Re-establish caching of results of parsing a tiddler

I switched this optimisation off back in
ed35d91be6, in October 2013, as part of a
big refactoring of the parsing and widget mechanism. I’ve been meaning
to switch it back on for some time.

My rough measurements suggest that this optimisation can reduce
rendering time by 5-10%.

core/modules/parsers/wikiparser/rules/macrodef.js

@@ -78,10 +78,13 @@ exports.parse = function() {
 	}
 	// Save the macro definition
 	return [{
-		type: "macrodef",
-		name: this.match[1],
-		params: params,
-		text: text
+		type: "set",
+		attributes: {
+			name: {type: "string", value: this.match[1]},
+			value: {type: "string", value: text}
+		},
+		children: [],
+		params: params
 	}];
 };
 

core/modules/parsers/wikiparser/wikiparser.js

@@ -48,12 +48,13 @@ var WikiParser = function(type,text,options) {
 	this.blockRules = this.instantiateRules(this.blockRuleClasses,"block",0);
 	this.inlineRules = this.instantiateRules(this.inlineRuleClasses,"inline",0);
 	// Parse any pragmas
-	this.tree = this.parsePragmas();
+	this.tree = [];
+	var topBranch = this.parsePragmas();
 	// Parse the text into inline runs or blocks
 	if(options.parseAsInline) {
-		this.tree.push.apply(this.tree,this.parseInlineRun());
+		topBranch.push.apply(topBranch,this.parseInlineRun());
 	} else {
-		this.tree.push.apply(this.tree,this.parseBlocks());
+		topBranch.push.apply(topBranch,this.parseBlocks());
 	}
 	// Return the parse tree
 };
@@ -122,7 +123,7 @@ WikiParser.prototype.findNextMatch = function(rules,startPos) {
 Parse any pragmas at the beginning of a block of parse text
 */
 WikiParser.prototype.parsePragmas = function() {
-	var tree = [];
+	var currentTreeBranch = this.tree;
 	while(true) {
 		// Skip whitespace
 		this.skipWhitespace();
@@ -137,9 +138,15 @@ WikiParser.prototype.parsePragmas = function() {
 			break;
 		}
 		// Process the pragma rule
-		tree.push.apply(tree,nextMatch.rule.parse());
+		var subTree = nextMatch.rule.parse();
+		if(subTree.length > 0) {
+			// Quick hack; we only cope with a single parse tree node being returned, which is true at the moment
+			currentTreeBranch.push.apply(currentTreeBranch,subTree);
+			subTree[0].children = [];
+			currentTreeBranch = subTree[0].children;
+		}
 	}
-	return tree;
+	return currentTreeBranch;
 };
 
 /*