pass:%d, Fail:%d)=%d Ave_msec:%d, workers(Running:%d, Reqired:%d) ",
         $analysis->logic_name,
@@ -358,7 +349,7 @@ sub determine_status {
             $self->status('EMPTY');
 
         } elsif( $self->total_job_count == $self->done_job_count + $self->failed_job_count ) {   # all jobs of the analysis have been finished
-            my $analysis = $self->get_analysis;
+            my $analysis = $self->analysis;
             my $absolute_tolerance = $analysis->failed_job_tolerance * $self->total_job_count / 100.0;
             if ($self->failed_job_count > $absolute_tolerance) {
                 $self->status('FAILED');

modules/Bio/EnsEMBL/Hive/PipeConfig/HiveGeneric_conf.pm

@@ -576,10 +576,9 @@ sub run {
                 'analysis_capacity'     => $analysis_capacity,
             );
             $analysis->get_compiled_module_name();  # check if it compiles and is named correctly
-            $analysis_adaptor->store($analysis);
 
             my $stats = Bio::EnsEMBL::Hive::AnalysisStats->new(
-                'analysis_id'           => $analysis->dbID,
+                'analysis'              => $analysis,
                 'batch_size'            => $batch_size,
                 'hive_capacity'         => $hive_capacity,
                 'status'                => $blocked ? 'BLOCKED' : 'EMPTY',  # be careful, as this "soft" way of blocking may be accidentally unblocked by deep sync
@@ -595,6 +594,8 @@ sub run {
                 'output_capacity'       => 4,
                 'sync_lock'             => 0,
             );
+
+            $analysis_adaptor->store($analysis);
             $analysis_stats_adaptor->store($stats);
         }
 
@@ -629,8 +630,9 @@ sub run {
                 # create control rules:
             foreach my $condition_url (@$wait_for) {
                 unless ($condition_url =~ m{^\w*://}) {
-                    my $condition_analysis = $analysis_adaptor->fetch_by_logic_name($condition_url);
-                    die "Could not fetch analysis '$condition_url' to create a control rule (in '".($analysis->logic_name)."')\n" unless defined $condition_analysis;
+                        # TODO: this should become a call to Pipeline_object:
+                    my $condition_analysis = $analysis_adaptor->fetch_by_logic_name($condition_url)
+                        or die "Could not fetch analysis '$condition_url' to create a control rule (in '".($analysis->logic_name)."')\n";
                 }
                 my $c_rule = Bio::EnsEMBL::Hive::AnalysisCtrlRule->new(
                         'condition_analysis_url'    => $condition_url,
@@ -680,8 +682,9 @@ sub run {
                 while(my ($heir_url, $input_id_template_list) = each %$heirs) {
 
                     unless ($heir_url =~ m{^\w*://}) {
-                        my $heir_analysis = $analysis_adaptor->fetch_by_logic_name($heir_url);
-                        die "No analysis named '$heir_url' (dataflow from analysis '".($analysis->logic_name)."')\n" unless defined $heir_analysis;
+                            # TODO: this should become a call to Pipeline_object:
+                        my $heir_analysis = $analysis_adaptor->fetch_by_logic_name($heir_url)
+                            or die "No analysis named '$heir_url' (dataflow from analysis '".($analysis->logic_name)."')\n";
                     }
                     
                     $input_id_template_list = [ $input_id_template_list ] unless(ref($input_id_template_list) eq 'ARRAY');  # allow for more than one template per analysis

modules/Bio/EnsEMBL/Hive/Queen.pm

@@ -634,7 +634,7 @@ sub synchronize_AnalysisStats {
 
 
         # compute the number of total required workers for this analysis (taking into account the jobs that are already running)
-    my $analysis              = $analysisStats->get_analysis();
+    my $analysis              = $analysisStats->analysis();
     my $scheduling_allowed    =  ( !defined( $analysisStats->hive_capacity ) or $analysisStats->hive_capacity )
                               && ( !defined( $analysis->analysis_capacity  ) or $analysis->analysis_capacity  );
     my $required_workers    = $scheduling_allowed

modules/Bio/EnsEMBL/Hive/Scheduler.pm

@@ -116,11 +116,11 @@ sub suggest_analysis_to_specialize_by_rc_id_meadow_type {
 sub schedule_workers {
     my ($queen, $submit_capacity, $default_meadow_type, $filter_rc_id, $filter_meadow_type, $filter_analysis, $meadow_capacity_limiter_hashed_by_type, $analysis_id2rc_name) = @_;
 
-    my @suitable_analyses   = $filter_analysis
+    my @suitable_analyses_stats   = $filter_analysis
                                 ? ( $filter_analysis->stats )
                                 : @{ $queen->db->get_AnalysisStatsAdaptor->fetch_all_by_suitability_rc_id_meadow_type($filter_rc_id, $filter_meadow_type) };
 
-    unless(@suitable_analyses) {
+    unless(@suitable_analyses_stats) {
         return $analysis_id2rc_name ? ({}, 0, "Scheduler could not find any suitable analyses to start with\n") : undef;    # FIXME: returns data in different format in "suggest analysis" mode
     }
 
@@ -132,10 +132,10 @@ sub schedule_workers {
     my $submit_capacity_limiter                     = Bio::EnsEMBL::Hive::Limiter->new( 'Max number of Workers scheduled this time', $submit_capacity );
     my $queen_capacity_limiter                      = Bio::EnsEMBL::Hive::Limiter->new( 'Total reciprocal capacity of the Hive', 1.0 - $queen->get_hive_current_load() );
 
-    foreach my $analysis_stats (@suitable_analyses) {
+    foreach my $analysis_stats (@suitable_analyses_stats) {
         last if( $submit_capacity_limiter->reached );
 
-        my $analysis            = $analysis_stats->get_analysis;    # FIXME: if it proves too expensive we may need to consider caching
+        my $analysis            = $analysis_stats->analysis();    # FIXME: if it proves too expensive we may need to consider caching
         my $this_meadow_type    = $analysis->meadow_type || $default_meadow_type;
 
         next if( $meadow_capacity_limiter_hashed_by_type && $meadow_capacity_limiter_hashed_by_type->{$this_meadow_type}->reached );