Aho-Corasick Freelinks Enhancement for Large Wikis and Non-Latin Titles

plugins/tiddlywiki/freelinks/aho-corasick.js

@@ -0,0 +1,242 @@
+/*\
+title: $:/core/modules/utils/aho-corasick.js
+type: application/javascript
+module-type: utils
+
+Optimized Aho-Corasick string matching algorithm implementation with enhanced performance
+and error handling for TiddlyWiki freelinking functionality.
+
+Useage:
+
+Initialization:
+ Create an AhoCorasick instance: var ac = new AhoCorasick();
+ After initialization, the trie and failure structures are automatically created to store patterns and failure links.
+
+Adding Patterns:
+ Call addPattern(pattern, index) to add a pattern, e.g., ac.addPattern("[[Link]]", 0);.
+ pattern is the string to match, and index is an identifier for tracking results.
+ Multiple patterns can be added, stored in the trie structure.
+
+Building Failure Links:
+ Call buildFailureLinks() to construct failure links for efficient multi-pattern matching.
+ Includes a maximum node limit (default 100,000 or 15 times the pattern count) to prevent excessive computation.
+
+Performing Search:
+ Use search(text, useWordBoundary) to find pattern matches in the text.
+ text is the input string, and useWordBoundary (boolean) controls whether to enforce word boundary checks.
+ Returns an array of match results, each containing pattern (matched pattern), index (start position), length (pattern length), and titleIndex (pattern identifier).
+
+Word Boundary Check:
+ If useWordBoundary is true, only matches surrounded by non-word characters (letters, digits, or underscores) are returned.
+
+Cleanup and Statistics:
+ Use clear() to reset the trie and failure links, freeing memory.
+ Use getStats() to retrieve statistics, including node count (nodeCount), pattern count (patternCount), and failure link count (failureLinks).
+
+Notes
+ Performance Considerations: The Aho-Corasick trie may consume significant memory with a large number of patterns. Limit the number of patterns (e.g., <10,000) for optimal performance.
+ Error Handling: The module includes maximum node and failure depth limits (maxFailureDepth) to prevent infinite loops or memory overflow.
+ Word Boundary: Enabling useWordBoundary ensures more precise matches, ideal for link detection scenarios.
+ Compatibility: Ensure compatibility with other TiddlyWiki modules (e.g., wikiparser.js) when processing WikiText.
+ Debugging: Use getStats() to inspect the trie structure's size and ensure it does not overload browser memory.
+
+\*/
+
+"use strict";
+
+function AhoCorasick() {
+	this.trie = {};
+	this.failure = {};
+	this.maxFailureDepth = 100;
+	this.patternCount = 0;
+}
+
+AhoCorasick.prototype.addPattern = function(pattern, index) {
+	if(!pattern || typeof pattern !== "string" || pattern.length === 0) {
+		return;
+	}
+
+	var node = this.trie;
+
+	for(var i = 0; i < pattern.length; i++) {
+		var char = pattern[i];
+		if(!node[char]) {
+			node[char] = {};
+		}
+		node = node[char];
+	}
+
+	if(!node.$) {
+		node.$ = [];
+	}
+	node.$.push({ 
+		pattern: pattern, 
+		index: index,
+		length: pattern.length
+	});
+
+	this.patternCount++;
+};
+
+AhoCorasick.prototype.buildFailureLinks = function() {
+	var queue = [];
+	var root = this.trie;
+	this.failure[root] = root;
+
+	for(var char in root) {
+		if(root[char] && char !== "$") {
+			this.failure[root[char]] = root;
+			queue.push(root[char]);
+		}
+	}
+
+	var processedNodes = 0;
+	var maxNodes = Math.max(100000, this.patternCount * 15);
+
+	while(queue.length > 0 && processedNodes < maxNodes) {
+		var node = queue.shift();
+		processedNodes++;
+
+		for(var char in node) {
+			if(node[char] && char !== "$") {
+				var child = node[char];
+				var fail = this.failure[node];
+				var failureDepth = 0;
+
+				while(fail && !fail[char] && failureDepth < this.maxFailureDepth) {
+					fail = this.failure[fail];
+					failureDepth++;
+				}
+
+				var failureLink = (fail && fail[char]) ? fail[char] : root;
+				this.failure[child] = failureLink;
+
+				var failureOutput = this.failure[child];
+				if(failureOutput && failureOutput.$) {
+					if(!child.$) {
+						child.$ = [];
+					}
+					child.$.push.apply(child.$, failureOutput.$);
+				}
+
+				queue.push(child);
+			}
+		}
+	}
+
+	if(processedNodes >= maxNodes) {
+		throw new Error("Aho-Corasick: buildFailureLinks exceeded maximum nodes (" + maxNodes + ")");
+	}
+};
+
+AhoCorasick.prototype.search = function(text, useWordBoundary) {
+	if(!text || typeof text !== "string" || text.length === 0) {
+		return [];
+	}
+
+	var matches = [];
+	var node = this.trie;
+	var textLength = text.length;
+	var maxMatches = Math.min(textLength * 2, 10000);
+
+	for(var i = 0; i < textLength; i++) {
+		var char = text[i];
+		var transitionCount = 0;
+
+		while(node && !node[char] && node !== this.trie && transitionCount < this.maxFailureDepth) {
+			node = this.failure[node] || this.trie;
+			transitionCount++;
+		}
+
+		if(node && node[char]) {
+			node = node[char];
+		} else {
+			node = this.trie;
+			if(this.trie[char]) {
+				node = this.trie[char];
+			}
+		}
+
+		var currentNode = node;
+		var collectCount = 0;
+		while(currentNode && collectCount < 10) {
+			if(currentNode.$) {
+				var outputs = currentNode.$;
+				for(var j = 0; j < outputs.length && matches.length < maxMatches; j++) {
+					var output = outputs[j];
+					var matchStart = i - output.length + 1;
+					var matchEnd = i + 1;
+
+					if(useWordBoundary && !this.isWordBoundaryMatch(text, matchStart, matchEnd)) {
+						continue;
+					}
+
+					matches.push({
+						pattern: output.pattern,
+						index: matchStart,
+						length: output.length,
+						titleIndex: output.index
+					});
+				}
+			}
+			currentNode = this.failure[currentNode];
+			if(currentNode === this.trie) break;
+			collectCount++;
+		}
+	}
+
+	return matches;
+};
+
+AhoCorasick.prototype.isWordBoundaryMatch = function(text, start, end) {
+	var beforeChar = start > 0 ? text[start - 1] : "";
+	var afterChar = end < text.length ? text[end] : "";
+
+	var isWordChar = function(char) {
+		return /[a-zA-Z0-9_\u00C0-\u00FF]/.test(char);
+	};
+
+	var beforeIsWord = beforeChar && isWordChar(beforeChar);
+	var afterIsWord = afterChar && isWordChar(afterChar);
+
+	return !beforeIsWord && !afterIsWord;
+};
+
+AhoCorasick.prototype.clear = function() {
+	this.trie = {};
+	this.failure = {};
+	this.patternCount = 0;
+};
+
+AhoCorasick.prototype.getStats = function() {
+	var nodeCount = 0;
+	var patternCount = 0;
+	var failureCount = 0;
+
+	function countNodes(node) {
+		if(!node) return;
+		nodeCount++;
+		if(node.$) {
+			patternCount += node.$.length;
+		}
+		for(var key in node) {
+			if(node[key] && typeof node[key] === "object" && key !== "$") {
+				countNodes(node[key]);
+			}
+		}
+	}
+
+	countNodes(this.trie);
+
+	for(var key in this.failure) {
+		failureCount++;
+	}
+
+	return {
+		nodeCount: nodeCount,
+		patternCount: this.patternCount,
+		failureLinks: failureCount
+	};
+};
+
+exports.AhoCorasick = AhoCorasick;

plugins/tiddlywiki/freelinks/config-Freelinks-WordBoundary.tid

@@ -0,0 +1,2 @@
+title: $:/config/Freelinks/WordBoundary
+text: yes

plugins/tiddlywiki/freelinks/readme.tid

@@ -2,26 +2,36 @@ title: $:/plugins/tiddlywiki/freelinks/readme
 
 This plugin adds automatic generation of links to tiddler titles.
 
-''Note that automatic link generation can be very slow when there are a large number of tiddlers''.
+The plugin uses the Aho-Corasick algorithm for efficient pattern matching with large numbers of tiddlers.
+
+!! Configuration
 
 Freelinking is activated for runs of text that have the following variables set:
 
 * `tv-wikilinks` is NOT equal to `no`
 * `tv-freelinks` is set to `yes`
 
-Freelinks are case sensitive by default but can be configured to ignore case in the settings tab.
+Freelinks are case sensitive by default but can be configured to ignore case in the settings panel.
+
+Word boundary checking can be configured in the settings panel. When enabled (default for Western languages), links are created only for complete words. When disabled, partial matches within words are also linked.
+
+When multiple tiddler titles match the same text, longer titles take precedence over shorter ones.
+
+Tiddlers do not create links to themselves.
+
+Use `$:/config/Freelinks/TargetFilter` to define which tiddlers are eligible for auto-linking.
 
-Within view templates, the variable `tv-freelinks` is automatically set to the content of $:/config/Freelinks/Enable, which can be set via the settings panel of this plugin.
+Within view templates, the variable `tv-freelinks` is automatically set to the content of `$:/config/Freelinks/Enable`, which can be set via the settings panel of this plugin.
 
 !! Notes
 
-To change within which tiddlers freelinking occurs requires customising the shadow tiddler [[$:/plugins/tiddlywiki/freelinks/macros/view]]. This tiddler is tagged $:/tags/Macro/View  which means that it will be included as a local macro in each view template. By default, its content is:
+To change within which tiddlers freelinking occurs requires customising the shadow tiddler [[$:/plugins/tiddlywiki/freelinks/macros/view]]. This tiddler is tagged `$:/tags/Macro/View` which means that it will be included as a local macro in each view template. By default, its content is:
 
 ```
 <$set name="tv-freelinks" value={{$:/config/Freelinks/Enable}}/>
 ```
 
-That means that for each tiddler the variable tv-freelinks will be set to the tiddler $:/config/Freelinks/Enable, which is set to "yes" or "no" by the settings in control panel.
+That means that for each tiddler the variable `tv-freelinks` will be set to the tiddler `$:/config/Freelinks/Enable`, which is set to "yes" or "no" by the settings in control panel.
 
 Instead, we can use a filter expression to, say, only freelink within the tiddler with the title "HelloThere":
 
@@ -40,3 +50,14 @@ Or we can combine both approaches:
 ```
 <$set name="tv-freelinks" value={{{ [<currentTiddler>match[HelloThere]] ~[<currentTiddler>tag[MyTag]] +[then[yes]else[no]] }}}/>
 ```
+
+!! Implementation Details
+
+The Aho-Corasick algorithm implementation includes:
+
+* Unicode character support for international text
+* Prevention of self-referential links within the current tiddler
+* Performance safeguards including depth protection and result limiting
+* Graceful fallback handling for invalid patterns
+
+Longer tiddler titles take precedence over shorter ones when multiple matches are possible.

plugins/tiddlywiki/freelinks/settings.tid

@@ -6,4 +6,6 @@ Filter defining tiddlers to which freelinks are made: <$edit-text tiddler="$:/co
 
 <$checkbox tiddler="$:/config/Freelinks/Enable" field="text" checked="yes" unchecked="no" default="no"> <$link to="$:/config/Freelinks/Enable">Enable freelinking within tiddler view templates</$link> </$checkbox>
 
+<$checkbox tiddler="$:/config/Freelinks/WordBoundary" field="text" checked="yes" unchecked="no" default="yes"> <$link to="$:/config/Freelinks/WordBoundary">Word Boundary Check</$link> </$checkbox>
+
 <$checkbox tiddler="$:/config/Freelinks/IgnoreCase" field="text" checked="yes" unchecked="no" default="no"> <$link to="$:/config/Freelinks/IgnoreCase">Ignore case</$link> </$checkbox>