be more resilient to bad data when updating contributors

When trying to update all contributor data, it will fail because there
are some broken releases in the index. While those should be fixed, they
shouldn't cause the contributor script to fail.

Author: haarg

lib/MetaCPAN/Script/Role/Contributor.pm

@@ -25,9 +25,28 @@ sub update_contributors {
         },
     );
 
-    my $bulk = $self->es->bulk_helper( es_doc_path('contributor') );
+    my $report = sub {
+        my ( $action, $result, $i ) = @_;
+        if ( $i == 0 ) {
+            log_info {'flushing contributor updates'};
+        }
+    };
+
+    my $bulk = $self->es->bulk_helper(
+        es_doc_path('contributor'),
+        on_success => $report,
+        on_error   => $report,
+    );
+
+    log_info { 'updating contributors for ' . $scroll->total . ' releases' };
 
     while ( my $release = $scroll->next ) {
+        my $source = $release->{_source};
+        my $name   = $source->{name};
+        if ( !( $name && $source->{author} && $source->{distribution} ) ) {
+            Dlog_warn {"found broken release: $_"} $release;
+            next;
+        }
         log_debug { 'updating contributors for ' . $release->{_source}{name} };
         my $actions = $self->release_contributor_update_actions(
             $release->{_source} );