Scribd Enhancer with OCR, export, print, parallel scraping, and clean UI. Auto language detection, smart filtering, output splitting — full feature set restored and refined. By Eliminater74.
Verzia zo dňa
// ==UserScript==
// @name Scribd Enhancer All-in-One (v2.7.3)
// @namespace https://greasyfork.org/users/Eliminater74
// @version 2.7.3
// @description Scribd Enhancer with OCR, export, print, parallel scraping, and clean UI. Auto language detection, smart filtering, output splitting — full feature set restored and refined. By Eliminater74.
// @author Eliminater74
// @license MIT
// @match *://*.scribd.com/*
// @grant none
// @icon https://s-f.scribdassets.com/favicon.ico
// ==/UserScript==
(function () {
'use strict';
const SETTINGS_KEY = 'scribdEnhancerSettings';
const defaultSettings = {
unblur: true,
autoScrape: false,
darkMode: false,
showPreview: true,
enableOCR: true,
ocrLang: 'auto',
splitEvery: 0
};
const settings = { ...defaultSettings, ...JSON.parse(localStorage.getItem(SETTINGS_KEY) || '{}') };
const saveSettings = () => localStorage.setItem(SETTINGS_KEY, JSON.stringify(settings));
const tesseractScript = document.createElement('script');
tesseractScript.src = 'https://cdn.jsdelivr.net/npm/[email protected]/dist/tesseract.min.js';
document.head.appendChild(tesseractScript);
const style = document.createElement('style');
style.textContent = `
#se-ui {
position: fixed; bottom: 20px; right: 20px; background: #222; color: #fff;
border-radius: 10px; padding: 10px; z-index: 9999; width: 320px;
font-family: sans-serif; font-size: 13px; box-shadow: 0 0 10px #000;
}
#se-ui label, #se-ui select, #se-ui button {
display: block; width: 100%; margin: 4px 0;
}
#se-ui input[type="checkbox"] { margin-right: 6px; }
#se-ui button {
background: #444; color: white; border: none; border-radius: 6px; padding: 6px;
}
#se-preview {
position: fixed; top: 10px; right: 20px; bottom: 140px; width: 360px;
background: #f4f4f4; color: #000; overflow: auto; padding: 10px;
font-family: monospace; font-size: 12px; white-space: pre-wrap;
border: 1px solid #999; z-index: 9998; border-radius: 10px;
}
.dark-mode #se-preview {
background: #222; color: #eee; border-color: #555;
}
.dark-mode * {
background-color: transparent !important;
color: #e0e0e0 !important;
border-color: #444 !important;
}
`;
document.head.appendChild(style);
function applyDarkMode() {
document.documentElement.classList.toggle('dark-mode', settings.darkMode);
document.body.classList.toggle('dark-mode', settings.darkMode);
}
function unblurContent() {
if (!settings.unblur) return;
const cleanup = () => {
document.querySelectorAll('.blurred_page, .promo_div, [unselectable="on"]').forEach(el => el.remove());
document.querySelectorAll('*').forEach(el => {
const cs = getComputedStyle(el);
if (cs.color === 'transparent') el.style.color = '#111';
if (cs.textShadow?.includes('white')) el.style.textShadow = 'none';
});
};
cleanup();
new MutationObserver(cleanup).observe(document.body, { childList: true, subtree: true });
}
function cleanOCRText(text) {
return text.split('\n').map(t => t.trim()).filter(line =>
line.length >= 3 && /[a-zA-Z]/.test(line) && !/^[^a-zA-Z0-9]{3,}$/.test(line)
).join('\n');
}
function detectLanguage(text) {
const map = { spa: /ñ|á|í|ó|ú/, fra: /é|è|ê|ç/, deu: /ä|ö|ü|ß/, ron: /ș|ț|ă|î|â/ };
for (const [lang, regex] of Object.entries(map)) {
if (regex.test(text)) return lang;
}
return 'eng';
}
async function preprocessImage(src) {
return new Promise(resolve => {
const img = new Image();
img.crossOrigin = 'anonymous';
img.onload = () => {
if (img.naturalWidth < 100 || img.naturalHeight < 100 || /logo|icon|watermark/i.test(src)) return resolve(null);
const canvas = document.createElement('canvas');
canvas.width = img.width; canvas.height = img.height;
const ctx = canvas.getContext('2d');
ctx.drawImage(img, 0, 0);
const imageData = ctx.getImageData(0, 0, canvas.width, canvas.height);
for (let i = 0; i < imageData.data.length; i += 4) {
const avg = (imageData.data[i] + imageData.data[i+1] + imageData.data[i+2]) / 3;
imageData.data[i] = imageData.data[i+1] = imageData.data[i+2] = avg;
}
ctx.putImageData(imageData, 0, 0);
resolve(canvas.toDataURL('image/png'));
};
img.src = src;
});
}
function createPreview() {
const preview = document.createElement('div');
preview.id = 'se-preview';
if (settings.showPreview) {
preview.textContent = '[Preview Initialized]\n';
document.body.appendChild(preview);
}
return preview;
}
function exportOutput(content, ext) {
const split = settings.splitEvery;
const lines = content.split(/(?=\[Page \d+])/);
if (!split || split < 1) {
const blob = new Blob([content], { type: `text/${ext}` });
const a = document.createElement('a');
a.href = URL.createObjectURL(blob);
a.download = `scribd_output.${ext}`;
a.click();
} else {
for (let i = 0; i < lines.length; i += split) {
const chunk = lines.slice(i, i + split).join('\n');
const blob = new Blob([chunk], { type: `text/${ext}` });
const a = document.createElement('a');
a.href = URL.createObjectURL(blob);
a.download = `scribd_part${Math.floor(i / split) + 1}.${ext}`;
a.click();
}
}
}
function printToPDF(content) {
const win = window.open('', 'PrintView');
win.document.write(`<html><head><title>Scribd Print</title></head><body><pre>${content}</pre></body></html>`);
win.document.close();
win.focus();
setTimeout(() => win.print(), 600);
}
async function scrapePages(pages, preview) {
const concurrency = 4;
let index = 0;
const firstText = [];
async function scrape(page, i) {
page.scrollIntoView();
await new Promise(r => setTimeout(r, 300));
let found = false;
const text = page.innerText.trim();
if (text) {
preview.textContent += `[Page ${i + 1}] ✅\n${text}\n\n`;
firstText.push(text);
found = true;
}
if (settings.enableOCR && window.Tesseract) {
const imgs = page.querySelectorAll('img');
for (let img of imgs) {
const src = img.src || '';
const processed = await preprocessImage(src);
if (!processed) continue;
const lang = settings.ocrLang === 'auto' ? detectLanguage(firstText.join(' ')) : settings.ocrLang;
const result = await window.Tesseract.recognize(processed, lang);
const ocrText = cleanOCRText(result.data.text || '');
if (ocrText) {
preview.textContent += `[OCR] ${ocrText}\n\n`;
found = true;
}
}
}
if (!found) preview.textContent += `[Page ${i + 1}] ❌ No content\n\n`;
}
const tasks = Array(concurrency).fill(null).map(async () => {
while (index < pages.length) {
const i = index++;
await scrape(pages[i], i);
}
});
await Promise.all(tasks);
alert(`✅ Scraped ${pages.length} pages.`);
}
function createUI(preview) {
const ui = document.createElement('div');
ui.id = 'se-ui';
ui.innerHTML = `
<label><input type="checkbox" id="opt-unblur"> Unblur</label>
<label><input type="checkbox" id="opt-autoscrape"> Auto Scrape</label>
<label><input type="checkbox" id="opt-dark"> Dark Mode</label>
<label><input type="checkbox" id="opt-preview"> Show Preview</label>
<label>OCR:
<select id="opt-lang">
<option value="auto">Auto</option>
<option value="eng">English</option>
<option value="spa">Spanish</option>
<option value="fra">French</option>
<option value="deu">German</option>
</select>
</label>
<label>Split Every:
<select id="opt-split">
<option value="0">Off</option>
<option value="100">100</option>
<option value="250">250</option>
<option value="500">500</option>
</select>
</label>
<button id="btn-scrape">📖 Scrape Pages</button>
<button id="btn-export">💾 Export TXT</button>
<button id="btn-html">🧾 Export HTML</button>
<button id="btn-print">🖨️ Print to PDF</button>
`;
document.body.appendChild(ui);
ui.querySelector('#opt-unblur').checked = settings.unblur;
ui.querySelector('#opt-autoscrape').checked = settings.autoScrape;
ui.querySelector('#opt-dark').checked = settings.darkMode;
ui.querySelector('#opt-preview').checked = settings.showPreview;
ui.querySelector('#opt-lang').value = settings.ocrLang;
ui.querySelector('#opt-split').value = settings.splitEvery;
ui.querySelectorAll('input, select').forEach(input => {
input.onchange = () => {
settings.unblur = ui.querySelector('#opt-unblur').checked;
settings.autoScrape = ui.querySelector('#opt-autoscrape').checked;
settings.darkMode = ui.querySelector('#opt-dark').checked;
settings.showPreview = ui.querySelector('#opt-preview').checked;
settings.ocrLang = ui.querySelector('#opt-lang').value;
settings.splitEvery = parseInt(ui.querySelector('#opt-split').value);
saveSettings();
applyDarkMode();
};
});
ui.querySelector('#btn-scrape').onclick = () => {
const pages = [...document.querySelectorAll(
'.page, .reader_column, [id^="page_container"], .outer_page, .abs_page, .scribd_page, .text_layer'
)];
if (!pages.length) return alert('❌ No pages found.');
scrapePages(pages, preview);
};
ui.querySelector('#btn-export').onclick = () => exportOutput(preview.textContent, 'txt');
ui.querySelector('#btn-html').onclick = () => exportOutput(`<html><body><pre>${preview.textContent}</pre></body></html>`, 'html');
ui.querySelector('#btn-print').onclick = () => printToPDF(preview.textContent);
}
window.addEventListener('load', () => {
applyDarkMode();
unblurContent();
const preview = createPreview();
createUI(preview);
if (settings.autoScrape) document.querySelector('#btn-scrape').click();
});
})();