Creatures are now cached in memory for computation

Also removed dependency group calculation because the optimisation isn't
as useful to reduce DB calls if the creature is in memory anyway

app/imports/server/publications/singleCharacter.js

@@ -5,6 +5,7 @@ import CreatureLogs from '/imports/api/creature/log/CreatureLogs.js';
 import { assertViewPermission } from '/imports/api/creature/creatures/creaturePermissions.js';
 import computeCreature from '/imports/api/engine/computeCreature.js';
 import VERSION from '/imports/constants/VERSION.js';
+import { loadCreature } from '/imports/api/engine/loadCreatures.js';
 
 let schema = new SimpleSchema({
   creatureId: {
@@ -13,7 +14,8 @@ let schema = new SimpleSchema({
   },
 });
 
-Meteor.publish('singleCharacter', function(creatureId){
+Meteor.publish('singleCharacter', function (creatureId) {
+  const self = this;
   try {
     schema.validate({ creatureId });
   } catch (e){
@@ -21,21 +23,24 @@ Meteor.publish('singleCharacter', function(creatureId){
   }
   this.autorun(function (computation){
     let userId = this.userId;
-    let creatureCursor
-    creatureCursor = Creatures.find({
+    let permissionCreature = Creatures.findOne({
       _id: creatureId,
+    }, {
+      fields: { owner: 1, readers: 1, writers: 1, public: 1, computeVersion: 1 }
     });
-    let creature = creatureCursor.fetch()[0];
-    try { assertViewPermission(creature, userId) }
-    catch(e){ return [] }
-    if (creature.computeVersion !== VERSION && computation.firstRun){
+    try { assertViewPermission(permissionCreature, userId) }
+    catch (e) { return [] }
+    loadCreature(creatureId, self);
+    if (permissionCreature.computeVersion !== VERSION && computation.firstRun){
       try {
         computeCreature(creatureId)
       }
       catch(e){ console.error(e) }
     }
     return [
-      creatureCursor,
+      Creatures.find({
+        _id: creatureId,
+      }),
       CreatureProperties.find({
         'ancestors.id': creatureId,
       }),