[SNOLAB] Speak latest telegram message With TTS technology just in your browser. @deprecated Use my new script [SNOLAB] I Heard Telegram Speaking
// ==UserScript==
// @deprecated Use my new script [SNOLAB] I Heard Telegram Speaking
// @name Telegram Speaker
// @namespace https://userscript.snomiao.com/
// @author snomiao@gmail.com
// @version 0.1.3
// @description [SNOLAB] Speak latest telegram message With TTS technology just in your browser. @deprecated Use my new script [SNOLAB] I Heard Telegram Speaking
// @match https://*.telegram.org/z/
// @grant none
// @run-at document-start
// @license GPL-3.0+
// @supportURL https://github.com/snomiao/userscript.js/issues
// @contributionURL https://snomiao.com/donate
// ==/UserScript==
/*
# the legacy way needs you to install and run an saying pipe service in your computer, in which situation that you don't have latest browser with TTS technologies.
npm i -g piserve snosay
piserve | snosay --voice "Microsoft Huihui Desktop"
*/
async function say(s) {
if (!s) return; // console.error('say empty msg')
// new method to say
console.log("saying " + s);
if (globalThis.speechSynthesis) {
// wait for voices
while (speechSynthesis.getVoices().length === 0) {
await new Promise((r) => setTimeout(r, 1e3));
}
const utter = new SpeechSynthesisUtterance(s);
utter.voice = speechSynthesis
.getVoices()
.filter(({ lang }) => navigator.languages.includes(lang))
.reverse()[0];
utter.rate = Math.min(Math.max(1, s.length / 60), 4);
if (speechSynthesis.speaking) speechSynthesis.cancel();
speechSynthesis.speak(utter);
} else await fetch("http://localhost:25971/?text=" + encodeURIComponent(s));
}
const lastMsg = () =>
[...document.querySelectorAll(".Message:not(.own) .text-content")]
.map((e) => e.textContent)
.reverse()[0];
const chagnedFilterMaker = (init) => (e) => e !== init ? (init = e) : undefined;
const changedFilter = chagnedFilterMaker("");
const looper = () => (say(changedFilter(lastMsg())), 1);
(async function () {
while (looper()) await new Promise((r) => setTimeout(r, 1000));
})();