feat(synced-lyrics): Added Hindi & Bengali romanization support for lyrics (#3933)

Co-authored-by: pranjol <pranjol@pranjols-iMac-Pro.local>

package.json

@@ -77,6 +77,7 @@
     "@hono/swagger-ui": "0.5.3",
     "@hono/zod-openapi": "1.2.1",
     "@hono/zod-validator": "0.7.6",
+    "@indic-transliteration/sanscript": "1.3.3",
     "@jellybrick/dbus-next": "0.10.3",
     "@jellybrick/electron-better-web-request": "2.0.0",
     "@jellybrick/mpris-service": "2.1.5",

pnpm-lock.yaml

@@ -78,6 +78,9 @@ importers:
       '@hono/zod-validator':
         specifier: 0.7.6
         version: 0.7.6(hono@4.11.7)(zod@4.3.6)
+      '@indic-transliteration/sanscript':
+        specifier: 1.3.3
+        version: 1.3.3
       '@jellybrick/dbus-next':
         specifier: 0.10.3
         version: 0.10.3
@@ -843,6 +846,12 @@ packages:
     resolution: {integrity: sha512-bV0Tgo9K4hfPCek+aMAn81RppFKv2ySDQeMoSZuvTASywNTnVJCArCZE2FWqpvIatKu7VMRLWlR1EazvVhDyhQ==}
     engines: {node: '>=18.18'}
 
+  '@indic-transliteration/common_maps@1.0.5':
+    resolution: {integrity: sha512-XbWDA5AXGE+Nh4uGr/yN9ZM8avRBy4F1KQL+DLgQGOdsQ390lcW4fga0NSjg4C/rOpMd0rHZv2YFV3Bq3UbpkQ==}
+
+  '@indic-transliteration/sanscript@1.3.3':
+    resolution: {integrity: sha512-zNGeARmQTPIlubwgEhl/JumpwTPHrdT/cNsQeCL+G67SQmjJe3qRnMIYghXiVt7+KDso/pU1Ky2ZfD/RBISfJQ==}
+
   '@isaacs/balanced-match@4.0.1':
     resolution: {integrity: sha512-yzMTt9lEb8Gv7zRioUilSglI0c0smZ9k5D65677DLWLtWJaXIS3CqcGyUFByYKlnUj6TkjLVs54fBl6+TiGQDQ==}
     engines: {node: 20 || >=22}
@@ -4338,6 +4347,9 @@ packages:
     resolution: {integrity: sha512-6udB24Q737UD/SDsKAHI9FCRP7Bqc9D/MQUV02ORQg5iskjtLJlZJNdN4kKtcdtwCeWIwIHDGaUsTsCCAa8sFQ==}
     engines: {node: '>=10'}
 
+  toml@2.3.6:
+    resolution: {integrity: sha512-gVweAectJU3ebq//Ferr2JUY4WKSDe5N+z0FvjDncLGyHmIDoxgY/2Ie4qfEIDm4IS7OA6Rmdm7pdEEdMcV/xQ==}
+
   totalist@3.0.1:
     resolution: {integrity: sha512-sf4i37nQ2LBx4m3wB74y+ubopq6W/dIzXg0FDGjsYnZHVa1Da8FH853wlL2gtUhg+xJXjfk3kUZS3BRoQeoQBQ==}
     engines: {node: '>=6'}
@@ -5206,6 +5218,13 @@ snapshots:
 
   '@humanwhocodes/retry@0.4.3': {}
 
+  '@indic-transliteration/common_maps@1.0.5': {}
+
+  '@indic-transliteration/sanscript@1.3.3':
+    dependencies:
+      '@indic-transliteration/common_maps': 1.0.5
+      toml: 2.3.6
+
   '@isaacs/balanced-match@4.0.1': {}
 
   '@isaacs/brace-expansion@5.0.0':
@@ -9085,6 +9104,8 @@ snapshots:
       '@tokenizer/token': 0.3.0
       ieee754: 1.2.1
 
+  toml@2.3.6: {}
+
   totalist@3.0.1: {}
 
   truncate-utf8-bytes@1.0.2:

src/plugins/synced-lyrics/renderer/utils.tsx

@@ -7,6 +7,7 @@ import * as pinyin from 'tiny-pinyin';
 import { romanize as romanizeThaiFrag } from '@dehoist/romanize-thai';
 import { lazy } from 'lazy-var';
 import { detect } from 'tinyld';
+import Sanscript from '@indic-transliteration/sanscript';
 
 import { waitForElement } from '@/utils/wait-for-element';
 import { LyricsRenderer, setIsVisible } from './renderer';
@@ -155,6 +156,12 @@ const hasChinese = (lines: string[]) =>
 const hasThai = (lines: string[]) =>
   lines.some((line) => /[\u0E00-\u0E7F]+/.test(line));
 
+const hasBengali = (lines: string[]) =>
+  lines.some((line) => /[\u0980-\u09FF]+/.test(line));
+
+const hasHindi = (lines: string[]) =>
+  lines.some((line) => /[\u0900-\u097F]+/.test(line));
+
 export const romanizeJapanese = async (line: string) =>
   (await kuroshiro.get()).convert(line, {
     to: 'romaji',
@@ -190,11 +197,35 @@ export const romanizeThai = (line: string) => {
   return latin;
 };
 
+export const romanizeBengali = (line: string) => {
+  try {
+    let out = Sanscript.t(line, 'bengali', 'iast');
+    out = out.normalize('NFD');
+    out = out.replace(/[\u0300-\u036f]/g, '');
+    out = out.replace(/[\u09BC\u09BE-\u09CD]/g, '');
+    return out.toLowerCase();
+  } catch {
+    return line;
+  }
+};
+
+export const romanizeHindi = (line: string) => {
+  try {
+    let out = Sanscript.t(line, 'devanagari', 'iast');
+    out = out.normalize('NFD').replace(/[\u0300-\u036f]/g, ''); // strip accents
+    return out.replace(/[^a-zA-Z\s]/g, '') || line; // remove any remaining symbols
+  } catch {
+    return line;
+  }
+};
+
 const handlers: Record<string, (line: string) => Promise<string> | string> = {
   ja: romanizeJapanese,
   ko: romanizeHangul,
   zh: romanizeChinese,
   th: romanizeThai,
+  bn: romanizeBengali,
+  hi: romanizeHindi,
 };
 
 export const romanize = async (line: string) => {
@@ -210,6 +241,8 @@ export const romanize = async (line: string) => {
   if (hasKorean([line])) line = romanizeHangul(line);
   if (hasChinese([line])) line = romanizeChinese(line);
   if (hasThai([line])) line = romanizeThai(line);
+  if (hasBengali([line])) line = romanizeBengali(line);
+  if (hasHindi([line])) line = romanizeHindi(line);
 
   return line;
 };