--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/includes/clientside/tinymce/plugins/devkit/jscripts/diff.js Wed Jun 13 16:07:17 2007 -0400
@@ -0,0 +1,1192 @@
+// Diff_Match_Patch v1.3
+// Computes the difference between two texts to create a patch.
+// Applies the patch onto another text, allowing for errors.
+// Copyright (C) 2006 Neil Fraser
+// http://neil.fraser.name/software/diff_match_patch/
+
+// This program is free software; you can redistribute it and/or
+// modify it under the terms of the GNU General Public License
+// as published by the Free Software Foundation.
+
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+// GNU General Public License (www.gnu.org) for more details.
+
+
+// Constants.
+// Redefine these in your program to override the defaults.
+
+// Number of seconds to map a diff before giving up. (0 for infinity)
+var DIFF_TIMEOUT = 1.0;
+// Cost of an empty edit operation in terms of edit characters.
+var DIFF_EDIT_COST = 4;
+// Tweak the relative importance (0.0 = accuracy, 1.0 = proximity)
+var MATCH_BALANCE = 0.5;
+// At what point is no match declared (0.0 = perfection, 1.0 = very loose)
+var MATCH_THRESHOLD = 0.5;
+// The min and max cutoffs used when computing text lengths.
+var MATCH_MINLENGTH = 100;
+var MATCH_MAXLENGTH = 1000;
+// Chunk size for context length.
+var PATCH_MARGIN = 4;
+
+
+ //////////////////////////////////////////////////////////////////////
+ // Diff //
+//////////////////////////////////////////////////////////////////////
+
+// The data structure representing a diff is an array of tuples:
+// [[-1, "Hello"], [1, "Goodbye"], [0, " world."]]
+// which means: delete "Hello", add "Goodbye" and keep " world."
+
+
+function diff_main(text1, text2, checklines) {
+ // Find the differences between two texts. Return an array of changes.
+ // If checklines is present and false, then don't run a line-level diff first to identify the changed areas.
+ // Check for equality (speedup)
+ if (text1 == text2)
+ return [[0, text1]];
+
+ if (typeof checklines == 'undefined')
+ checklines = true;
+
+ var a;
+ // Trim off common prefix (speedup)
+ a = diff_prefix(text1, text2);
+ text1 = a[0];
+ text2 = a[1];
+ var commonprefix = a[2];
+
+ // Trim off common suffix (speedup)
+ a = diff_suffix(text1, text2);
+ text1 = a[0];
+ text2 = a[1];
+ var commonsuffix = a[2];
+
+ var diff, i;
+ var longtext = text1.length > text2.length ? text1 : text2;
+ var shorttext = text1.length > text2.length ? text2 : text1;
+
+ if (!text1) { // Just add some text (speedup)
+ diff = [[1, text2]];
+ } else if (!text2) { // Just delete some text (speedup)
+ diff = [[-1, text1]];
+ } else if ((i = longtext.indexOf(shorttext)) != -1) {
+ // Shorter text is inside the longer text (speedup)
+ diff = [[1, longtext.substring(0, i)], [0, shorttext], [1, longtext.substring(i+shorttext.length)]];
+ // Swap insertions for deletions if diff is reversed.
+ if (text1.length > text2.length)
+ diff[0][0] = diff[2][0] = -1;
+ } else {
+ longtext = shorttext = null; // Garbage collect
+ // Check to see if the problem can be split in two.
+ var hm = diff_halfmatch(text1, text2);
+ if (hm) {
+ // A half-match was found, sort out the return data.
+ var text1_a = hm[0];
+ var text1_b = hm[1];
+ var text2_a = hm[2];
+ var text2_b = hm[3];
+ var mid_common = hm[4];
+ // Send both pairs off for separate processing.
+ var diff_a = diff_main(text1_a, text2_a, checklines);
+ var diff_b = diff_main(text1_b, text2_b, checklines);
+ // Merge the results.
+ diff = diff_a.concat([[0, mid_common]], diff_b);
+ } else {
+ // Perform a real diff.
+ if (checklines && text1.length + text2.length < 250)
+ checklines = false; // Too trivial for the overhead.
+ if (checklines) {
+ // Scan the text on a line-by-line basis first.
+ a = diff_lines2chars(text1, text2);
+ text1 = a[0];
+ text2 = a[1];
+ var linearray = a[2];
+ }
+ diff = diff_map(text1, text2);
+ if (!diff) // No acceptable result.
+ diff = [[-1, text1], [1, text2]];
+ if (checklines) {
+ diff_chars2lines(diff, linearray); // Convert the diff back to original text.
+ diff_cleanup_semantic(diff); // Eliminate freak matches (e.g. blank lines)
+
+ // Rediff any replacement blocks, this time on character-by-character basis.
+ diff.push([0, '']); // Add a dummy entry at the end.
+ var pointer = 0;
+ var count_delete = 0;
+ var count_insert = 0;
+ var text_delete = '';
+ var text_insert = '';
+ while(pointer < diff.length) {
+ if (diff[pointer][0] == 1) {
+ count_insert++;
+ text_insert += diff[pointer][1];
+ } else if (diff[pointer][0] == -1) {
+ count_delete++;
+ text_delete += diff[pointer][1];
+ } else { // Upon reaching an equality, check for prior redundancies.
+ if (count_delete >= 1 && count_insert >= 1) {
+ // Delete the offending records and add the merged ones.
+ a = diff_main(text_delete, text_insert, false);
+ diff.splice(pointer - count_delete - count_insert, count_delete + count_insert);
+ pointer = pointer - count_delete - count_insert;
+ for (i=a.length-1; i>=0; i--)
+ diff.splice(pointer, 0, a[i]);
+ pointer = pointer + a.length;
+ }
+ count_insert = 0;
+ count_delete = 0;
+ text_delete = '';
+ text_insert = '';
+ }
+ pointer++;
+ }
+ diff.pop(); // Remove the dummy entry at the end.
+
+ }
+ }
+ }
+
+ if (commonprefix)
+ diff.unshift([0, commonprefix]);
+ if (commonsuffix)
+ diff.push([0, commonsuffix]);
+ diff_cleanup_merge(diff);
+ return diff;
+}
+
+
+function diff_lines2chars(text1, text2) {
+ // Split text into an array of strings.
+ // Reduce the texts to a string of hashes where each character represents one line.
+ var linearray = new Array(); // linearray[4] == "Hello\n"
+ var linehash = new Object(); // linehash["Hello\n"] == 4
+
+ // "\x00" is a valid JavaScript character, but the Venkman debugger doesn't like it (bug 335098)
+ // So we'll insert a junk entry to avoid generating a null character.
+ linearray.push('');
+
+ function diff_lines2chars_munge(text) {
+ // My first ever closure!
+ var i, line;
+ var chars = '';
+ while (text) {
+ i = text.indexOf('\n');
+ if (i == -1)
+ i = text.length;
+ line = text.substring(0, i+1);
+ text = text.substring(i+1);
+ if (linehash.hasOwnProperty ? linehash.hasOwnProperty(line) : (linehash[line] !== undefined)) {
+ chars += String.fromCharCode(linehash[line]);
+ } else {
+ linearray.push(line);
+ linehash[line] = linearray.length - 1;
+ chars += String.fromCharCode(linearray.length - 1);
+ }
+ }
+ return chars;
+ }
+
+ var chars1 = diff_lines2chars_munge(text1);
+ var chars2 = diff_lines2chars_munge(text2);
+ return [chars1, chars2, linearray];
+}
+
+
+function diff_chars2lines(diff, linearray) {
+ // Rehydrate the text in a diff from a string of line hashes to real lines of text.
+ var chars, text;
+ for (var x=0; x<diff.length; x++) {
+ chars = diff[x][1];
+ text = '';
+ for (var y=0; y<chars.length; y++)
+ text += linearray[chars.charCodeAt(y)];
+ diff[x][1] = text;
+ }
+}
+
+
+function diff_map(text1, text2) {
+ // Explore the intersection points between the two texts.
+ var now = new Date();
+ var ms_end = now.getTime() + DIFF_TIMEOUT * 1000; // Don't run for too long.
+ var max = (text1.length + text2.length) / 2;
+ var v_map1 = new Array();
+ var v_map2 = new Array();
+ var v1 = new Object();
+ var v2 = new Object();
+ v1[1] = 0;
+ v2[1] = 0;
+ var x, y;
+ var footstep; // Used to track overlapping paths.
+ var footsteps = new Object();
+ var done = false;
+ var hasOwnProperty = !!(footsteps.hasOwnProperty);
+ // If the total number of characters is odd, then the front path will collide with the reverse path.
+ var front = (text1.length + text2.length) % 2;
+ for (var d=0; d<max; d++) {
+ now = new Date();
+ if (DIFF_TIMEOUT > 0 && now.getTime() > ms_end) // Timeout reached
+ return null;
+
+ // Walk the front path one step.
+ v_map1[d] = new Object();
+ for (var k=-d; k<=d; k+=2) {
+ if (k == -d || k != d && v1[k-1] < v1[k+1])
+ x = v1[k+1];
+ else
+ x = v1[k-1]+1;
+ y = x - k;
+ footstep = x+","+y;
+ if (front && (hasOwnProperty ? footsteps.hasOwnProperty(footstep) : (footsteps[footstep] !== undefined)))
+ done = true;
+ if (!front)
+ footsteps[footstep] = d;
+ while (!done && x < text1.length && y < text2.length && text1.charAt(x) == text2.charAt(y)) {
+ x++; y++;
+ footstep = x+","+y;
+ if (front && (hasOwnProperty ? footsteps.hasOwnProperty(footstep) : (footsteps[footstep] !== undefined)))
+ done = true;
+ if (!front)
+ footsteps[footstep] = d;
+ }
+ v1[k] = x;
+ v_map1[d][x+","+y] = true;
+ if (done) {
+ // Front path ran over reverse path.
+ v_map2 = v_map2.slice(0, footsteps[footstep]+1);
+ var a = diff_path1(v_map1, text1.substring(0, x), text2.substring(0, y));
+ return a.concat(diff_path2(v_map2, text1.substring(x), text2.substring(y)));
+ }
+ }
+
+ // Walk the reverse path one step.
+ v_map2[d] = new Object();
+ for (var k=-d; k<=d; k+=2) {
+ if (k == -d || k != d && v2[k-1] < v2[k+1])
+ x = v2[k+1];
+ else
+ x = v2[k-1]+1;
+ y = x - k;
+ footstep = (text1.length-x)+","+(text2.length-y);
+ if (!front && (hasOwnProperty ? footsteps.hasOwnProperty(footstep) : (footsteps[footstep] !== undefined)))
+ done = true;
+ if (front)
+ footsteps[footstep] = d;
+ while (!done && x < text1.length && y < text2.length && text1.charAt(text1.length-x-1) == text2.charAt(text2.length-y-1)) {
+ x++; y++;
+ footstep = (text1.length-x)+","+(text2.length-y);
+ if (!front && (hasOwnProperty ? footsteps.hasOwnProperty(footstep) : (footsteps[footstep] !== undefined)))
+ done = true;
+ if (front)
+ footsteps[footstep] = d;
+ }
+ v2[k] = x;
+ v_map2[d][x+","+y] = true;
+ if (done) {
+ // Reverse path ran over front path.
+ v_map1 = v_map1.slice(0, footsteps[footstep]+1);
+ var a = diff_path1(v_map1, text1.substring(0, text1.length-x), text2.substring(0, text2.length-y));
+ return a.concat(diff_path2(v_map2, text1.substring(text1.length-x), text2.substring(text2.length-y)));
+ }
+ }
+ }
+ // Number of diffs equals number of characters, no commonality at all.
+ return null;
+}
+
+
+function diff_path1(v_map, text1, text2) {
+ // Work from the middle back to the start to determine the path.
+ var path = [];
+ var x = text1.length;
+ var y = text2.length;
+ var last_op = null;
+ for (var d=v_map.length-2; d>=0; d--) {
+ while(1) {
+ if (v_map[d].hasOwnProperty ? v_map[d].hasOwnProperty((x-1)+","+y) : (v_map[d][(x-1)+","+y] !== undefined)) {
+ x--;
+ if (last_op === -1)
+ path[0][1] = text1.charAt(x) + path[0][1];
+ else
+ path.unshift([-1, text1.charAt(x)]);
+ last_op = -1;
+ break;
+ } else if (v_map[d].hasOwnProperty ? v_map[d].hasOwnProperty(x+","+(y-1)) : (v_map[d][x+","+(y-1)] !== undefined)) {
+ y--;
+ if (last_op === 1)
+ path[0][1] = text2.charAt(y) + path[0][1];
+ else
+ path.unshift([1, text2.charAt(y)]);
+ last_op = 1;
+ break;
+ } else {
+ x--;
+ y--;
+ //if (text1.charAt(x) != text2.charAt(y))
+ // return alert("No diagonal. Can't happen. (diff_path1)");
+ if (last_op === 0)
+ path[0][1] = text1.charAt(x) + path[0][1];
+ else
+ path.unshift([0, text1.charAt(x)]);
+ last_op = 0;
+ }
+ }
+ }
+ return path;
+}
+
+
+function diff_path2(v_map, text1, text2) {
+ // Work from the middle back to the end to determine the path.
+ var path = [];
+ var x = text1.length;
+ var y = text2.length;
+ var last_op = null;
+ for (var d=v_map.length-2; d>=0; d--) {
+ while(1) {
+ if (v_map[d].hasOwnProperty ? v_map[d].hasOwnProperty((x-1)+","+y) : (v_map[d][(x-1)+","+y] !== undefined)) {
+ x--;
+ if (last_op === -1)
+ path[path.length-1][1] += text1.charAt(text1.length-x-1);
+ else
+ path.push([-1, text1.charAt(text1.length-x-1)]);
+ last_op = -1;
+ break;
+ } else if (v_map[d].hasOwnProperty ? v_map[d].hasOwnProperty(x+","+(y-1)) : (v_map[d][x+","+(y-1)] !== undefined)) {
+ y--;
+ if (last_op === 1)
+ path[path.length-1][1] += text2.charAt(text2.length-y-1);
+ else
+ path.push([1, text2.charAt(text2.length-y-1)]);
+ last_op = 1;
+ break;
+ } else {
+ x--;
+ y--;
+ //if (text1.charAt(text1.length-x-1) != text2.charAt(text2.length-y-1))
+ // return alert("No diagonal. Can't happen. (diff_path2)");
+ if (last_op === 0)
+ path[path.length-1][1] += text1.charAt(text1.length-x-1);
+ else
+ path.push([0, text1.charAt(text1.length-x-1)]);
+ last_op = 0;
+ }
+ }
+ }
+ return path;
+}
+
+
+function diff_prefix(text1, text2) {
+ // Trim off common prefix
+ var pointermin = 0;
+ var pointermax = Math.min(text1.length, text2.length);
+ var pointermid = pointermax;
+ while(pointermin < pointermid) {
+ if (text1.substring(0, pointermid) == text2.substring(0, pointermid))
+ pointermin = pointermid;
+ else
+ pointermax = pointermid;
+ pointermid = Math.floor((pointermax - pointermin) / 2 + pointermin);
+ }
+ var commonprefix = text1.substring(0, pointermid);
+ text1 = text1.substring(pointermid);
+ text2 = text2.substring(pointermid);
+ return [text1, text2, commonprefix];
+}
+
+
+function diff_suffix(text1, text2) {
+ // Trim off common suffix
+ var pointermin = 0;
+ var pointermax = Math.min(text1.length, text2.length);
+ var pointermid = pointermax;
+ while(pointermin < pointermid) {
+ if (text1.substring(text1.length-pointermid) == text2.substring(text2.length-pointermid))
+ pointermin = pointermid;
+ else
+ pointermax = pointermid;
+ pointermid = Math.floor((pointermax - pointermin) / 2 + pointermin);
+ }
+ var commonsuffix = text1.substring(text1.length-pointermid);
+ text1 = text1.substring(0, text1.length-pointermid);
+ text2 = text2.substring(0, text2.length-pointermid);
+ return [text1, text2, commonsuffix];
+}
+
+
+function diff_halfmatch(text1, text2) {
+ // Do the two texts share a substring which is at least half the length of the longer text?
+ var longtext = text1.length > text2.length ? text1 : text2;
+ var shorttext = text1.length > text2.length ? text2 : text1;
+ if (longtext.length < 10 || shorttext.length < 1)
+ return null; // Pointless.
+
+ function diff_halfmatch_i(longtext, shorttext, i) {
+ // Start with a 1/4 length substring at position i as a seed.
+ var seed = longtext.substring(i, i+Math.floor(longtext.length/4));
+ var j = -1;
+ var best_common = '';
+ var best_longtext_a, best_longtext_b, best_shorttext_a, best_shorttext_b;
+ while ((j = shorttext.indexOf(seed, j+1)) != -1) {
+ var my_prefix = diff_prefix(longtext.substring(i), shorttext.substring(j));
+ var my_suffix = diff_suffix(longtext.substring(0, i), shorttext.substring(0, j));
+ if (best_common.length < (my_suffix[2] + my_prefix[2]).length) {
+ best_common = my_suffix[2] + my_prefix[2];
+ best_longtext_a = my_suffix[0];
+ best_longtext_b = my_prefix[0];
+ best_shorttext_a = my_suffix[1];
+ best_shorttext_b = my_prefix[1];
+ }
+ }
+ if (best_common.length >= longtext.length/2)
+ return [best_longtext_a, best_longtext_b, best_shorttext_a, best_shorttext_b, best_common];
+ else
+ return null;
+ }
+
+ // First check if the second quarter is the seed for a half-match.
+ var hm1 = diff_halfmatch_i(longtext, shorttext, Math.ceil(longtext.length/4));
+ // Check again based on the third quarter.
+ var hm2 = diff_halfmatch_i(longtext, shorttext, Math.ceil(longtext.length/2));
+ var hm;
+ if (!hm1 && !hm2)
+ return null;
+ else if (!hm2)
+ hm = hm1;
+ else if (!hm1)
+ hm = hm2;
+ else // Both matched. Select the longest.
+ hm = hm1[4].length > hm2[4].length ? hm1 : hm2;
+
+ // A half-match was found, sort out the return data.
+ if (text1.length > text2.length) {
+ var text1_a = hm[0];
+ var text1_b = hm[1];
+ var text2_a = hm[2];
+ var text2_b = hm[3];
+ } else {
+ var text2_a = hm[0];
+ var text2_b = hm[1];
+ var text1_a = hm[2];
+ var text1_b = hm[3];
+ }
+ var mid_common = hm[4];
+ return [text1_a, text1_b, text2_a, text2_b, mid_common];
+}
+
+
+function diff_cleanup_semantic(diff) {
+ // Reduce the number of edits by eliminating semantically trivial equalities.
+ var changes = false;
+ var equalities = []; // Stack of indices where equalities are found.
+ var lastequality = null; // Always equal to equalities[equalities.length-1][1]
+ var pointer = 0; // Index of current position.
+ var length_changes1 = 0; // Number of characters that changed prior to the equality.
+ var length_changes2 = 0; // Number of characters that changed after the equality.
+ while (pointer < diff.length) {
+ if (diff[pointer][0] == 0) { // equality found
+ equalities.push(pointer);
+ length_changes1 = length_changes2;
+ length_changes2 = 0;
+ lastequality = diff[pointer][1];
+ } else { // an insertion or deletion
+ length_changes2 += diff[pointer][1].length;
+ if (lastequality != null && (lastequality.length <= length_changes1) && (lastequality.length <= length_changes2)) {
+ //alert("Splitting: '"+lastequality+"'");
+ diff.splice(equalities[equalities.length-1], 0, [-1, lastequality]); // Duplicate record
+ diff[equalities[equalities.length-1]+1][0] = 1; // Change second copy to insert.
+ equalities.pop(); // Throw away the equality we just deleted;
+ equalities.pop(); // Throw away the previous equality;
+ pointer = equalities.length ? equalities[equalities.length-1] : -1;
+ length_changes1 = 0; // Reset the counters.
+ length_changes2 = 0;
+ lastequality = null;
+ changes = true;
+ }
+ }
+ pointer++;
+ }
+
+ if (changes)
+ diff_cleanup_merge(diff);
+}
+
+
+function diff_cleanup_efficiency(diff) {
+ // Reduce the number of edits by eliminating operationally trivial equalities.
+ var changes = false;
+ var equalities = []; // Stack of indices where equalities are found.
+ var lastequality = ''; // Always equal to equalities[equalities.length-1][1]
+ var pointer = 0; // Index of current position.
+ var pre_ins = false; // Is there an insertion operation before the last equality.
+ var pre_del = false; // Is there an deletion operation before the last equality.
+ var post_ins = false; // Is there an insertion operation after the last equality.
+ var post_del = false; // Is there an deletion operation after the last equality.
+ while (pointer < diff.length) {
+ if (diff[pointer][0] == 0) { // equality found
+ if (diff[pointer][1].length < DIFF_EDIT_COST && (post_ins || post_del)) {
+ // Candidate found.
+ equalities.push(pointer);
+ pre_ins = post_ins;
+ pre_del = post_del;
+ lastequality = diff[pointer][1];
+ } else {
+ // Not a candidate, and can never become one.
+ equalities = [];
+ lastequality = '';
+ }
+ post_ins = post_del = false;
+ } else { // an insertion or deletion
+ if (diff[pointer][0] == -1)
+ post_del = true;
+ else
+ post_ins = true;
+ // Five types to be split:
+ // <ins>A</ins><del>B</del>XY<ins>C</ins><del>D</del>
+ // <ins>A</ins>X<ins>C</ins><del>D</del>
+ // <ins>A</ins><del>B</del>X<ins>C</ins>
+ // <ins>A</del>X<ins>C</ins><del>D</del>
+ // <ins>A</ins><del>B</del>X<del>C</del>
+ if (lastequality && ((pre_ins && pre_del && post_ins && post_del) || ((lastequality.length < DIFF_EDIT_COST/2) && (pre_ins + pre_del + post_ins + post_del) == 3))) {
+ //alert("Splitting: '"+lastequality+"'");
+ diff.splice(equalities[equalities.length-1], 0, [-1, lastequality]); // Duplicate record
+ diff[equalities[equalities.length-1]+1][0] = 1; // Change second copy to insert.
+ equalities.pop(); // Throw away the equality we just deleted;
+ lastequality = '';
+ if (pre_ins && pre_del) {
+ // No changes made which could affect previous entry, keep going.
+ post_ins = post_del = true;
+ equalities = [];
+ } else {
+ equalities.pop(); // Throw away the previous equality;
+ pointer = equalities.length ? equalities[equalities.length-1] : -1;
+ post_ins = post_del = false;
+ }
+ changes = true;
+ }
+ }
+ pointer++;
+ }
+
+ if (changes)
+ diff_cleanup_merge(diff);
+}
+
+
+function diff_cleanup_merge(diff) {
+ // Reorder and merge like edit sections. Merge equalities.
+ // Any edit section can move as long as it doesn't cross an equality.
+ diff.push([0, '']); // Add a dummy entry at the end.
+ var pointer = 0;
+ var count_delete = 0;
+ var count_insert = 0;
+ var text_delete = '';
+ var text_insert = '';
+ var record_insert, record_delete;
+ var my_xfix;
+ while(pointer < diff.length) {
+ if (diff[pointer][0] == 1) {
+ count_insert++;
+ text_insert += diff[pointer][1];
+ pointer++;
+ } else if (diff[pointer][0] == -1) {
+ count_delete++;
+ text_delete += diff[pointer][1];
+ pointer++;
+ } else { // Upon reaching an equality, check for prior redundancies.
+ if (count_delete > 1 || count_insert > 1) {
+ if (count_delete > 1 && count_insert > 1) {
+ // Factor out any common prefixies.
+ my_xfix = diff_prefix(text_insert, text_delete);
+ if (my_xfix[2] != '') {
+ if ((pointer - count_delete - count_insert) > 0 && diff[pointer - count_delete - count_insert - 1][0] == 0) {
+ text_insert = my_xfix[0];
+ text_delete = my_xfix[1];
+ diff[pointer - count_delete - count_insert - 1][1] += my_xfix[2];
+ }
+ }
+ // Factor out any common suffixies.
+ my_xfix = diff_suffix(text_insert, text_delete);
+ if (my_xfix[2] != '') {
+ text_insert = my_xfix[0];
+ text_delete = my_xfix[1];
+ diff[pointer][1] = my_xfix[2] + diff[pointer][1];
+ }
+ }
+ // Delete the offending records and add the merged ones.
+ if (count_delete == 0)
+ diff.splice(pointer - count_delete - count_insert, count_delete + count_insert, [1, text_insert]);
+ else if (count_insert == 0)
+ diff.splice(pointer - count_delete - count_insert, count_delete + count_insert, [-1, text_delete]);
+ else
+ diff.splice(pointer - count_delete - count_insert, count_delete + count_insert, [-1, text_delete], [1, text_insert]);
+ pointer = pointer - count_delete - count_insert + (count_delete ? 1 : 0) + (count_insert ? 1 : 0) + 1;
+ } else if (pointer != 0 && diff[pointer-1][0] == 0) {
+ // Merge this equality with the previous one.
+ diff[pointer-1][1] += diff[pointer][1];
+ diff.splice(pointer, 1);
+ } else {
+ pointer++;
+ }
+ count_insert = 0;
+ count_delete = 0;
+ text_delete = '';
+ text_insert = '';
+ }
+ }
+ if (diff[diff.length-1][1] == '')
+ diff.pop(); // Remove the dummy entry at the end.
+}
+
+
+function diff_addindex(diff) {
+ // Add an index to each tuple, represents where the tuple is located in text2.
+ // e.g. [[-1, 'h', 0], [1, 'c', 0], [0, 'at', 1]]
+ var i = 0;
+ for (var x=0; x<diff.length; x++) {
+ diff[x].push(i);
+ if (diff[x][0] != -1)
+ i += diff[x][1].length;
+ }
+}
+
+
+function diff_xindex(diff, loc) {
+ // loc is a location in text1, compute and return the equivalent location in text2.
+ // e.g. "The cat" vs "The big cat", 1->1, 5->8
+ var chars1 = 0;
+ var chars2 = 0;
+ var last_chars1 = 0;
+ var last_chars2 = 0;
+ for (var x=0; x<diff.length; x++) {
+ if (diff[x][0] != 1) // Equality or deletion.
+ chars1 += diff[x][1].length;
+ if (diff[x][0] != -1) // Equality or insertion.
+ chars2 += diff[x][1].length;
+ if (chars1 > loc) // Overshot the location.
+ break;
+ last_chars1 = chars1;
+ last_chars2 = chars2;
+ }
+ if (diff.length != x && diff[x][0] == -1) // The location was deleted.
+ return last_chars2;
+ // Add the remaining character length.
+ return last_chars2 + (loc - last_chars1);
+}
+
+
+function diff_prettyhtml(diff) {
+ // Convert a diff array into a pretty HTML report.
+ diff_addindex(diff);
+ var html = '';
+ for (var x=0; x<diff.length; x++) {
+ var m = diff[x][0]; // Mode (-1=delete, 0=copy, 1=add)
+ var t = diff[x][1]; // Text of change.
+ var i = diff[x][2]; // Index of change.
+ t = t.replace(/&/g, "&").replace(/</g, "<").replace(/>/g, ">");
+ t = t.replace(/\n/g, "¶<BR>");
+ if (m == -1)
+ html += "<DEL STYLE='background:#FFE6E6;' TITLE='i="+i+"'>"+t+"</DEL>";
+ else if (m == 1)
+ html += "<INS STYLE='background:#E6FFE6;' TITLE='i="+i+"'>"+t+"</INS>";
+ else
+ html += "<SPAN TITLE='i="+i+"'>"+t+"</SPAN>";
+ }
+ return html;
+}
+
+
+ //////////////////////////////////////////////////////////////////////
+ // Match //
+//////////////////////////////////////////////////////////////////////
+
+
+function match_getmaxbits() {
+ // Compute the number of bits in an int.
+ // The normal answer for JavaScript is 32.
+ var maxbits = 0;
+ var oldi = 1;
+ var newi = 2;
+ while (oldi != newi) {
+ maxbits++;
+ oldi = newi;
+ newi = newi << 1;
+ }
+ return maxbits;
+}
+var MATCH_MAXBITS = match_getmaxbits();
+
+
+function match_main(text, pattern, loc) {
+ // Locate the best instance of 'pattern' in 'text' near 'loc'.
+ loc = Math.max(0, Math.min(loc, text.length-pattern.length));
+ if (text == pattern) {
+ // Shortcut (potentially not guaranteed by the algorithm)
+ return 0;
+ } else if (text.length == 0) {
+ // Nothing to match.
+ return null;
+ } else if (text.substring(loc, loc + pattern.length) == pattern) {
+ // Perfect match at the perfect spot! (Includes case of null pattern)
+ return loc;
+ } else {
+ // Do a fuzzy compare.
+ var match = match_bitap(text, pattern, loc);
+ return match;
+ }
+}
+
+
+function match_bitap(text, pattern, loc) {
+ // Locate the best instance of 'pattern' in 'text' near 'loc' using the Bitap algorithm.
+ if (pattern.length > MATCH_MAXBITS)
+ return alert("Pattern too long for this browser.");
+
+ // Initialise the alphabet.
+ var s = match_alphabet(pattern);
+
+ var score_text_length = text.length;
+ // Coerce the text length between reasonable maximums and minimums.
+ score_text_length = Math.max(score_text_length, MATCH_MINLENGTH);
+ score_text_length = Math.min(score_text_length, MATCH_MAXLENGTH);
+
+ function match_bitap_score (e, x) {
+ // Compute and return the score for a match with e errors and x location.
+ var d = Math.abs(loc-x);
+ return (e / pattern.length / MATCH_BALANCE) + (d / score_text_length / (1.0 - MATCH_BALANCE));
+ }
+
+ // Highest score beyond which we give up.
+ var score_threshold = MATCH_THRESHOLD;
+ // Is there a nearby exact match? (speedup)
+ var best_loc = text.indexOf(pattern, loc);
+ if (best_loc != -1)
+ score_threshold = Math.min(match_bitap_score(0, best_loc), score_threshold);
+ // What about in the other direction? (speedup)
+ best_loc = text.lastIndexOf(pattern, loc+pattern.length);
+ if (best_loc != -1)
+ score_threshold = Math.min(match_bitap_score(0, best_loc), score_threshold);
+
+ // Initialise the bit arrays.
+ var r = Array();
+ var d = -1;
+ var matchmask = Math.pow(2, pattern.length-1);
+ best_loc = null;
+
+ var bin_min, bin_mid;
+ var bin_max = Math.max(loc+loc, text.length);
+ var last_rd;
+ for (var d=0; d<pattern.length; d++) {
+ // Scan for the best match; each iteration allows for one more error.
+ var rd = Array(text.length);
+
+ // Run a binary search to determine how far from 'loc' we can stray at this error level.
+ bin_min = loc;
+ bin_mid = bin_max;
+ while(bin_min < bin_mid) {
+ if (match_bitap_score(d, bin_mid) < score_threshold)
+ bin_min = bin_mid;
+ else
+ bin_max = bin_mid;
+ bin_mid = Math.floor((bin_max - bin_min) / 2 + bin_min);
+ }
+ bin_max = bin_mid; // Use the result from this iteration as the maximum for the next.
+ var start = Math.max(0, loc - (bin_mid - loc) - 1);
+ var finish = Math.min(text.length-1, pattern.length + bin_mid);
+
+ if (text.charAt(finish) == pattern.charAt(pattern.length-1))
+ rd[finish] = Math.pow(2, d+1)-1;
+ else
+ rd[finish] = Math.pow(2, d)-1;
+ for (var j=finish-1; j>=start; j--) {
+ // The alphabet (s) is a sparse hash, so the following lines generate warnings.
+ if (d == 0) // First pass: exact match.
+ rd[j] = ((rd[j+1] << 1) | 1) & s[text.charAt(j)];
+ else // Subsequent passes: fuzzy match.
+ rd[j] = ((rd[j+1] << 1) | 1) & s[text.charAt(j)] | ((last_rd[j+1] << 1) | 1) | ((last_rd[j] << 1) | 1) | last_rd[j+1];
+ if (rd[j] & matchmask) {
+ var score = match_bitap_score(d, j);
+ // This match will almost certainly be better than any existing match. But check anyway.
+ if (score <= score_threshold) {
+ // Told you so.
+ score_threshold = score;
+ best_loc = j;
+ if (j > loc) {
+ // When passing loc, don't exceed our current distance from loc.
+ start = Math.max(0, loc - (j - loc));
+ } else {
+ // Already passed loc, downhill from here on in.
+ break;
+ }
+ }
+ }
+ }
+ if (match_bitap_score(d+1, loc) > score_threshold) // No hope for a (better) match at greater error levels.
+ break;
+ last_rd = rd;
+ }
+ return best_loc;
+}
+
+
+function match_alphabet(pattern) {
+ // Initialise the alphabet for the Bitap algorithm.
+ var s = Object();
+ for (var i=0; i<pattern.length; i++)
+ s[pattern.charAt(i)] = 0;
+ for (var i=0; i<pattern.length; i++)
+ s[pattern.charAt(i)] |= Math.pow(2, pattern.length-i-1);
+ return s;
+}
+
+
+ //////////////////////////////////////////////////////////////////////
+ // Patch //
+//////////////////////////////////////////////////////////////////////
+
+
+function patch_obj() {
+ // Constructor for a patch object.
+ this.diffs = [];
+ this.start1 = null;
+ this.start2 = null;
+ this.length1 = 0;
+ this.length2 = 0;
+
+ this.toString = function() {
+ // Emmulate GNU diff's format.
+ // Header: @@ -382,8 +481,9 @@
+ // Indicies are printed as 1-based, not 0-based.
+ var coords1, coords2;
+ if (this.length1 == 0)
+ coords1 = this.start1+",0";
+ else if (this.length1 == 1)
+ coords1 = this.start1+1;
+ else
+ coords1 = (this.start1+1)+","+this.length1;
+ if (this.length2 == 0)
+ coords2 = this.start2+",0";
+ else if (this.length2 == 1)
+ coords2 = this.start2+1;
+ else
+ coords2 = (this.start2+1)+","+this.length2;
+ var txt = "@@ -"+coords1+" +"+coords2+" @@\n";
+ // Escape the body of the patch with %xx notation.
+ for (var x=0; x<this.diffs.length; x++)
+ txt += ("- +".charAt(this.diffs[x][0]+1)) + encodeURI(this.diffs[x][1]) + "\n";
+ return txt.replace(/%20/g, ' ');
+ }
+
+ this.text1 = function() {
+ // Compute and return the source text (all equalities and deletions).
+ var txt = '';
+ for (var x=0; x<this.diffs.length; x++)
+ if (this.diffs[x][0] == 0 || this.diffs[x][0] == -1)
+ txt += this.diffs[x][1];
+ return txt;
+ }
+
+ this.text2 = function() {
+ // Compute and return the destination text (all equalities and insertions).
+ var txt = '';
+ for (var x=0; x<this.diffs.length; x++)
+ if (this.diffs[x][0] == 0 || this.diffs[x][0] == 1)
+ txt += this.diffs[x][1];
+ return txt;
+ }
+}
+
+
+function patch_addcontext(patch, text) {
+ var pattern = text.substring(patch.start2, patch.start2+patch.length1);
+ var padding = 0;
+ // Increase the context until we're unique (but don't let the pattern expand beyond MATCH_MAXBITS).
+ while (text.indexOf(pattern) != text.lastIndexOf(pattern) && pattern.length < MATCH_MAXBITS-PATCH_MARGIN-PATCH_MARGIN) {
+ padding += PATCH_MARGIN;
+ pattern = text.substring(patch.start2 - padding, patch.start2+patch.length1 + padding);
+ }
+ // Add one chunk for good luck.
+ padding += PATCH_MARGIN;
+ // Add the prefix.
+ var prefix = text.substring(patch.start2 - padding, patch.start2);
+ if (prefix != '')
+ patch.diffs.unshift([0, prefix]);
+ // Add the suffix
+ var suffix = text.substring(patch.start2+patch.length1, patch.start2+patch.length1 + padding);
+ if (suffix != '')
+ patch.diffs.push([0, suffix]);
+
+ // Roll back the start points.
+ patch.start1 -= prefix.length;
+ patch.start2 -= prefix.length;
+ // Extend the lengths.
+ patch.length1 += prefix.length + suffix.length;
+ patch.length2 += prefix.length + suffix.length;
+}
+
+
+function patch_make(text1, text2, diff) {
+ // Compute a list of patches to turn text1 into text2.
+ // Use diff if provided, otherwise compute it ourselves.
+ if (typeof diff == 'undefined') {
+ diff = diff_main(text1, text2, true);
+ if (diff.length > 2) {
+ diff_cleanup_semantic(diff);
+ diff_cleanup_efficiency(diff);
+ }
+ }
+ if (diff.length == 0)
+ return []; // Get rid of the null case.
+ var patches = [];
+ var patch = new patch_obj();
+ var char_count1 = 0; // Number of characters into the text1 string.
+ var char_count2 = 0; // Number of characters into the text2 string.
+ var last_type = null;
+ var prepatch_text = text1; // Recreate the patches to determine context info.
+ var postpatch_text = text1;
+ for (var x=0; x<diff.length; x++) {
+ var diff_type = diff[x][0];
+ var diff_text = diff[x][1];
+
+ if (patch.diffs.length == 0 && diff_type != 0) {
+ // A new patch starts here.
+ patch.start1 = char_count1;
+ patch.start2 = char_count2;
+ }
+
+ if (diff_type == 1) {
+ // Insertion
+ patch.diffs.push(diff[x]);
+ patch.length2 += diff_text.length;
+ postpatch_text = postpatch_text.substring(0, char_count2) + diff_text + postpatch_text.substring(char_count2);
+ } else if (diff_type == -1) {
+ // Deletion.
+ patch.length1 += diff_text.length;
+ patch.diffs.push(diff[x]);
+ postpatch_text = postpatch_text.substring(0, char_count2) + postpatch_text.substring(char_count2 + diff_text.length);
+ } else if (diff_type == 0 && diff_text.length <= 2*PATCH_MARGIN && patch.diffs.length != 0 && diff.length != x+1) {
+ // Small equality inside a patch.
+ patch.diffs.push(diff[x]);
+ patch.length1 += diff_text.length;
+ patch.length2 += diff_text.length;
+ }
+
+ last_type = diff_type;
+ if (diff_type == 0 && diff_text.length >= 2*PATCH_MARGIN) {
+ // Time for a new patch.
+ if (patch.diffs.length != 0) {
+ patch_addcontext(patch, prepatch_text);
+ patches.push(patch);
+ var patch = new patch_obj();
+ last_type = null;
+ prepatch_text = postpatch_text;
+ }
+ }
+
+ // Update the current character count.
+ if (diff_type != 1)
+ char_count1 += diff_text.length;
+ if (diff_type != -1)
+ char_count2 += diff_text.length;
+ }
+ // Pick up the leftover patch if not empty.
+ if (patch.diffs.length != 0) {
+ patch_addcontext(patch, prepatch_text);
+ patches.push(patch);
+ }
+
+ return patches;
+}
+
+
+function patch_apply(patches, text) {
+ // Merge a set of patches onto the text.
+ // Return a patched text, as well as a list of true/false values indicating which patches were applied.
+ patch_splitmax(patches);
+ var results = [];
+ var delta = 0;
+ var expected_loc, start_loc;
+ var text1, text2;
+ var diff, mod, index1, index2;
+ for (var x=0; x<patches.length; x++) {
+ expected_loc = patches[x].start2 + delta;
+ text1 = patches[x].text1();
+ start_loc = match_main(text, text1, expected_loc);
+ if (start_loc == null) {
+ // No match found. :(
+ results.push(false);
+ } else {
+ // Found a match. :)
+ results.push(true);
+ delta = start_loc - expected_loc;
+ text2 = text.substring(start_loc, start_loc + text1.length);
+ if (text1 == text2) {
+ // Perfect match, just shove the replacement text in.
+ text = text.substring(0, start_loc) + patches[x].text2() + text.substring(start_loc + text1.length);
+ } else {
+ // Imperfect match. Run a diff to get a framework of equivalent indicies.
+ diff = diff_main(text1, text2, false);
+ index1 = 0;
+ for (var y=0; y<patches[x].diffs.length; y++) {
+ mod = patches[x].diffs[y];
+ if (mod[0] != 0)
+ index2 = diff_xindex(diff, index1);
+ if (mod[0] == 1) // Insertion
+ text = text.substring(0, start_loc + index2) + mod[1] + text.substring(start_loc + index2);
+ else if (mod[0] == -1) // Deletion
+ text = text.substring(0, start_loc + index2) + text.substring(start_loc + diff_xindex(diff, index1 + mod[1].length));
+ if (mod[0] != -1)
+ index1 += mod[1].length;
+ }
+ }
+ }
+ }
+ return [text, results];
+}
+
+
+function patch_splitmax(patches) {
+ // Look through the patches and break up any which are longer than the maximum limit of the match algorithm.
+ var bigpatch, patch, patch_size, start1, start2, diff_type, diff_text, precontext, postcontext, empty;
+ for (var x=0; x<patches.length; x++) {
+ if (patches[x].length1 > MATCH_MAXBITS) {
+ bigpatch = patches[x];
+ // Remove the big old patch.
+ patches.splice(x, 1);
+ patch_size = MATCH_MAXBITS;
+ start1 = bigpatch.start1;
+ start2 = bigpatch.start2;
+ precontext = '';
+ while (bigpatch.diffs.length != 0) {
+ // Create one of several smaller patches.
+ patch = new patch_obj();
+ empty = true;
+ patch.start1 = start1 - precontext.length;
+ patch.start2 = start2 - precontext.length;
+ if (precontext != '') {
+ patch.length1 = patch.length2 = precontext.length;
+ patch.diffs.push([0, precontext]);
+ }
+ while (bigpatch.diffs.length != 0 && patch.length1 < patch_size - PATCH_MARGIN) {
+ diff_type = bigpatch.diffs[0][0];
+ diff_text = bigpatch.diffs[0][1];
+ if (diff_type == 1) {
+ // Insertions are harmless.
+ patch.length2 += diff_text.length;
+ start2 += diff_text.length;
+ patch.diffs.push(bigpatch.diffs.shift());
+ empty = false;
+ } else {
+ // Deletion or equality. Only take as much as we can stomach.
+ diff_text = diff_text.substring(0, patch_size - patch.length1 - PATCH_MARGIN);
+ patch.length1 += diff_text.length;
+ start1 += diff_text.length;
+ if (diff_type == 0) {
+ patch.length2 += diff_text.length;
+ start2 += diff_text.length;
+ } else {
+ empty = false;
+ }
+ patch.diffs.push([diff_type, diff_text]);
+ if (diff_text == bigpatch.diffs[0][1])
+ bigpatch.diffs.shift();
+ else
+ bigpatch.diffs[0][1] = bigpatch.diffs[0][1].substring(diff_text.length);
+ }
+ }
+ // Compute the head context for the next patch.
+ precontext = patch.text2();
+ precontext = precontext.substring(precontext.length - PATCH_MARGIN);
+ // Append the end context for this patch.
+ postcontext = bigpatch.text1().substring(0, PATCH_MARGIN);
+ if (postcontext != '') {
+ patch.length1 += postcontext.length;
+ patch.length2 += postcontext.length;
+ if (patch.diffs.length > 0 && patch.diffs[patch.diffs.length-1][0] == 0)
+ patch.diffs[patch.diffs.length-1][1] += postcontext;
+ else
+ patch.diffs.push([0, postcontext]);
+ }
+ if (!empty)
+ patches.splice(x++, 0, patch);
+ }
+ }
+ }
+}
+
+
+function patch_totext(patches) {
+ // Take a list of patches and return a textual representation.
+ var text = '';
+ for (var x=0; x<patches.length; x++)
+ text += patches[x];
+ return text;
+}
+
+
+function patch_fromtext(text) {
+ // Take a textual representation of patches and return a list of patch objects.
+ var patches = [];
+ text = text.split('\n');
+ var patch, m, chars1, chars2, sign, line;
+ while (text.length != 0) {
+ m = text[0].match(/^@@ -(\d+),?(\d*) \+(\d+),?(\d*) @@$/);
+ if (!m)
+ return alert("Invalid patch string:\n"+text[0]);
+ patch = new patch_obj();
+ patches.push(patch);
+ patch.start1 = parseInt(m[1]);
+ if (m[2] == '') {
+ patch.start1--;
+ patch.length1 = 1;
+ } else if (m[2] == '0') {
+ patch.length1 = 0;
+ } else {
+ patch.start1--;
+ patch.length1 = parseInt(m[2]);
+ }
+
+ patch.start2 = parseInt(m[3]);
+ if (m[4] == '') {
+ patch.start2--;
+ patch.length2 = 1;
+ } else if (m[4] == '0') {
+ patch.length2 = 0;
+ } else {
+ patch.start2--;
+ patch.length2 = parseInt(m[4]);
+ }
+ text.shift();
+
+ while (text.length != 0) {
+ sign = text[0].charAt(0);
+ line = decodeURIComponent(text[0].substring(1));
+ if (sign == '-') {
+ // Deletion.
+ patch.diffs.push([-1, line]);
+ } else if (sign == '+') {
+ // Insertion.
+ patch.diffs.push([1, line]);
+ } else if (sign == ' ') {
+ // Minor equality.
+ patch.diffs.push([0, line]);
+ } else if (sign == '@') {
+ // Start of next patch.
+ break;
+ } else if (sign == '') {
+ // Blank line? Whatever.
+ } else {
+ // WTF?
+ return alert("Invalid patch mode: '"+sign+"'\n"+line);
+ }
+ text.shift();
+ }
+ }
+ return patches;
+}
+
+// EOF