New unicode-graphemes addon.

Handle grapheme cluster lookup as well as wcwidth.
This is based on Unicode 15, and could replace the unicode11 addon.

src/common/services/Services.ts

@@ -294,6 +294,11 @@ export interface IOscLinkService {
   getLinkData(linkId: number): IOscLinkData | undefined;
 }
 
+/** Width and Grapheme_Cluster_Break properties of a character. */
+export type UnicodeCharProperties = number;
+export const UnicodeInitialProperties: UnicodeCharProperties = 0; // UNEEDED?
+export type UnicodeCharWidth = 0 | 1 | 2;
+
 export const IUnicodeService = createDecorator<IUnicodeService>('UnicodeService');
 export interface IUnicodeService {
   serviceBrand: undefined;
@@ -309,13 +314,18 @@ export interface IUnicodeService {
   /**
    * Unicode version dependent
    */
-  wcwidth(codepoint: number): number;
+  wcwidth(codepoint: number): UnicodeCharWidth;
   getStringCellWidth(s: string): number;
+  /** Return character width, character type (for grapheme clustering).
+   * If preceding!=0, it is return code from previous character;
+   * in that case result specifies if characters should be joined. */
+  charProperties(codepoint: number, preceding: UnicodeCharProperties): UnicodeCharProperties;
 }
 
 export interface IUnicodeVersionProvider {
   readonly version: string;
-  wcwidth(ucs: number): 0 | 1 | 2;
+  wcwidth(ucs: number): UnicodeCharWidth;
+  charProperties(codepoint: number, preceding: UnicodeCharProperties): UnicodeCharProperties;
 }
 
 export const IDecorationService = createDecorator<IDecorationService>('DecorationService');

src/common/services/UnicodeService.test.ts

@@ -12,6 +12,9 @@ class DummyProvider implements IUnicodeVersionProvider {
   public wcwidth(n: number): 0 | 1 | 2 {
     return 2;
   }
+  charProperties(codepoint: number): number {
+    return UnicodeService.createPropertyValue(0, this.wcwidth(codepoint));
+  }
 }
 
 describe('unicode provider', () => {

src/common/services/UnicodeService.ts

@@ -2,7 +2,7 @@
  * Copyright (c) 2019 The xterm.js authors. All rights reserved.
  * @license MIT
  */
-import { IUnicodeService, IUnicodeVersionProvider } from 'common/services/Services';
+import { IUnicodeService, IUnicodeVersionProvider, UnicodeCharProperties, UnicodeCharWidth  } from 'common/services/Services';
 import { EventEmitter, IEvent } from 'common/EventEmitter';
 import { UnicodeV6 } from 'common/input/UnicodeV6';
 
@@ -16,6 +16,19 @@ export class UnicodeService implements IUnicodeService {
   private readonly _onChange = new EventEmitter<string>();
   public readonly onChange = this._onChange.event;
 
+  public static extractShouldJoin(value: UnicodeCharProperties): boolean {
+    return (value & 1) !== 0;
+  }
+  public static extractWidth(value: UnicodeCharProperties): UnicodeCharWidth {
+    return ((value >> 1) & 0x3) as UnicodeCharWidth;
+  }
+  public static extractCharKind(value: UnicodeCharProperties): number {
+    return value >> 3;
+  }
+  public static createPropertyValue(state: number, width: number, shouldJoin: boolean = false): UnicodeCharProperties {
+    return ((state & 0xffffff) << 3) | ((width & 3) << 1) | (shouldJoin?1:0);
+  }
+
   constructor() {
     const defaultProvider = new UnicodeV6();
     this.register(defaultProvider);
@@ -51,7 +64,7 @@ export class UnicodeService implements IUnicodeService {
   /**
    * Unicode version dependent interface.
    */
-  public wcwidth(num: number): number {
+  public wcwidth(num: number): UnicodeCharWidth {
     return this._activeProvider.wcwidth(num);
   }
 
@@ -83,4 +96,8 @@ export class UnicodeService implements IUnicodeService {
     }
     return result;
   }
+
+  charProperties(codepoint: number, preceding: UnicodeCharProperties): UnicodeCharProperties {
+    return this._activeProvider.charProperties(codepoint, preceding);
+  }
 }