Support RTL languages in the canvas renderer

src/Terminal.ts

@@ -52,6 +52,7 @@ import { IKeyboardEvent } from './common/Types';
 import { evaluateKeyboardEvent } from './core/input/Keyboard';
 import { KeyboardResultType, ICharset } from './core/Types';
 import { clone } from './common/Clone';
+import { registerRtlCharacterJoiners } from './Unicode';
 
 // Let it work inside Node.js for automated testing purposes.
 const document = (typeof window !== 'undefined') ? window.document : null;
@@ -752,7 +753,6 @@ export class Terminal extends EventEmitter implements ITerminal, IDisposable, II
     // Listen for mouse events and translate
     // them into terminal mouse protocols.
     this.bindMouse();
-
   }
 
   private _setupRenderer(): void {
@@ -762,6 +762,7 @@ export class Terminal extends EventEmitter implements ITerminal, IDisposable, II
       default: throw new Error(`Unrecognized rendererType "${this.options.rendererType}"`);
     }
     this.register(this.renderer);
+    registerRtlCharacterJoiners(this);
   }
 
   /**

src/Unicode.test.ts

@@ -0,0 +1,19 @@
+/**
+ * Copyright (c) 2017 The xterm.js authors. All rights reserved.
+ * @license MIT
+ */
+
+import { assert } from 'chai';
+import { createUnicodeRangeJoiner } from './Unicode';
+
+describe('createUnicodeRangeJoiner', () => {
+  it('should join charcaters within the range', () => {
+    // Join a-z
+    const joiner = createUnicodeRangeJoiner(97, 122);
+    assert.deepEqual(joiner('ABCabc aBc abC Abc'), [
+      [3, 6],   // abc
+      [11, 13], // ab
+      [16, 18]  // bc
+    ]);
+  });
+});

src/Unicode.ts

@@ -0,0 +1,56 @@
+/**
+ * Copyright (c) 2019 The xterm.js authors. All rights reserved.
+ */
+
+import { Terminal } from 'xterm';
+
+export function registerRtlCharacterJoiners(terminal: Terminal): void {
+  // Hebrew
+  terminal.registerCharacterJoiner(createUnicodeRangeJoiner(0x0590, 0x05FF));
+  // Arabic (Kurdish, Persian, Urdu)
+  terminal.registerCharacterJoiner(createUnicodeRangeJoiner(0x0600, 0x06FF));
+  // Arabic Supplement
+  terminal.registerCharacterJoiner(createUnicodeRangeJoiner(0x0750, 0x077F));
+  // Thanana (Dhivehi, Maldavian)
+  terminal.registerCharacterJoiner(createUnicodeRangeJoiner(0x0780, 0x07BF));
+  // Arabic Extended-A
+  terminal.registerCharacterJoiner(createUnicodeRangeJoiner(0x08A0, 0x08FF));
+  // Arabic Presentation Forms-A
+  terminal.registerCharacterJoiner(createUnicodeRangeJoiner(0xFB50, 0xFDFF));
+  // Arabic Presentation Forms-B
+  terminal.registerCharacterJoiner(createUnicodeRangeJoiner(0xFE70, 0xFEFF));
+  // Rumi Numeral Symbols
+  terminal.registerCharacterJoiner(createUnicodeRangeJoiner(0x10E60, 0x10E60));
+  // Arabic Mathematical Alphabetic Symbols
+  terminal.registerCharacterJoiner(createUnicodeRangeJoiner(0x1EE00, 0x1EEFF));
+}
+
+export function createUnicodeRangeJoiner(rangeStart: number, rangeEnd: number): (text: string) => [number, number][] {
+  return (text: string) => {
+    let start = -1;
+    let length = 0;
+    const result: [number, number][] = [];
+    for (let i = 0; i < text.length; i++) {
+      const code = text.charCodeAt(i);
+      if (code >= rangeStart && code <= rangeEnd || code === 32) {
+        // Add to range
+        if (start === -1) {
+          start = i;
+        }
+        length++;
+      } else {
+        if (length > 1) {
+          // Finish range
+          result.push([start, start + length]);
+        }
+        // Clear a single match
+        start = -1;
+        length = 0;
+      }
+    }
+    if (length > 0) {
+      result.push([start, start + length]);
+    }
+    return result;
+  };
+}