feat: Supports unicode diacritical marks when rendering line diff (fixes #169)

# [5.2.0](https://github.com/wickedest/Mergely/compare/v5.1.4...v5.2.0) (2024-06-09)

### Features

* Allows height to be not explicit height, e.g. 'inherit' or '100%' ([#196](https://github.com/wickedest/Mergely/issues/196)) ([b9e3641](b9e3641c85))

CHANGELOG.md

@@ -1,3 +1,10 @@
+# [5.2.0](https://github.com/wickedest/Mergely/compare/v5.1.4...v5.2.0) (2024-06-09)
+
+
+### Features
+
+* Allows height to be not explicit height, e.g. 'inherit' or '100%' ([#196](https://github.com/wickedest/Mergely/issues/196)) ([b9e3641](https://github.com/wickedest/Mergely/commit/b9e3641c852a8926db5efdf33e65a607d5f2df5e))
+
 ## [5.1.4](https://github.com/wickedest/Mergely/compare/v5.1.3...v5.1.4) (2024-05-17)
 
 

package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "mergely",
-  "version": "5.1.4",
+  "version": "5.2.0",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "mergely",
-      "version": "5.1.4",
+      "version": "5.2.0",
       "license": "(GPL-3.0 OR LGPL-3.0 OR MPL-1.1 OR SEE LICENSE IN LICENSE)",
       "devDependencies": {
         "@babel/core": "^7.1.6",

package.json

@@ -1,6 +1,6 @@
 {
   "name": "mergely",
-  "version": "5.1.4",
+  "version": "5.2.0",
   "description": "A javascript UI for diff/merge",
   "license": "(GPL-3.0 OR LGPL-3.0 OR MPL-1.1 OR SEE LICENSE IN LICENSE)",
   "author": {

src/diff-view.js

@@ -65,7 +65,6 @@ CodeMirrorDiffView.prototype.unbind = function() {
 		this.el.removeChild(this.el.lastChild);
 	}
 	if (this._origEl) {
-		this.el.style = this._origEl.style;
 		this.el.className = this._origEl.className;
 	}
 	this._unbound = true;
@@ -257,16 +256,17 @@ CodeMirrorDiffView.prototype.resize = function() {
 
 CodeMirrorDiffView.prototype.bind = function(container) {
 	this.trace('api#bind', container);
-	this._origEl = {
-		style: container.style,
-		className: container.className
-	};
 	const el = dom.getMergelyContainer({ clazz: container.className });
 	const computedStyle = window.getComputedStyle(container);
-	if (!computedStyle.height || computedStyle.height === '0px') {
+	if (!el.style.height
+		&& (!computedStyle.height || computedStyle.height === '0px')
+	) {
 		throw new Error(
 			`The element "${container.id}" requires an explicit height`);
 	}
+	this._origEl = {
+		className: container.className
+	};
 	this.id = `${container.id}`;
 	this.lhsId = `${container.id}-lhs`;
 	this.rhsId = `${container.id}-rhs`;
@@ -753,6 +753,9 @@ CodeMirrorDiffView.prototype._set_top_offset = function (side) {
 	// this is the distance from the top of the screen to the top of the
 	// content of the first codemirror editor
 	const topnode = this._queryElement('.CodeMirror-measure');
+	if (!topnode.offsetParent) {
+		return false;
+	}
 	const top_offset = topnode.offsetParent.offsetTop + 4;
 
 	// restore editor's scroll position

src/diff.js

@@ -244,7 +244,9 @@ function CodeifyText(lhs, rhs, options) {
 
 	if (typeof lhs === 'string') {
 		if (this.options.split === 'chars') {
-			this.lhs = lhs.split('');
+			// split characters and include their diacritical marks
+			this.lhs = lhs.match(/\p{Letter}\p{Mark}*|\p{White_Space}/gu) || [];
+			// this.lhs = [...lhs];
 		} else if (this.options.split === 'words') {
 			this.lhs = lhs.split(/\s/);
 		} else if (this.options.split === 'lines') {
@@ -255,7 +257,9 @@ function CodeifyText(lhs, rhs, options) {
 	}
 	if (typeof rhs === 'string') {
 		if (this.options.split === 'chars') {
-			this.rhs = rhs.split('');
+			// split characters and include their diacritical marks
+			this.rhs = rhs.match(/\p{Letter}\p{Mark}*|\p{White_Space}/gu) || [];
+			// this.rhs = [...rhs];
 		} else if (this.options.split === 'words') {
 			this.rhs = rhs.split(/\s/);
 		} else if (this.options.split === 'lines') {

src/mergely.js

@@ -46,7 +46,9 @@ class Mergely {
 		}
 
 		const computedStyle = window.getComputedStyle(element);
-		if (!computedStyle.height || computedStyle.height === '0px') {
+		if (!element.style.height
+			&& (!computedStyle.height || computedStyle.height === '0px')
+		) {
 			throw new Error(
 				`The element "${selector}" requires an explicit height`);
 		}

src/vdoc.js

@@ -2,6 +2,8 @@ const diff = require('./diff');
 
 const trace = console.log;
 
+const expLetters = new RegExp(/\p{Letter}\p{Mark}*|\p{White_Space}/gu);
+
 class VDoc {
 	constructor(options) {
 		this.options = options;
@@ -275,15 +277,18 @@ class VLine {
 				editor.setGutterMarker(this.id, name, item);
 			}
 			if (this.markup.length) {
+				// while Mergely diffs unicode chars (letters+mark), CM is by character,
+				// so diffs need to be mapped.
+				const mapped = mapLettersToChars(editor.getValue());
 				for (const markup of this.markup) {
 					const [ charFrom, charTo, className ] = markup;
 					const fromPos = { line: this.id };
 					const toPos = { line: this.id };
 					if (charFrom >= 0) {
-						fromPos.ch = charFrom;
+						fromPos.ch = mapped[charFrom];
 					}
 					if (charTo >= 0) {
-						toPos.ch = charTo;
+						toPos.ch = mapped[charTo];
 					}
 					this._clearMarkup.push(
 						editor.markText(fromPos, toPos, { className }));
@@ -334,4 +339,15 @@ function getExtents(side, change) {
 	};
 }
 
+function mapLettersToChars(text) {
+	let match;
+	let mapped = {};
+	let index = 0;
+	expLetters.lastIndex = 0;
+	while ((match = expLetters.exec(text)) !== null) {
+		mapped[index++] = match.index;
+	}
+	return mapped;
+}
+
 module.exports = VDoc;

test/markup.spec.js

@@ -233,7 +233,109 @@ describe('markup', () => {
 				expect(rhs_spans[1].innerText).to.equal('h');
 				expect(rhs_spans[2].innerText).to.equal('ir');
 			}
-		}
+		},
+		{
+			name: 'single word single diacritic non-spacing marks',
+			lhs: 'كلمة',
+			rhs: 'كَلمة',
+			check: (editor) => {
+				expect(editor.querySelectorAll(LHS_CHANGE_START + '.cid-0')).to.have.length(1);
+				expect(editor.querySelectorAll(LHS_CHANGE_END + '.cid-0')).to.have.length(1);
+				expect(editor.querySelectorAll(RHS_CHANGE_START + '.cid-0')).to.have.length(1);
+				expect(editor.querySelectorAll(RHS_CHANGE_END + '.cid-0')).to.have.length(1);
+				const lhs_spans = editor.querySelectorAll(LHS_INLINE_TEXT + '.cid-0');
+				expect(lhs_spans).to.have.length(1);
+				expect(lhs_spans[0].innerText).to.equal('ك');
+				const rhs_spans = editor.querySelectorAll(RHS_INLINE_TEXT + '.cid-0');
+				expect(rhs_spans).to.have.length(1);
+				expect(rhs_spans[0].innerText).to.equal('كَ');
+			}
+		},
+		{
+			name: 'single word multiple diacritic non-spacing marks',
+			lhs: ['\u006E', '\u0061', '\u0314', '\u0065'].join(''), // na̔e
+			rhs: ['\u006E', '\u0061', '\u0314', '\u034A', '\u0065'].join(''), // na̔͊e
+			check: (editor) => {
+				expect(editor.querySelectorAll(LHS_CHANGE_START + '.cid-0')).to.have.length(1);
+				expect(editor.querySelectorAll(LHS_CHANGE_END + '.cid-0')).to.have.length(1);
+				expect(editor.querySelectorAll(RHS_CHANGE_START + '.cid-0')).to.have.length(1);
+				expect(editor.querySelectorAll(RHS_CHANGE_END + '.cid-0')).to.have.length(1);
+				const lhs_spans = editor.querySelectorAll(LHS_INLINE_TEXT + '.cid-0');
+				expect(lhs_spans).to.have.length(1);
+				expect(lhs_spans[0].innerText).to.equal(['\u0061', '\u0314'].join(''));
+				const rhs_spans = editor.querySelectorAll(RHS_INLINE_TEXT + '.cid-0');
+				expect(rhs_spans).to.have.length(1);
+				expect(rhs_spans[0].innerText).to.equal('a̔͊');
+			}
+		},
+		{
+			name: 'multiple words diacritic non-spacing marks',
+			lhs: 'كلمة اخرى',
+			rhs: 'كْلمة اخرى',
+			check: (editor) => {
+				expect(editor.querySelectorAll(LHS_CHANGE_START + '.cid-0')).to.have.length(1);
+				expect(editor.querySelectorAll(LHS_CHANGE_END + '.cid-0')).to.have.length(1);
+				expect(editor.querySelectorAll(RHS_CHANGE_START + '.cid-0')).to.have.length(1);
+				expect(editor.querySelectorAll(RHS_CHANGE_END + '.cid-0')).to.have.length(1);
+				const lhs_spans = editor.querySelectorAll(LHS_INLINE_TEXT + '.cid-0');
+				expect(lhs_spans).to.have.length(1);
+				expect(lhs_spans[0].innerText).to.equal('ك');
+				const rhs_spans = editor.querySelectorAll(RHS_INLINE_TEXT + '.cid-0');
+				expect(rhs_spans).to.have.length(1);
+				expect(rhs_spans[0].innerText).to.equal('كْ');
+			}
+		},
+		{
+			name: 'nonnormalizable diacritic non-spacing marks',
+			lhs: 'naeg',
+			// there are 2 marks on 'e', tilde (0303) and x (0353)
+			rhs: ['\u006E', '\u0061', '\u0353', '\u0065', '\u0353', '\u0303', '\u0067'].join(''),
+			check: (editor) => {
+				expect(editor.querySelectorAll(LHS_CHANGE_START + '.cid-0')).to.have.length(1);
+				expect(editor.querySelectorAll(LHS_CHANGE_END + '.cid-0')).to.have.length(1);
+				expect(editor.querySelectorAll(RHS_CHANGE_START + '.cid-0')).to.have.length(1);
+				expect(editor.querySelectorAll(RHS_CHANGE_END + '.cid-0')).to.have.length(1);
+				const lhs_spans = editor.querySelectorAll(LHS_INLINE_TEXT + '.cid-0');
+				expect(lhs_spans).to.have.length(1);
+				expect(lhs_spans[0].innerText).to.equal('ae');
+				const rhs_spans = editor.querySelectorAll(RHS_INLINE_TEXT + '.cid-0');
+				expect(rhs_spans).to.have.length(1);
+				expect(rhs_spans[0].innerText).to.equal(
+					['\u0061', '\u0353', '\u0065', '\u0353', '\u0303'].join('')
+				);
+			}
+		},
+		{
+			name: 'nonnormalizable diacritic non-spacing marks',
+			lhs: [
+				'\u0065', '\u0353', '\u0303',
+				'\u0065', '\u0353', '\u0303',
+				'\u0065', '\u0353', '\u0303',
+				'x',
+				'\u0065', '\u0353', '\u0303',
+			].join(''),
+			// there are 2 marks on 'e', tilde (0303) and x (0353)
+			rhs: [
+				'\u0065', '\u0353', '\u0303',
+				'\u0065', '\u0353', '\u0303',
+				'\u0065', '\u0353', '\u0303',
+				'y',
+				'\u0065', '\u0353', '\u0303',
+			].join(''),
+			check: (editor) => {
+				expect(editor.querySelectorAll(LHS_CHANGE_START + '.cid-0')).to.have.length(1);
+				expect(editor.querySelectorAll(LHS_CHANGE_END + '.cid-0')).to.have.length(1);
+				expect(editor.querySelectorAll(RHS_CHANGE_START + '.cid-0')).to.have.length(1);
+				expect(editor.querySelectorAll(RHS_CHANGE_END + '.cid-0')).to.have.length(1);
+				const lhs_spans = editor.querySelectorAll(LHS_INLINE_TEXT + '.cid-0');
+				expect(lhs_spans).to.have.length(1);
+				expect(lhs_spans[0].innerText).to.equal('x');
+				const rhs_spans = editor.querySelectorAll(RHS_INLINE_TEXT + '.cid-0');
+				expect(rhs_spans).to.have.length(1);
+				expect(rhs_spans[0].innerText).to.equal('y');
+			}
+		},
+
 	];
 
 	// to debug, add `only: true` to the test `opts` above, and run `npm run debug`

webpack.dev.js

@@ -1,4 +1,5 @@
 const path = require('path')
+const chalk = require('chalk');
 const HtmlWebpackPlugin = require('html-webpack-plugin');
 
 module.exports = {
@@ -50,8 +51,8 @@ module.exports = {
 				compiler.hooks.entryOption.tap('MyPlugin', (context, entry) => {
 					console.log('-'.repeat(78));
 					console.log('Applications:');
-					console.log('http://localhost:8080/app.html');
+					console.log(chalk.bold(chalk.underline(chalk.cyan('http://localhost:8080/app.html'))));
-					console.log('http://localhost:8080/app-styles.html');
+					console.log(chalk.bold(chalk.underline(chalk.cyan('http://localhost:8080/app-styles.html'))));
 					console.log('-'.repeat(78));
 				});
 			}

webpack.prod.js

@@ -15,6 +15,10 @@ module.exports = (mode) => {
 			...webpackDevConfig.output,
 			path: path.join(__dirname, 'lib'),
 			filename: './[name].js',
+			library: {
+				name: 'mergely',
+				type: 'umd',
+			}
 		},
 		optimization: {
 			minimize: true,