tampermonkey_export_lark_transcript.js

// ==UserScript==
// @name         Export Meeting Transcripts
// @namespace    https://example.com
// @version      1.0
// @description  Extract transcript info (speaker/time/text) from a web page
// @match        https://*.larksuite.com/*
// @grant        none
// ==/UserScript==

(function() {
    'use strict';

    // --- Configuration ---
    // Adjust these selectors if the page structure changes:
    const PARAGRAPH_WRAPPER_SELECTOR = '.subtitle-item-wrapper .paragraph-editor-wrapper';
    const SPEAKER_SELECTOR = '.p-user-name';
    const TIME_SELECTOR    = '.p-time';
    const TEXT_SELECTOR    = '.mm-paragraph-content .ace-line';

    // Create and insert an "Export" button
    const exportButton = document.createElement('button');
    exportButton.textContent = 'Export Transcripts';
    exportButton.style.position = 'fixed';
    exportButton.style.top = '10px';
    exportButton.style.right = '10px';
    exportButton.style.zIndex = '99999';
    exportButton.style.padding = '6px 10px';
    exportButton.style.backgroundColor = '#007bff';
    exportButton.style.color = '#fff';
    exportButton.style.border = 'none';
    exportButton.style.borderRadius = '4px';
    exportButton.style.cursor = 'pointer';
    document.body.appendChild(exportButton);

    // Download helper
    function downloadFile(filename, text) {
        const element = document.createElement('a');
        element.setAttribute('href', 'data:text/plain;charset=utf-8,' + encodeURIComponent(text));
        element.setAttribute('download', filename);
        element.style.display = 'none';
        document.body.appendChild(element);
        element.click();
        document.body.removeChild(element);
    }

    // The main export function
    function exportTranscripts() {
        const paragraphWrappers = document.querySelectorAll(PARAGRAPH_WRAPPER_SELECTOR);

        const results = [];
        paragraphWrappers.forEach(wrapper => {
            // Extract speaker
            const speakerEl = wrapper.querySelector(SPEAKER_SELECTOR);
            const speaker = speakerEl ? speakerEl.getAttribute('user-name-content') || speakerEl.textContent.trim() : '';

            // Extract time
            const timeEl = wrapper.querySelector(TIME_SELECTOR);
            const time = timeEl ? timeEl.getAttribute('time-content') || timeEl.textContent.trim() : '';

            // Extract raw text from the lines
            // Because each transcript can have multiple segments, we gather them all
            let textContent = '';
            const textEls = wrapper.querySelectorAll(TEXT_SELECTOR);
            textEls.forEach(tEl => {
                textContent += tEl.innerText.trim() + ' ';
            });
            textContent = textContent.trim();

            // Store the result
            results.push({
                speaker,
                time,
                text: textContent
            });
        });

        // Convert to JSON (or CSV, if you prefer)
        const jsonString = JSON.stringify(results, null, 2);

        // Download the file
        downloadFile('transcripts.json', jsonString);
    }

    // Attach the click handler
    exportButton.addEventListener('click', exportTranscripts);

})();