Finishes work on not using the ParsingState objects as the identifier…

… of a parsing process but an empty token object. Should reduce memory footprint even further and tackle for example #66.

packages/Ohm-Core.package/OhmPExprApply.class/instance/ruleBodyOn..st

@@ -1,10 +1,10 @@
 accessing
 ruleBodyOn: state
 
-	cachedRuleBodyGrammar ~= state  
+	cachedRuleBodyParseId ~= state idOfParseProcess  
 		ifTrue: [cachedRuleBody := nil].	
 	cachedRuleBody ifNil: [
 		cachedRuleBody := (state grammar ruleNamed: ruleName asSymbol)
 			replaceParametersWithArguments: self arguments.
-		cachedRuleBodyGrammar := state.].
+		cachedRuleBodyParseId := state idOfParseProcess.].
 	^ cachedRuleBody

packages/Ohm-Core.package/OhmPExprApply.class/methodProperties.json

@@ -15,7 +15,7 @@
 		"innerNodesAt:" : "pre 7/20/2015 14:39",
 		"nodesResultingInCST" : "pre 6/1/2015 14:48",
 		"replaceParametersWithArguments:" : "pre 11/17/2017 20:48",
-		"ruleBodyOn:" : "pre 7/26/2019 10:26",
+		"ruleBodyOn:" : "pre 5/4/2021 17:16",
 		"ruleName" : "pre 12/3/2014 16:53",
 		"ruleName:" : "pre 12/3/2014 16:53",
 		"toRecipe" : "pre 11/17/2017 20:20" } }

packages/Ohm-Core.package/OhmPExprApply.class/properties.json

@@ -9,7 +9,6 @@
 		"ruleName",
 		"arguments",
 		"cachedRuleBody",
-		"cachedRuleBodyGrammar",
 		"cachedRuleBodyParseId" ],
 	"name" : "OhmPExprApply",
 	"pools" : [

packages/Ohm-Core.package/OhmParsingState.class/instance/idOfParseProcess.st

@@ -1,4 +1,4 @@
 accessing
 idOfParseProcess
 
-	^ idOfParseProcess ifNil: [self] 
+	^ idOfParseProcess ifNil: [idOfParseProcess := Object new] 

packages/Ohm-Core.package/OhmParsingState.class/methodProperties.json

@@ -9,7 +9,7 @@
 		"cachedSpacesRule" : "pre 6/3/2019 17:02",
 		"grammar" : "pre 1/13/2015 10:40",
 		"grammar:" : "pre 1/13/2015 10:40",
-		"idOfParseProcess" : "pre 6/4/2019 18:21",
+		"idOfParseProcess" : "pre 5/4/2021 17:15",
 		"idOfParseProcess:" : "pre 6/4/2019 18:07",
 		"initialize" : "pre 1/21/2015 10:31",
 		"inputStream" : "pre 1/13/2015 12:35",