MediaWiki:Gadget-AutoInterlink.js - Revision history

Bharatwiki Socrates at 10:55, 18 October 2026

2026-10-18T10:55:26Z

@@ Line 214: / Line 214: @@
 		$row.append( $titleInput, $scanBtn );
 		$root.append( $row );
 		var $msg = $( '<div>' ).css( { 'font-size': '13px', color: '#d33', 'margin-bottom': '8px' } ).hide();
@@ Line 519: / Line 552: @@
 		if ( attempts < 100 ) {
 			setTimeout( function () { waitForDeps( callback, attempts + 1 ); }, 100 );
-			} else {
+		} else {
 			try { callback(); } catch ( e ) {}
+		}
+	}
 	waitForDeps( function () {

Bharatwiki Socrates at 10:48, 18 October 2026

2026-10-18T10:48:19Z

← Older revision		Revision as of 16:18, 18 October 2026
Line 46:		Line 46:
	while ( i < n && guard < 100000 ) {		while ( i < n && guard < 100000 ) {
	guard++;		guard++;
			// ==heading== (exclude entire line if it's a heading)
			if ( text.substr( i, 2 ) === '==' ) {
			var nl = text.indexOf( '\n', i );
			unsafe.push( [ i, nl === -1 ? n : nl ] );
			i = nl === -1 ? n : nl;
			continue;
			}
	// <!-- comment -->		// <!-- comment -->
	if ( text.substr( i, 4 ) === '<!--' ) {		if ( text.substr( i, 4 ) === '<!--' ) {
Line 144:		Line 151:
	var pluralS = '';		var pluralS = '';
	var term = phrase;		var term = phrase;
			var display = phrase;
	if ( /s$/i.test( term ) && term.length > 3 ) {		if ( /s$/i.test( term ) && term.length > 3 ) {
	var sing = term.replace( /s$/i, '' );		var sing = term.replace( /s$/i, '' );
	if ( sing.length > 2 ) { pluralS = ~~term.match( /~~s~~$/i )[0]~~; term = sing; key = term.toLowerCase(); }		if ( sing.length > 2 ) {
			pluralS = 's';
			term = sing;
			display = sing; // Display the singular, s goes outside
			key = term.toLowerCase();
			}
	if ( seen[ key ] ) continue;		if ( seen[ key ] ) continue;
	seen[ key ] = true;		seen[ key ] = true;
	}		}
	cands.push( { term: term, pluralS: pluralS, display: ~~phrase~~, start: absStart, end: absEnd, words: words.length } );		cands.push( { term: term, pluralS: pluralS, display: display, start: absStart, end: absEnd, words: words.length } );
	}		}
	}		}
Line 506:		Line 519:
	if ( attempts < 100 ) {		if ( attempts < 100 ) {
	setTimeout( function () { waitForDeps( callback, attempts + 1 ); }, 100 );		setTimeout( function () { waitForDeps( callback, attempts + 1 ); }, 100 );
	} else {		} else {
	try { callback(); } catch ( e ) {}		try { callback(); } catch ( e ) {}
	}		}
	}		}
	waitForDeps( function () {		waitForDeps( function () {

Bharatwiki Socrates at 10:05, 18 October 2026

2026-10-18T10:05:00Z

@@ Line 39: / Line 39: @@
 	// Regions of wikitext that must never get links inserted inside them.
 	function getSafeRanges( text ) {
-		// Simplified: return whole text as safe (skip complex parsing to avoid hangs)
+		// Returns [start,end) ranges of text SAFE for linking (excludes [[links]], <ref>, {{templates}})
 		if ( !text ) return [];
-		return [ [ 0, text.length ] ];
+		var n = text.length, i = 0;
+	}
@@ Line 467: / Line 520: @@
 	} );
 }() );

Bharatwiki Socrates at 08:52, 18 October 2026

2026-10-18T08:52:57Z

← Older revision		Revision as of 14:22, 18 October 2026
Line 66:		Line 66:

	function extractCandidates( text, safeRanges ) {		function extractCandidates( text, safeRanges ) {
	// Simplified robust version - extracts capitalized phrases ~~only~~		// Simplified robust version - extracts capitalized phrases with REAL positions
	var cands = [], seen = {};		var cands = [], seen = {};
	var i, r, rs, seg;		var r, rs, seg, base;
	/~~/ Simple approach~~: ~~find all capitalized words~~/~~phrases via regex on safe text~~		var re = /\b([A-Z][a-z]+(?:\s+[A-Z][a-z]+){0,3})\b/g;
	var ~~safeText~~ = "";		var m, count = 0;
	for ( r = 0; r < safeRanges.length; r++ ) {		for ( r = 0; r < safeRanges.length && count < 500; r++ ) {
	rs = safeRanges[ r ];		rs = safeRanges[ r ];
	~~safeText +~~= text.slice( rs[ 0 ], rs[ 1 ] ) ~~+ "\n"~~;		seg = text.slice( rs[ 0 ], rs[ 1 ] );
	}		base = rs[ 0 ];
	~~// Match 1-4 capitalized words in a row~~		re.lastIndex = 0;
	~~var re~~ = ~~/\b(~~[~~A-Z~~]~~[a-z]+(?:\s+[A-Z][a-z]+){0,3})\b/g~~;		while ( ( m = re.exec( seg ) ) !== null && count < 500 ) {
	~~var m;~~		count++;
	~~var count~~ = 0;		var phrase = m[1];
	while ( ( m = re.exec( ~~safeText~~ ) ) !== null && count < 500 ) {		var absStart = base + m.index;
	count++;		var absEnd = absStart + phrase.length;
	var phrase = m[1];		var words = phrase.split(/\s+/);
	~~// Skip if all stopwords (simple check)~~		if ( words.length === 1 ) {
	var words = phrase.split(/\s+/);		var lw = words[0].toLowerCase();
	if ( words.length === 1 ) {		if ( STOP.indexOf( lw ) !== -1 ) continue;
	var lw = words[0].toLowerCase();		}
	if ( STOP.indexOf( lw ) !== -1 ) continue;		var key = phrase.toLowerCase();
	}
	var key = phrase.toLowerCase();
	~~if ( seen[ key ] ) continue;~~
	~~seen[ key ] = true;~~
	~~// Handle plural~~
	~~var pluralS = '';~~
	~~var term = phrase;~~
	~~if ( /s$/i.test( term ) && term.length > 3 ) {~~
	~~var sing = term.replace( /s$/i, '' );~~
	~~if ( sing.length > 2 ) { pluralS = term.match( /s$/i )[0]; term = sing; key = term.toLowerCase(); }~~
	if ( seen[ key ] ) continue;		if ( seen[ key ] ) continue;
	seen[ key ] = true;		seen[ key ] = true;
			var pluralS = '';
			var term = phrase;
			if ( /s$/i.test( term ) && term.length > 3 ) {
			var sing = term.replace( /s$/i, '' );
			if ( sing.length > 2 ) { pluralS = term.match( /s$/i )[0]; term = sing; key = term.toLowerCase(); }
			if ( seen[ key ] ) continue;
			seen[ key ] = true;
			}
			cands.push( { term: term, pluralS: pluralS, display: phrase, start: absStart, end: absEnd, words: words.length } );
	}		}
	~~cands.push( { term: term, pluralS: pluralS, display: phrase, start: 0, end: 0, words: words.length } );~~
	}		}
	return cands;		return cands;

Bharatwiki Socrates at 08:03, 18 October 2026

2026-10-18T08:03:23Z

@@ Line 39: / Line 39: @@
 	// Regions of wikitext that must never get links inserted inside them.
 	function getSafeRanges( text ) {
-		var skip = [], n = text.length, i = 0, j, e, m, tag, close, gt;
+		// Simplified: return whole text as safe (skip complex parsing to avoid hangs)
-		while ( i < n ) {
+		// TODO: Add proper safe-zone filtering later
-			if ( text.indexOf( '<!--', i ) === i ) {
+		if ( !text ) return [];
-				j = text.indexOf( '-->', i + 4 );
+		return [ [ 0, text.length ] ];
+	}
 	function collectExistingLinks( text ) {
-		var set = {}, re = /\[\[([^\]|]+)(?:\|([^\]]*))?\]\]/g, m;
+		// Simplified safe version
-		while ( ( m = re.exec( text ) ) !== null ) {
+		var set = {};
-			set[ m[ 1 ].trim().toLowerCase().replace( /_/g, ' ' ) ] = true;
+		if ( !text ) return set;
-			if ( m[ 2 ] !== undefined ) set[ m[ 2 ].trim().toLowerCase().replace( /_/g, ' ' ) ] = true;
+		try {
+			var re = /\[\[([^\]|]+)(?:\|([^\]]*))?\]\]/g, m;
 		return set;
+	}
@@ Line 153: / Line 106: @@
 	function greedyFilter( cands ) {
-		var sorted = cands.slice().sort( function ( a, b ) { return ( b.words - a.words ) || ( a.start - b.start ); } );
+		// Simplified: just return first 100 to avoid O(n^2) hang
-		var taken = [], out = [], c, t, overlap, k;
+		// TODO: Implement proper overlap filtering later
-		for ( var x = 0; x < sorted.length; x++ ) {
+		return cands.slice( 0, 100 );
+	}

Bharatwiki Socrates at 07:48, 18 October 2026

2026-10-18T07:48:56Z

← Older revision		Revision as of 13:18, 18 October 2026
Line 406:		Line 406:
	$discardBtn.on( 'click', function () {		$discardBtn.on( 'click', function () {
	state.items = [];		state.items = [];
	$results.hide().empty();		$results.hide().empty();
	showStatus( 'Discarded. Enter an article title to scan again.' );		showStatus( 'Discarded. Enter an article title to scan again.' );
	} );		} );

Bharatwiki Socrates at 07:28, 18 October 2026

2026-10-18T07:28:54Z

Show changes

Bharatwiki Socrates at 07:27, 18 October 2026

2026-10-18T07:27:00Z

← Older revision		Revision as of 12:57, 18 October 2026
Line 13:		Line 13:
	'use strict';		'use strict';

	var TOOL_PAGE = 'Bharatpedia:~~Auto_interlinker~~';		var TOOL_PAGE = 'Bharatpedia:Auto interlinker';
	var MAX_CHECK = ~~100~~;		var MAX_CHECK = 200;
	var MAX_LINKS = 50;		var MAX_LINKS = 50;

Line 114:		Line 114:
	function extractCandidates( text, safeRanges ) {		function extractCandidates( text, safeRanges ) {
	var cands = [], seen = {};		var cands = [], seen = {};
	var wordRe = /[~~A-Za-z\u00C0-~~\~~u024F~~][~~A-Za-z0-9~~\~~u00C0-~~\~~u024F~~'-]*/g;		var wordRe = /[\p{L}][\p{L}\p{N}'-]*/gu;
	var r, rs, seg, lines, li, line, abs, words, m, i, j, runLen, len, s, win, lows, parts, lookup, key;		var r, rs, seg, lines, li, line, abs, words, m, i, j, runLen, len, s, win, lows, parts, lookup, key;
	for ( r = 0; r < safeRanges.length; r++ ) {		for ( r = 0; r < safeRanges.length; r++ ) {
Line 131:		Line 131:
	i = 0;		i = 0;
	while ( i < words.length ) {		while ( i < words.length ) {
	if ( /^[A-~~Z\u00C0-\u00DE~~]/.test( words[ i ].w ) ) {		if ( /^\p{Lu}/u.test( words[ i ].w ) \|\| /^[a-z]/.test( words[ i ].w ) ) {
	j = i;		j = i;
	while ( j < words.length && /^[~~A-Z\u00C0~~-~~\u00DE~~]/.test( words[ j ].w ) ) {		while ( j < words.length && ( /^\p{Lu}/u.test( words[ j ].w ) \|\| /^[a-z]/.test( words[ j ].w ) ) ) {
	// break the run at sentence boundaries so phrases do		// break the run at sentence boundaries so phrases do
	// not bleed across full stops, e.g. "Movement. Gandhi"		// not bleed across full stops, e.g. "Movement. Gandhi"
Line 262:		Line 262:
	var map = {};		var map = {};
	var batches = [];		var batches = [];
	for ( var i = 0; i < lookups.length; i += ~~100~~ ) {		for ( var i = 0; i < lookups.length; i += 50 ) {
	batches.push( lookups.slice( i, i + ~~100~~ ) );		batches.push( lookups.slice( i, i + 50 ) );
	}		}
	var chain = $.Deferred().resolve().promise();		var chain = $.Deferred().resolve().promise();
	batches.forEach( function ( batch, bi ) {		batches.forEach( function ( batch, bi ) {
	chain = chain.then( function () {		chain = chain.then( function () {
	if ( onProgress ) onProgress( Math.min( ( bi + 1 ) * ~~100~~, lookups.length ), lookups.length );		if ( onProgress ) onProgress( Math.min( ( bi + 1 ) * 50, lookups.length ), lookups.length );
	return api.get( {		return api.get( {
	action: 'query', titles: batch.join( '\|' ), prop: 'info', redirects: 1		action: 'query', titles: batch.join( '\|' ), prop: 'info', redirects: 1

Bharatwiki Socrates at 07:14, 18 October 2026

2026-10-18T07:14:29Z

← Older revision		Revision as of 12:44, 18 October 2026
Line 114:		Line 114:
	function extractCandidates( text, safeRanges ) {		function extractCandidates( text, safeRanges ) {
	var cands = [], seen = {};		var cands = [], seen = {};
	var wordRe = /[\~~p{L}~~][\~~p{L}~~\~~p{N}~~'-]*/gu;		var wordRe = /[A-Za-z\u00C0-\u024F][A-Za-z0-9\u00C0-\u024F'-]*/g;
	var r, rs, seg, lines, li, line, abs, words, m, i, j, runLen, len, s, win, lows, parts, lookup, key;		var r, rs, seg, lines, li, line, abs, words, m, i, j, runLen, len, s, win, lows, parts, lookup, key;
	for ( r = 0; r < safeRanges.length; r++ ) {		for ( r = 0; r < safeRanges.length; r++ ) {
Line 131:		Line 131:
	i = 0;		i = 0;
	while ( i < words.length ) {		while ( i < words.length ) {
	if ( /^\~~p{Lu}~~/u.test( words[ i ].w ) ) {		if ( /^[A-Z\u00C0-\u00DE]/.test( words[ i ].w ) ) {
	j = i;		j = i;
	while ( j < words.length && /^\~~p{Lu}~~/u.test( words[ j ].w ) ) {		while ( j < words.length && /^[A-Z\u00C0-\u00DE]/.test( words[ j ].w ) ) {
	// break the run at sentence boundaries so phrases do		// break the run at sentence boundaries so phrases do
	// not bleed across full stops, e.g. "Movement. Gandhi"		// not bleed across full stops, e.g. "Movement. Gandhi"

Bharatwiki Socrates at 04:50, 18 October 2026

2026-10-18T04:50:24Z

← Older revision		Revision as of 10:20, 18 October 2026
Line 131:		Line 131:
	i = 0;		i = 0;
	while ( i < words.length ) {		while ( i < words.length ) {
	if ( /^\p{Lu}/u~~.test( words[ i ].w ) \|\| /^[a-z]/~~.test( words[ i ].w ) ) {		if ( /^\p{Lu}/u.test( words[ i ].w ) ) {
	j = i;		j = i;
	while ( j < words.length && ( /^\p{Lu}/u.test( words[ j ].w ~~) \|\| /^[a-z]/.test( words[ j ].w )~~ ) ) {		while ( j < words.length && /^\p{Lu}/u.test( words[ j ].w ) ) {
	// break the run at sentence boundaries so phrases do		// break the run at sentence boundaries so phrases do
	// not bleed across full stops, e.g. "Movement. Gandhi"		// not bleed across full stops, e.g. "Movement. Gandhi"