prosemirror-changeset - npm Package Compare versions

Comparing version 2.2.1 to 2.3.0

CHANGELOG.md

@@ -0,1 +1,7 @@
+## 2.3.0 (2025-05-05)
+
+### New features
+
+Change sets can now be passed a custom token encoder that controls the way changed content is diffed.
+
 ## 2.2.1 (2023-05-17)
@@ -2,0 +8,0 @@

dist/index.cjs

 'use strict';
 
 function _toConsumableArray(arr) { return _arrayWithoutHoles(arr) || _iterableToArray(arr) || _unsupportedIterableToArray(arr) || _nonIterableSpread(); }
-
 function _nonIterableSpread() { throw new TypeError("Invalid attempt to spread non-iterable instance.\nIn order to be iterable, non-array objects must have a [Symbol.iterator]() method."); }
-
 function _unsupportedIterableToArray(o, minLen) { if (!o) return; if (typeof o === "string") return _arrayLikeToArray(o, minLen); var n = Object.prototype.toString.call(o).slice(8, -1); if (n === "Object" && o.constructor) n = o.constructor.name; if (n === "Map" || n === "Set") return Array.from(o); if (n === "Arguments" || /^(?:Ui|I)nt(?:8|16|32)(?:Clamped)?Array$/.test(n)) return _arrayLikeToArray(o, minLen); }
-
 function _iterableToArray(iter) { if (typeof Symbol !== "undefined" && iter[Symbol.iterator] != null || iter["@@iterator"] != null) return Array.from(iter); }
-
 function _arrayWithoutHoles(arr) { if (Array.isArray(arr)) return _arrayLikeToArray(arr); }
-
-function _arrayLikeToArray(arr, len) { if (len == null || len > arr.length) len = arr.length; for (var i = 0, arr2 = new Array(len); i < len; i++) { arr2[i] = arr[i]; } return arr2; }
-
+function _arrayLikeToArray(arr, len) { if (len == null || len > arr.length) len = arr.length; for (var i = 0, arr2 = new Array(len); i < len; i++) arr2[i] = arr[i]; return arr2; }
+function _typeof(o) { "@babel/helpers - typeof"; return _typeof = "function" == typeof Symbol && "symbol" == typeof Symbol.iterator ? function (o) { return typeof o; } : function (o) { return o && "function" == typeof Symbol && o.constructor === Symbol && o !== Symbol.prototype ? "symbol" : typeof o; }, _typeof(o); }
 function _classCallCheck(instance, Constructor) { if (!(instance instanceof Constructor)) { throw new TypeError("Cannot call a class as a function"); } }
-
-function _defineProperties(target, props) { for (var i = 0; i < props.length; i++) { var descriptor = props[i]; descriptor.enumerable = descriptor.enumerable || false; descriptor.configurable = true; if ("value" in descriptor) descriptor.writable = true; Object.defineProperty(target, descriptor.key, descriptor); } }
-
+function _defineProperties(target, props) { for (var i = 0; i < props.length; i++) { var descriptor = props[i]; descriptor.enumerable = descriptor.enumerable || false; descriptor.configurable = true; if ("value" in descriptor) descriptor.writable = true; Object.defineProperty(target, _toPropertyKey(descriptor.key), descriptor); } }
 function _createClass(Constructor, protoProps, staticProps) { if (protoProps) _defineProperties(Constructor.prototype, protoProps); if (staticProps) _defineProperties(Constructor, staticProps); Object.defineProperty(Constructor, "prototype", { writable: false }); return Constructor; }
-
-function _typeof(obj) { "@babel/helpers - typeof"; return _typeof = "function" == typeof Symbol && "symbol" == typeof Symbol.iterator ? function (obj) { return typeof obj; } : function (obj) { return obj && "function" == typeof Symbol && obj.constructor === Symbol && obj !== Symbol.prototype ? "symbol" : typeof obj; }, _typeof(obj); }
-
-Object.defineProperty(exports, '__esModule', {
-  value: true
-});
-
-function tokens(frag, start, end, target) {
+function _toPropertyKey(arg) { var key = _toPrimitive(arg, "string"); return _typeof(key) === "symbol" ? key : String(key); }
+function _toPrimitive(input, hint) { if (_typeof(input) !== "object" || input === null) return input; var prim = input[Symbol.toPrimitive]; if (prim !== undefined) { var res = prim.call(input, hint || "default"); if (_typeof(res) !== "object") return res; throw new TypeError("@@toPrimitive must return a primitive value."); } return (hint === "string" ? String : Number)(input); }
+var DefaultEncoder = {
+  encodeCharacter: function encodeCharacter(_char) {
+    return _char;
+  },
+  encodeNodeStart: function encodeNodeStart(node) {
+    return node.type.name;
+  },
+  encodeNodeEnd: function encodeNodeEnd() {
+    return -1;
+  },
+  compareTokens: function compareTokens(a, b) {
+    return a === b;
+  }
+};
+function tokens(frag, encoder, start, end, target) {
   for (var i = 0, off = 0; i < frag.childCount; i++) {
     var child = frag.child(i),
-        endOff = off + child.nodeSize;
+      endOff = off + child.nodeSize;
     var from = Math.max(off, start),
-        to = Math.min(endOff, end);
-
+      to = Math.min(endOff, end);
     if (from < to) {
       if (child.isText) {
-        for (var j = from; j < to; j++) {
-          target.push(child.text.charCodeAt(j - off));
-        }
+        for (var j = from; j < to; j++) target.push(encoder.encodeCharacter(child.text.charCodeAt(j - off), child.marks));
       } else if (child.isLeaf) {
-        target.push(child.type.name);
+        target.push(encoder.encodeNodeStart(child));
       } else {
-        if (from == off) target.push(child.type.name);
-        tokens(child.content, Math.max(off + 1, from) - off - 1, Math.min(endOff - 1, to) - off - 1, target);
-        if (to == endOff) target.push(-1);
+        if (from == off) target.push(encoder.encodeNodeStart(child));
+        tokens(child.content, encoder, Math.max(off + 1, from) - off - 1, Math.min(endOff - 1, to) - off - 1, target);
+        if (to == endOff) target.push(encoder.encodeNodeEnd(child));
       }
     }
-
     off = endOff;
   }
-
   return target;
 }
-
 var MAX_DIFF_SIZE = 5000;
-
 function minUnchanged(sizeA, sizeB) {
   return Math.min(15, Math.max(2, Math.floor(Math.max(sizeA, sizeB) / 10)));
 }
-
 function computeDiff(fragA, fragB, range) {
-  var tokA = tokens(fragA, range.fromA, range.toA, []);
-  var tokB = tokens(fragB, range.fromB, range.toB, []);
+  var encoder = arguments.length > 3 && arguments[3] !== undefined ? arguments[3] : DefaultEncoder;
+  var tokA = tokens(fragA, encoder, range.fromA, range.toA, []);
+  var tokB = tokens(fragB, encoder, range.fromB, range.toB, []);
   var start = 0,
-      endA = tokA.length,
-      endB = tokB.length;
-
-  while (start < tokA.length && start < tokB.length && tokA[start] === tokB[start]) {
-    start++;
-  }
-
+    endA = tokA.length,
+    endB = tokB.length;
+  var cmp = encoder.compareTokens;
+  while (start < tokA.length && start < tokB.length && cmp(tokA[start], tokB[start])) start++;
   if (start == tokA.length && start == tokB.length) return [];
-
-  while (endA > start && endB > start && tokA[endA - 1] === tokB[endB - 1]) {
-    endA--, endB--;
-  }
-
+  while (endA > start && endB > start && cmp(tokA[endA - 1], tokB[endB - 1])) endA--, endB--;
   if (endA == start || endB == start || endA == endB && endA == start + 1) return [range.slice(start, endA, start, endB)];
   var lenA = endA - start,
-      lenB = endB - start;
+    lenB = endB - start;
   var max = Math.min(MAX_DIFF_SIZE, lenA + lenB),
-      off = max + 1;
+    off = max + 1;
   var history = [];
   var frontier = [];
-
-  for (var len = off * 2, i = 0; i < len; i++) {
-    frontier[i] = -1;
-  }
-
+  for (var len = off * 2, i = 0; i < len; i++) frontier[i] = -1;
   for (var size = 0; size <= max; size++) {
-    for (var diag = -size; diag <= size; diag += 2) {
-      var next = frontier[diag + 1 + max],
-          prev = frontier[diag - 1 + max];
-      var x = next < prev ? prev : next + 1,
-          y = x + diag;
-
-      while (x < lenA && y < lenB && tokA[start + x] === tokB[start + y]) {
-        x++, y++;
-      }
-
-      frontier[diag + max] = x;
-
-      if (x >= lenA && y >= lenB) {
-        var _ret = function () {
+    var _loop = function _loop(_diag) {
+        var next = frontier[_diag + 1 + max],
+          prev = frontier[_diag - 1 + max];
+        var x = next < prev ? prev : next + 1,
+          y = x + _diag;
+        while (x < lenA && y < lenB && cmp(tokA[start + x], tokB[start + y])) x++, y++;
+        frontier[_diag + max] = x;
+        if (x >= lenA && y >= lenB) {
           var diff = [],
-              minSpan = minUnchanged(endA - start, endB - start);
+            minSpan = minUnchanged(endA - start, endB - start);
           var fromA = -1,
-              toA = -1,
-              fromB = -1,
-              toB = -1;
-
+            toA = -1,
+            fromB = -1,
+            toB = -1;
           var add = function add(fA, tA, fB, tB) {
@@ -123,22 +100,18 @@ if (fromA > -1 && fromA < tA + minSpan) {
           };
-
           for (var _i = size - 1; _i >= 0; _i--) {
-            var _next = frontier[diag + 1 + max],
-                _prev = frontier[diag - 1 + max];
-
+            var _next = frontier[_diag + 1 + max],
+              _prev = frontier[_diag - 1 + max];
             if (_next < _prev) {
-              diag--;
+              _diag--;
               x = _prev + start;
-              y = x + diag;
+              y = x + _diag;
               add(x, x, y, y + 1);
             } else {
-              diag++;
+              _diag++;
               x = _next + start;
-              y = x + diag;
+              y = x + _diag;
               add(x, x + 1, y, y);
             }
-
             frontier = history[_i >> 1];
           }
-
           if (fromA > -1) diff.push(range.slice(fromA, toA, fromB, toB));
@@ -148,22 +121,20 @@ return {
           };
-        }();
-
-        if (_typeof(_ret) === "object") return _ret.v;
-      }
+        }
+        diag = _diag;
+      },
+      _ret;
+    for (var diag = -size; diag <= size; diag += 2) {
+      _ret = _loop(diag);
+      if (_ret) return _ret.v;
     }
-
     if (size % 2 == 0) history.push(frontier.slice());
   }
-
   return [range.slice(start, endA, start, endB)];
 }
-
 var Span = function () {
   function Span(length, data) {
     _classCallCheck(this, Span);
-
     this.length = length;
     this.data = data;
   }
-
   _createClass(Span, [{
@@ -180,6 +151,5 @@ key: "cut",
       var result = [];
-
       for (var i = 0, off = 0; off < to; i++) {
         var span = spans[i],
-            end = off + span.length;
+          end = off + span.length;
         var overlap = Math.min(to, end) - Math.max(from, off);
@@ -189,3 +159,2 @@ if (overlap > 0) result.push(span.cut(overlap));
       }
-
       return result;
@@ -202,7 +171,3 @@ }
       result.push(new Span(a[a.length - 1].length + b[0].length, combined));
-
-      for (var i = 1; i < b.length; i++) {
-        result.push(b[i]);
-      }
-
+      for (var i = 1; i < b.length; i++) result.push(b[i]);
       return result;
@@ -214,20 +179,12 @@ }
       var len = 0;
-
-      for (var i = 0; i < spans.length; i++) {
-        len += spans[i].length;
-      }
-
+      for (var i = 0; i < spans.length; i++) len += spans[i].length;
       return len;
     }
   }]);
-
   return Span;
 }();
-
 Span.none = [];
-
 var Change = function () {
   function Change(fromA, toA, fromB, toB, deleted, inserted) {
     _classCallCheck(this, Change);
-
     this.fromA = fromA;
@@ -240,3 +197,2 @@ this.toA = toA;
   }
-
   _createClass(Change, [{
@@ -264,3 +220,2 @@ key: "lenA",
       var result = [];
-
       for (var iX = 0, iY = 0, curX = x[0], curY = y[0];;) {
@@ -275,3 +230,2 @@ if (!curX && !curY) {
           var _off = iX ? x[iX - 1].toB - x[iX - 1].toA : 0;
-
           result.push(_off == 0 ? curY : new Change(curY.fromA - _off, curY.toA - _off, curY.fromB, curY.toB, curY.deleted, curY.inserted));
@@ -282,10 +236,9 @@ curY = iY++ == y.length ? null : y[iY];
           var fromA = Math.min(curX.fromA, curY.fromA - (iX ? x[iX - 1].toB - x[iX - 1].toA : 0)),
-              toA = fromA;
+            toA = fromA;
           var fromB = Math.min(curY.fromB, curX.fromB + (iY ? y[iY - 1].toB - y[iY - 1].toA : 0)),
-              toB = fromB;
+            toB = fromB;
           var deleted = Span.none,
-              inserted = Span.none;
+            inserted = Span.none;
           var enteredX = false,
-              enteredY = false;
-
+            enteredY = false;
           for (;;) {
@@ -296,5 +249,4 @@ var nextX = !curX ? 2e8 : pos >= curX.fromB ? curX.toB : curX.fromB;
             var inX = curX && pos >= curX.fromB,
-                inY = curY && pos >= curY.fromA;
+              inY = curY && pos >= curY.fromA;
             if (!inX && !inY) break;
-
             if (inX && pos == curX.fromB && !enteredX) {
@@ -305,3 +257,2 @@ deleted = Span.join(deleted, curX.deleted, combine);
             }
-
             if (inX && !inY) {
@@ -311,3 +262,2 @@ inserted = Span.join(inserted, Span.slice(curX.inserted, pos - curX.fromB, next - curX.fromB), combine);
             }
-
             if (inY && pos == curY.fromA && !enteredY) {
@@ -318,3 +268,2 @@ inserted = Span.join(inserted, curY.inserted, combine);
             }
-
             if (inY && !inX) {
@@ -324,3 +273,2 @@ deleted = Span.join(deleted, Span.slice(curY.deleted, pos - curY.fromA, next - curY.fromA), combine);
             }
-
             if (inX && next == curX.toB) {
@@ -330,3 +278,2 @@ curX = iX++ == x.length ? null : x[iX];
             }
-
             if (inY && next == curY.toA) {
@@ -336,6 +283,4 @@ curY = iY++ == y.length ? null : y[iY];
             }
-
             pos = next;
           }
-
           if (fromA < toA || fromB < toB) result.push(new Change(fromA, toA, fromB, toB, deleted, inserted));
@@ -346,14 +291,9 @@ }
   }]);
-
   return Change;
 }();
-
 var letter;
-
 try {
   letter = new RegExp("[\\p{Alphabetic}_]", "u");
 } catch (_) {}
-
 var nonASCIISingleCaseWordChar = /[\u00df\u0587\u0590-\u05f4\u0600-\u06ff\u3040-\u309f\u30a0-\u30ff\u3400-\u4db5\u4e00-\u9fcc\uac00-\ud7af]/;
-
 function isLetter(code) {
@@ -365,13 +305,10 @@ if (code < 128) return code >= 48 && code <= 57 || code >= 65 && code <= 90 || code >= 79 && code <= 122;
 }
-
 function getText(frag, start, end) {
   var out = "";
-
   function convert(frag, start, end) {
     for (var i = 0, off = 0; i < frag.childCount; i++) {
       var child = frag.child(i),
-          endOff = off + child.nodeSize;
+        endOff = off + child.nodeSize;
       var from = Math.max(off, start),
-          to = Math.min(endOff, end);
-
+        to = Math.min(endOff, end);
       if (from < to) {
@@ -388,30 +325,19 @@ if (child.isText) {
       }
-
       off = endOff;
     }
   }
-
   convert(frag, start, end);
   return out;
 }
-
 var MAX_SIMPLIFY_DISTANCE = 30;
-
 function simplifyChanges(changes, doc) {
   var result = [];
-
   for (var i = 0; i < changes.length; i++) {
     var end = changes[i].toB,
-        start = i;
-
-    while (i < changes.length - 1 && changes[i + 1].fromB <= end + MAX_SIMPLIFY_DISTANCE) {
-      end = changes[++i].toB;
-    }
-
+      start = i;
+    while (i < changes.length - 1 && changes[i + 1].fromB <= end + MAX_SIMPLIFY_DISTANCE) end = changes[++i].toB;
     simplifyAdjacentChanges(changes, start, i + 1, doc, result);
   }
-
   return result;
 }
-
 function simplifyAdjacentChanges(changes, from, to, doc, target) {
@@ -421,14 +347,11 @@ var start = Math.max(0, changes[from].fromB - MAX_SIMPLIFY_DISTANCE);
   var text = getText(doc.content, start, end);
-
   for (var i = from; i < to; i++) {
     var startI = i,
-        last = changes[i],
-        deleted = last.lenA,
-        inserted = last.lenB;
-
+      last = changes[i],
+      deleted = last.lenA,
+      inserted = last.lenB;
     while (i < to - 1) {
       var next = changes[i + 1],
-          boundary = false;
+        boundary = false;
       var prevLetter = last.toB == end ? false : isLetter(text.charCodeAt(last.toB - 1 - start));
-
       for (var pos = last.toB; !boundary && pos < next.fromB; pos++) {
@@ -439,3 +362,2 @@ var nextLetter = pos == end ? false : isLetter(text.charCodeAt(pos - start));
       }
-
       if (boundary) break;
@@ -447,31 +369,19 @@ deleted += next.lenA;
     }
-
     if (inserted > 0 && deleted > 0 && !(inserted == 1 && deleted == 1)) {
       var _from = changes[startI].fromB,
-          _to = changes[i].toB;
-      if (_from < end && isLetter(text.charCodeAt(_from - start))) while (_from > start && isLetter(text.charCodeAt(_from - 1 - start))) {
-        _from--;
-      }
-      if (_to > start && isLetter(text.charCodeAt(_to - 1 - start))) while (_to < end && isLetter(text.charCodeAt(_to - start))) {
-        _to++;
-      }
+        _to = changes[i].toB;
+      if (_from < end && isLetter(text.charCodeAt(_from - start))) while (_from > start && isLetter(text.charCodeAt(_from - 1 - start))) _from--;
+      if (_to > start && isLetter(text.charCodeAt(_to - 1 - start))) while (_to < end && isLetter(text.charCodeAt(_to - start))) _to++;
       var joined = fillChange(changes.slice(startI, i + 1), _from, _to);
-
       var _last = target.length ? target[target.length - 1] : null;
-
       if (_last && _last.toA == joined.fromA) target[target.length - 1] = new Change(_last.fromA, joined.toA, _last.fromB, joined.toB, _last.deleted.concat(joined.deleted), _last.inserted.concat(joined.inserted));else target.push(joined);
     } else {
-      for (var j = startI; j <= i; j++) {
-        target.push(changes[j]);
-      }
+      for (var j = startI; j <= i; j++) target.push(changes[j]);
     }
   }
-
   return changes;
 }
-
 function combine(a, b) {
   return a === b ? a : null;
 }
-
 function fillChange(changes, fromB, toB) {
@@ -482,10 +392,9 @@ var fromA = changes[0].fromA - (changes[0].fromB - fromB);
   var deleted = Span.none,
-      inserted = Span.none;
+    inserted = Span.none;
   var delData = (changes[0].deleted.length ? changes[0].deleted : changes[0].inserted)[0].data;
   var insData = (changes[0].inserted.length ? changes[0].inserted : changes[0].deleted)[0].data;
-
   for (var posA = fromA, posB = fromB, i = 0;; i++) {
     var next = i == changes.length ? null : changes[i];
     var endA = next ? next.fromA : toA,
-        endB = next ? next.fromB : toB;
+      endB = next ? next.fromB : toB;
     if (endA > posA) deleted = Span.join(deleted, [new Span(endA - posA, delData)], combine);
@@ -501,14 +410,10 @@ if (endB > posB) inserted = Span.join(inserted, [new Span(endB - posB, insData)], combine);
   }
-
   return new Change(fromA, toA, fromB, toB, deleted, inserted);
 }
-
 var ChangeSet = function () {
   function ChangeSet(config, changes) {
     _classCallCheck(this, ChangeSet);
-
     this.config = config;
     this.changes = changes;
   }
-
   _createClass(ChangeSet, [{
@@ -518,6 +423,4 @@ key: "addSteps",
       var _this = this;
-
       var stepChanges = [];
-
-      var _loop = function _loop(i) {
+      var _loop2 = function _loop2() {
         var d = Array.isArray(data) ? data[i] : data;
@@ -530,7 +433,5 @@ var off = 0;
       };
-
       for (var i = 0; i < maps.length; i++) {
-        _loop(i);
+        _loop2();
       }
-
       if (stepChanges.length == 0) return this;
@@ -540,41 +441,30 @@ var newChanges = mergeAll(stepChanges, this.config.combine);
       var updated = changes;
-
-      var _loop2 = function _loop2(_i3) {
-        var change = updated[_i3];
-
-        if (change.fromA == change.toA || change.fromB == change.toB || !newChanges.some(function (r) {
-          return r.toB > change.fromB && r.fromB < change.toB;
-        })) {
+      var _loop3 = function _loop3(_i3) {
+          var change = updated[_i3];
+          if (change.fromA == change.toA || change.fromB == change.toB || !newChanges.some(function (r) {
+            return r.toB > change.fromB && r.fromB < change.toB;
+          })) {
+            _i2 = _i3;
+            return 0;
+          }
+          var diff = computeDiff(_this.config.doc.content, newDoc.content, change, _this.config.encoder);
+          if (diff.length == 1 && diff[0].fromB == 0 && diff[0].toB == change.toB - change.fromB) {
+            _i2 = _i3;
+            return 0;
+          }
+          if (updated == changes) updated = changes.slice();
+          if (diff.length == 1) {
+            updated[_i3] = diff[0];
+          } else {
+            var _updated;
+            (_updated = updated).splice.apply(_updated, [_i3, 1].concat(_toConsumableArray(diff)));
+            _i3 += diff.length - 1;
+          }
           _i2 = _i3;
-          return "continue";
-        }
-
-        var diff = computeDiff(_this.config.doc.content, newDoc.content, change);
-
-        if (diff.length == 1 && diff[0].fromB == 0 && diff[0].toB == change.toB - change.fromB) {
-          _i2 = _i3;
-          return "continue";
-        }
-
-        if (updated == changes) updated = changes.slice();
-
-        if (diff.length == 1) {
-          updated[_i3] = diff[0];
-        } else {
-          var _updated;
-
-          (_updated = updated).splice.apply(_updated, [_i3, 1].concat(_toConsumableArray(diff)));
-
-          _i3 += diff.length - 1;
-        }
-
-        _i2 = _i3;
-      };
-
+        },
+        _ret2;
       for (var _i2 = 0; _i2 < updated.length; _i2++) {
-        var _ret2 = _loop2(_i2);
-
-        if (_ret2 === "continue") continue;
+        _ret2 = _loop3(_i2);
+        if (_ret2 === 0) continue;
       }
-
       return new ChangeSet(this.config, updated);
@@ -594,3 +484,2 @@ }
       };
-
       return new ChangeSet(this.config, this.changes.map(function (ch) {
@@ -606,10 +495,7 @@ return new Change(ch.fromA, ch.toA, ch.fromB, ch.toB, ch.deleted.map(mapSpan), ch.inserted.map(mapSpan));
       var moved = touched ? touched.toB - touched.fromB - (touched.toA - touched.fromA) : 0;
-
       function map(p) {
         return !touched || p <= touched.fromA ? p : p + moved;
       }
-
       var from = touched ? touched.fromB : 2e8,
-          to = touched ? touched.toB : -2e8;
-
+        to = touched ? touched.toB : -2e8;
       function add(start) {
@@ -620,10 +506,7 @@ var end = arguments.length > 1 && arguments[1] !== undefined ? arguments[1] : start;
       }
-
       var rA = this.changes,
-          rB = b.changes;
-
+        rB = b.changes;
       for (var iA = 0, iB = 0; iA < rA.length && iB < rB.length;) {
         var rangeA = rA[iA],
-            rangeB = rB[iB];
-
+          rangeB = rB[iB];
         if (rangeA && rangeB && sameRanges(rangeA, rangeB, map)) {
@@ -640,3 +523,2 @@ iA++;
       }
-
       return from <= to ? {
@@ -653,14 +535,13 @@ from: from,
       };
+      var tokenEncoder = arguments.length > 2 && arguments[2] !== undefined ? arguments[2] : DefaultEncoder;
       return new ChangeSet({
         combine: combine,
-        doc: doc
+        doc: doc,
+        encoder: tokenEncoder
       }, []);
     }
   }]);
-
   return ChangeSet;
 }();
-
 ChangeSet.computeDiff = computeDiff;
-
 function mergeAll(ranges, combine) {
@@ -673,10 +554,7 @@ var start = arguments.length > 2 && arguments[2] !== undefined ? arguments[2] : 0;
 }
-
 function endRange(maps) {
   var from = 2e8,
-      to = -2e8;
-
+    to = -2e8;
   for (var i = 0; i < maps.length; i++) {
     var map = maps[i];
-
     if (from != 2e8) {
@@ -686,3 +564,2 @@ from = map.map(from, -1);
     }
-
     map.forEach(function (_s, _e, start, end) {
@@ -693,3 +570,2 @@ from = Math.min(from, start);
   }
-
   return from == 2e8 ? null : {
@@ -700,3 +576,2 @@ from: from,
 }
-
 function touchedRange(maps) {
@@ -715,17 +590,10 @@ var b = endRange(maps);
 }
-
 function sameRanges(a, b, map) {
   return map(a.fromB) == b.fromB && map(a.toB) == b.toB && sameSpans(a.deleted, b.deleted) && sameSpans(a.inserted, b.inserted);
 }
-
 function sameSpans(a, b) {
   if (a.length != b.length) return false;
-
-  for (var i = 0; i < a.length; i++) {
-    if (a[i].length != b[i].length || a[i].data !== b[i].data) return false;
-  }
-
+  for (var i = 0; i < a.length; i++) if (a[i].length != b[i].length || a[i].data !== b[i].data) return false;
   return true;
 }
-
 exports.Change = Change;
@@ -732,0 +600,0 @@ exports.ChangeSet = ChangeSet;

dist/index.d.cts

@@ -1,2 +0,2 @@
-import { Node } from 'prosemirror-model';
+import { Mark, Node } from 'prosemirror-model';
 import { StepMap } from 'prosemirror-transform';
@@ -44,3 +44,3 @@
     Data associated with the inserted content. Length adds up to
-    `this.toB - this.toA`.
+    `this.toB - this.fromB`.
     */
@@ -57,2 +57,34 @@ readonly inserted: readonly Span<Data>[];
 /**
+A token encoder can be passed when creating a `ChangeSet` in order
+to influence the way the library runs its diffing algorithm. The
+encoder determines how document tokens (such as nodes and
+characters) are encoded and compared.
+
+Note that both the encoding and the comparison may run a lot, and
+doing non-trivial work in these functions could impact
+performance.
+*/
+interface TokenEncoder<T> {
+    /**
+    Encode a given character, with the given marks applied.
+    */
+    encodeCharacter(char: number, marks: readonly Mark[]): T;
+    /**
+    Encode the start of a node or, if this is a leaf node, the
+    entire node.
+    */
+    encodeNodeStart(node: Node): T;
+    /**
+    Encode the end token for the given node. It is valid to encode
+    every end token in the same way.
+    */
+    encodeNodeEnd(node: Node): T;
+    /**
+    Compare the given tokens. Should return true when they count as
+    equal.
+    */
+    compareTokens(a: T, b: T): boolean;
+}
+
+/**
 Simplifies a set of changes for presentation. This makes the
@@ -113,9 +145,15 @@ assumption that having both insertions and deletions within a word
     Create a changeset with the given base object and configuration.
+    
     The `combine` function is used to compare and combine metadata—it
     should return null when metadata isn't compatible, and a combined
     version for a merged range when it is.
+    
+    When given, a token encoder determines how document tokens are
+    serialized and compared when diffing the content produced by
+    changes. The default is to just compare nodes by name and text
+    by character, ignoring marks and attributes.
     */
-    static create<Data = any>(doc: Node, combine?: (dataA: Data, dataB: Data) => Data): ChangeSet<Data>;
+    static create<Data = any>(doc: Node, combine?: (dataA: Data, dataB: Data) => Data, tokenEncoder?: TokenEncoder<any>): ChangeSet<Data>;
 }
 
-export { Change, ChangeSet, Span, simplifyChanges };
+export { Change, ChangeSet, Span, type TokenEncoder, simplifyChanges };

dist/index.d.ts

@@ -1,2 +0,2 @@
-import { Node } from 'prosemirror-model';
+import { Mark, Node } from 'prosemirror-model';
 import { StepMap } from 'prosemirror-transform';
@@ -44,3 +44,3 @@
     Data associated with the inserted content. Length adds up to
-    `this.toB - this.toA`.
+    `this.toB - this.fromB`.
     */
@@ -57,2 +57,34 @@ readonly inserted: readonly Span<Data>[];
 /**
+A token encoder can be passed when creating a `ChangeSet` in order
+to influence the way the library runs its diffing algorithm. The
+encoder determines how document tokens (such as nodes and
+characters) are encoded and compared.
+
+Note that both the encoding and the comparison may run a lot, and
+doing non-trivial work in these functions could impact
+performance.
+*/
+interface TokenEncoder<T> {
+    /**
+    Encode a given character, with the given marks applied.
+    */
+    encodeCharacter(char: number, marks: readonly Mark[]): T;
+    /**
+    Encode the start of a node or, if this is a leaf node, the
+    entire node.
+    */
+    encodeNodeStart(node: Node): T;
+    /**
+    Encode the end token for the given node. It is valid to encode
+    every end token in the same way.
+    */
+    encodeNodeEnd(node: Node): T;
+    /**
+    Compare the given tokens. Should return true when they count as
+    equal.
+    */
+    compareTokens(a: T, b: T): boolean;
+}
+
+/**
 Simplifies a set of changes for presentation. This makes the
@@ -113,9 +145,15 @@ assumption that having both insertions and deletions within a word
     Create a changeset with the given base object and configuration.
+    
     The `combine` function is used to compare and combine metadata—it
     should return null when metadata isn't compatible, and a combined
     version for a merged range when it is.
+    
+    When given, a token encoder determines how document tokens are
+    serialized and compared when diffing the content produced by
+    changes. The default is to just compare nodes by name and text
+    by character, ignoring marks and attributes.
     */
-    static create<Data = any>(doc: Node, combine?: (dataA: Data, dataB: Data) => Data): ChangeSet<Data>;
+    static create<Data = any>(doc: Node, combine?: (dataA: Data, dataB: Data) => Data, tokenEncoder?: TokenEncoder<any>): ChangeSet<Data>;
 }
 
-export { Change, ChangeSet, Span, simplifyChanges };
+export { Change, ChangeSet, Span, type TokenEncoder, simplifyChanges };

dist/index.js

@@ -0,5 +1,11 @@
+const DefaultEncoder = {
+    encodeCharacter: char => char,
+    encodeNodeStart: node => node.type.name,
+    encodeNodeEnd: () => -1,
+    compareTokens: (a, b) => a === b
+};
 // Convert the given range of a fragment to tokens, where node open
 // tokens are encoded as strings holding the node name, characters as
 // their character code, and node close tokens as -1.
-function tokens(frag, start, end, target) {
+function tokens(frag, encoder, start, end, target) {
     for (let i = 0, off = 0; i < frag.childCount; i++) {
@@ -11,13 +17,13 @@ let child = frag.child(i), endOff = off + child.nodeSize;
                 for (let j = from; j < to; j++)
-                    target.push(child.text.charCodeAt(j - off));
+                    target.push(encoder.encodeCharacter(child.text.charCodeAt(j - off), child.marks));
             }
             else if (child.isLeaf) {
-                target.push(child.type.name);
+                target.push(encoder.encodeNodeStart(child));
             }
             else {
                 if (from == off)
-                    target.push(child.type.name);
-                tokens(child.content, Math.max(off + 1, from) - off - 1, Math.min(endOff - 1, to) - off - 1, target);
+                    target.push(encoder.encodeNodeStart(child));
+                tokens(child.content, encoder, Math.max(off + 1, from) - off - 1, Math.min(endOff - 1, to) - off - 1, target);
                 if (to == endOff)
-                    target.push(-1);
+                    target.push(encoder.encodeNodeEnd(child));
             }
@@ -41,12 +47,13 @@ }
 }
-function computeDiff(fragA, fragB, range) {
-    let tokA = tokens(fragA, range.fromA, range.toA, []);
-    let tokB = tokens(fragB, range.fromB, range.toB, []);
+function computeDiff(fragA, fragB, range, encoder = DefaultEncoder) {
+    let tokA = tokens(fragA, encoder, range.fromA, range.toA, []);
+    let tokB = tokens(fragB, encoder, range.fromB, range.toB, []);
     // Scan from both sides to cheaply eliminate work
     let start = 0, endA = tokA.length, endB = tokB.length;
-    while (start < tokA.length && start < tokB.length && tokA[start] === tokB[start])
+    let cmp = encoder.compareTokens;
+    while (start < tokA.length && start < tokB.length && cmp(tokA[start], tokB[start]))
         start++;
     if (start == tokA.length && start == tokB.length)
         return [];
-    while (endA > start && endB > start && tokA[endA - 1] === tokB[endB - 1])
+    while (endA > start && endB > start && cmp(tokA[endA - 1], tokB[endB - 1]))
         endA--, endB--;
@@ -70,3 +77,3 @@ // If the result is simple _or_ too big to cheaply compute, return
             let x = next < prev ? prev : next + 1, y = x + diag;
-            while (x < lenA && y < lenB && tokA[start + x] === tokB[start + y])
+            while (x < lenA && y < lenB && cmp(tokA[start + x], tokB[start + y]))
                 x++, y++;
@@ -231,3 +238,3 @@ frontier[diag + max] = x;
     Data associated with the inserted content. Length adds up to
-    `this.toB - this.toA`.
+    `this.toB - this.fromB`.
     */
@@ -556,3 +563,3 @@ inserted) {
                 continue;
-            let diff = computeDiff(this.config.doc.content, newDoc.content, change);
+            let diff = computeDiff(this.config.doc.content, newDoc.content, change, this.config.encoder);
             // Fast path: If they are completely different, don't do anything
@@ -630,8 +637,14 @@ if (diff.length == 1 && diff[0].fromB == 0 && diff[0].toB == change.toB - change.fromB)
     Create a changeset with the given base object and configuration.
+    
     The `combine` function is used to compare and combine metadata—it
     should return null when metadata isn't compatible, and a combined
     version for a merged range when it is.
+    
+    When given, a token encoder determines how document tokens are
+    serialized and compared when diffing the content produced by
+    changes. The default is to just compare nodes by name and text
+    by character, ignoring marks and attributes.
     */
-    static create(doc, combine = (a, b) => a === b ? a : null) {
-        return new ChangeSet({ combine, doc }, []);
+    static create(doc, combine = (a, b) => a === b ? a : null, tokenEncoder = DefaultEncoder) {
+        return new ChangeSet({ combine, doc, encoder: tokenEncoder }, []);
     }
@@ -638,0 +651,0 @@ }

package.json

 {
   "name": "prosemirror-changeset",
-  "version": "2.2.1",
+  "version": "2.3.0",
   "description": "Distills a series of editing steps into deleted and added ranges",
@@ -32,3 +32,4 @@ "type": "module",
     "prosemirror-model": "^1.0.0",
-    "prosemirror-test-builder": "^1.0.0"
+    "prosemirror-test-builder": "^1.0.0",
+    "builddocs": "^1.0.8"
   },
@@ -38,4 +39,4 @@ "scripts": {
     "prepare": "pm-buildhelper src/changeset.ts",
-    "build-readme": "FIXME"
+    "build-readme": "builddocs --format markdown --main src/README.md src/changeset.ts  > README.md"
   }
 }

README.md

@@ -44,3 +44,3 @@ # prosemirror-changeset
    Data associated with the inserted content. Length adds up to
-   `this.toB - this.toA`.
+   `this.toB - this.fromB`.
 
@@ -100,4 +100,5 @@ * `static `**`merge`**`<Data>(x: readonly Change[], y: readonly Change[], combine: fn(dataA: Data, dataB: Data) → Data) → readonly Change[]`\
 
- * `static `**`create`**`<Data = any>(doc: Node, combine?: fn(dataA: Data, dataB: Data) → Data = (a, b) => a === b ? a : null as any) → ChangeSet`\
+ * `static `**`create`**`<Data = any>(doc: Node, combine?: fn(dataA: Data, dataB: Data) → Data = (a, b) => a === b ? a : null as any, tokenEncoder?: TokenEncoder = DefaultEncoder) → ChangeSet`\
    Create a changeset with the given base object and configuration.
+
    The `combine` function is used to compare and combine metadata—it
@@ -107,3 +108,8 @@ should return null when metadata isn't compatible, and a combined
 
+   When given, a token encoder determines how document tokens are
+   serialized and compared when diffing the content produced by
+   changes. The default is to just compare nodes by name and text
+   by character, ignoring marks and attributes.
 
+
  * **`simplifyChanges`**`(changes: readonly Change[], doc: Node) → Change[]`\
@@ -117,1 +123,28 @@ Simplifies a set of changes for presentation. This makes the
 
+
+### interface TokenEncoder`<T>`
+
+A token encoder can be passed when creating a `ChangeSet` in order
+to influence the way the library runs its diffing algorithm. The
+encoder determines how document tokens (such as nodes and
+characters) are encoded and compared.
+
+Note that both the encoding and the comparison may run a lot, and
+doing non-trivial work in these functions could impact
+performance.
+
+ * **`encodeCharacter`**`(char: number, marks: readonly Mark[]) → T`\
+   Encode a given character, with the given marks applied.
+
+ * **`encodeNodeStart`**`(node: Node) → T`\
+   Encode the start of a node or, if this is a leaf node, the
+   entire node.
+
+ * **`encodeNodeEnd`**`(node: Node) → T`\
+   Encode the end token for the given node. It is valid to encode
+   every end token in the same way.
+
+ * **`compareTokens`**`(a: T, b: T) → boolean`\
+   Compare the given tokens. Should return true when they count as
+   equal.
+

src/change.ts

@@ -69,3 +69,3 @@ /// Stores metadata for a part of a change.
     /// Data associated with the inserted content. Length adds up to
-    /// `this.toB - this.toA`.
+    /// `this.toB - this.fromB`.
     readonly inserted: readonly Span<Data>[]
@@ -72,0 +72,0 @@ ) {}

src/changeset.ts

 import {Node} from "prosemirror-model"
 import {StepMap} from "prosemirror-transform"
-import {computeDiff} from "./diff"
+import {computeDiff, TokenEncoder, DefaultEncoder} from "./diff"
 import {Change, Span} from "./change"
 export {Change, Span}
 export {simplifyChanges} from "./simplify"
+export {TokenEncoder}
 
@@ -16,3 +17,7 @@ /// A change set tracks the changes to a document from a given point
     /// @internal
-    readonly config: {doc: Node, combine: (dataA: Data, dataB: Data) => Data},
+    readonly config: {
+      doc: Node,
+      combine: (dataA: Data, dataB: Data) => Data,
+      encoder: TokenEncoder<any>
+    },
     /// Replaced regions.
@@ -73,3 +78,3 @@ readonly changes: readonly Change<Data>[]
           !newChanges.some(r => r.toB > change.fromB && r.fromB < change.toB)) continue
-      let diff = computeDiff(this.config.doc.content, newDoc.content, change)
+      let diff = computeDiff(this.config.doc.content, newDoc.content, change, this.config.encoder)
 
@@ -137,7 +142,17 @@ // Fast path: If they are completely different, don't do anything
   /// Create a changeset with the given base object and configuration.
+  ///
   /// The `combine` function is used to compare and combine metadata—it
   /// should return null when metadata isn't compatible, and a combined
   /// version for a merged range when it is.
-  static create<Data = any>(doc: Node, combine: (dataA: Data, dataB: Data) => Data = (a, b) => a === b ? a : null as any) {
-    return new ChangeSet({combine, doc}, [])
+  ///
+  /// When given, a token encoder determines how document tokens are
+  /// serialized and compared when diffing the content produced by
+  /// changes. The default is to just compare nodes by name and text
+  /// by character, ignoring marks and attributes.
+  static create<Data = any>(
+    doc: Node,
+    combine: (dataA: Data, dataB: Data) => Data = (a, b) => a === b ? a : null as any,
+    tokenEncoder: TokenEncoder<any> = DefaultEncoder
+  ) {
+    return new ChangeSet({combine, doc, encoder: tokenEncoder}, [])
   }
@@ -144,0 +159,0 @@

src/diff.ts

@@ -1,8 +0,37 @@
-import {Fragment} from "prosemirror-model"
+import {Fragment, Node, Mark} from "prosemirror-model"
 import {Change} from "./change"
 
+/// A token encoder can be passed when creating a `ChangeSet` in order
+/// to influence the way the library runs its diffing algorithm. The
+/// encoder determines how document tokens (such as nodes and
+/// characters) are encoded and compared.
+///
+/// Note that both the encoding and the comparison may run a lot, and
+/// doing non-trivial work in these functions could impact
+/// performance.
+export interface TokenEncoder<T> {
+  /// Encode a given character, with the given marks applied.
+  encodeCharacter(char: number, marks: readonly Mark[]): T
+  /// Encode the start of a node or, if this is a leaf node, the
+  /// entire node.
+  encodeNodeStart(node: Node): T
+  /// Encode the end token for the given node. It is valid to encode
+  /// every end token in the same way.
+  encodeNodeEnd(node: Node): T
+  /// Compare the given tokens. Should return true when they count as
+  /// equal.
+  compareTokens(a: T, b: T): boolean
+}
+
+export const DefaultEncoder: TokenEncoder<number | string> = {
+  encodeCharacter: char => char,
+  encodeNodeStart: node => node.type.name,
+  encodeNodeEnd: () => -1,
+  compareTokens: (a, b) => a === b
+}
+
 // Convert the given range of a fragment to tokens, where node open
 // tokens are encoded as strings holding the node name, characters as
 // their character code, and node close tokens as -1.
-function tokens(frag: Fragment, start: number, end: number, target: (number | string)[]) {
+function tokens<T>(frag: Fragment, encoder: TokenEncoder<T>, start: number, end: number, target: T[]) {
   for (let i = 0, off = 0; i < frag.childCount; i++) {
@@ -13,9 +42,9 @@ let child = frag.child(i), endOff = off + child.nodeSize
       if (child.isText) {
-        for (let j = from; j < to; j++) target.push(child.text!.charCodeAt(j - off))
+        for (let j = from; j < to; j++) target.push(encoder.encodeCharacter(child.text!.charCodeAt(j - off), child.marks))
       } else if (child.isLeaf) {
-        target.push(child.type.name)
+        target.push(encoder.encodeNodeStart(child))
       } else {
-        if (from == off) target.push(child.type.name)
-        tokens(child.content, Math.max(off + 1, from) - off - 1, Math.min(endOff - 1, to) - off - 1, target)
-        if (to == endOff) target.push(-1)
+        if (from == off) target.push(encoder.encodeNodeStart(child))
+        tokens(child.content, encoder, Math.max(off + 1, from) - off - 1, Math.min(endOff - 1, to) - off - 1, target)
+        if (to == endOff) target.push(encoder.encodeNodeEnd(child))
       }
@@ -42,11 +71,12 @@ }
 
-export function computeDiff(fragA: Fragment, fragB: Fragment, range: Change) {
-  let tokA = tokens(fragA, range.fromA, range.toA, [])
-  let tokB = tokens(fragB, range.fromB, range.toB, [])
+export function computeDiff(fragA: Fragment, fragB: Fragment, range: Change, encoder: TokenEncoder<any> = DefaultEncoder) {
+  let tokA = tokens(fragA, encoder, range.fromA, range.toA, [])
+  let tokB = tokens(fragB, encoder, range.fromB, range.toB, [])
 
   // Scan from both sides to cheaply eliminate work
   let start = 0, endA = tokA.length, endB = tokB.length
-  while (start < tokA.length && start < tokB.length && tokA[start] === tokB[start]) start++
+  let cmp = encoder.compareTokens
+  while (start < tokA.length && start < tokB.length && cmp(tokA[start], tokB[start])) start++
   if (start == tokA.length && start == tokB.length) return []
-  while (endA > start && endB > start && tokA[endA - 1] === tokB[endB - 1]) endA--, endB--
+  while (endA > start && endB > start && cmp(tokA[endA - 1], tokB[endB - 1])) endA--, endB--
   // If the result is simple _or_ too big to cheaply compute, return
@@ -71,3 +101,3 @@ // the remaining region as the diff
       let x = next < prev ? prev : next + 1, y = x + diag
-      while (x < lenA && y < lenB && tokA[start + x] === tokB[start + y]) x++, y++
+      while (x < lenA && y < lenB && cmp(tokA[start + x], tokB[start + y])) x++, y++
       frontier[diag + max] = x
@@ -74,0 +104,0 @@ // Found a match

src/README.md

@@ -28,2 +28,4 @@ # prosemirror-changeset
 
-@simplifyChanges
+@simplifyChanges
+
+@TokenEncoder

No alert changes

Improved metrics

Total package byte prevSize

119668

6.69%

Lines of code

2360

3.78%

Number of lines in readme file

147

28.95%

Worsened metrics

Dev dependency count

4

33.33%

No dependency changes