feat: add validator for Lao (lo) (#651)

server/lib/validation/index.js

@@ -8,6 +8,7 @@ const ig = require('./languages/ig');
 const it = require('./languages/it');
 const kab = require( './languages/kab');
 const ko = require( './languages/ko');
+const lo = require( './languages/lo');
 const ne = require('./languages/ne');
 const or = require('./languages/or');
 const ru = require('./languages/ru');
@@ -27,6 +28,7 @@ const VALIDATORS = {
   it,
   kab,
   ko,
+  lo,
   ne,
   or,
   ru,

server/lib/validation/languages/lo.js

@@ -0,0 +1,34 @@
+// Lao rules
+// use any rule from Thai rules https://github.com/common-voice/sentence-collector/blob/main/server/lib/validation/languages/th.js
+const MIN_LENGTH = 2;
+const MAX_LENGTH = 140;
+
+const INVALIDATIONS = [{
+  fn: (sentence) => {
+    return sentence.length < MIN_LENGTH || sentence.length > MAX_LENGTH;
+  },
+  error: `ຈຳນວນຕົວອັກສອນຕ້ອງຢູ່ລະຫວ່າງ ${MIN_LENGTH} ຫາ ${MAX_LENGTH} (ລວມ)`,
+}, {
+  // Lao digits and Thai digits
+  regex: /[0-9໑໒໓໔໕໖໗໘໙໐๐-๙]/,
+  error: 'ປະໂຫຍກບໍ່ຄວນມີຕົວເລກ',
+}, {
+  // English and Thai characters are not allowed
+  regex: /[A-Za-z\u0E00-\u0E7F]/,
+  error: 'ປະໂຫຍກບໍ່ຄວນມີຕົວອັກສອນລາຕິນ ຫຼືຕົວອັກສອນໄທ',
+}, {
+  // < > + * \ # @ ^ [ ] ( ) /
+  // ellipsis: \u0EAF ຯ
+  // repetition: \u0EC6 ໆ
+  regex: /[<>+*\\#@^[\]()/\u0EAF\u0EC6]/,
+  error: 'ປະໂຫຍກບໍ່ຄວນມີສັນຍາລັກ, ລວມທັງ ຯ ແລະ ໆ',
+}, {
+  // Emoji range from https://www.regextester.com/106421 and
+  // https://stackoverflow.com/questions/10992921/how-to-remove-emoji-code-using-javascript
+  regex: /(\u00a9|\u00ae|[\u2000-\u3300]|[\u2580-\u27bf]|\ud83c[\ud000-\udfff]|\ud83d[\ud000-\udfff]|\ud83e[\ud000-\udfff]|[\ue000-\uf8ff])/,
+  error: 'ປະໂຫຍກບໍ່ຄວນມີ ອີໂມຈິ ຫຼືສັນຍາລັກຂອງ Unicode ພິເສດອື່ນໆ',
+}];
+
+module.exports = {
+  INVALIDATIONS,
+};