Added archive support! (Beta) + UI change

This commit is contained in:
Daniel
2022-05-26 20:45:07 +04:30
parent f3f36a9b96
commit d010e351b5
13 changed files with 1168 additions and 1936 deletions
+35
View File
@@ -0,0 +1,35 @@
const puppeteer = require('puppeteer');
const { PuppeteerBlocker } = require('@cliqz/adblocker-puppeteer');
const fetch = require('cross-fetch');
const config = require('../../src/config.json');
const fs = require('fs');
const screenshotDirectory = config.api.storage_location + '/Webmarker/screenshot\'s/';
const pdfDirectory = config.api.storage_location + '/Webmarker/pdf\'s/';
if (!fs.existsSync(screenshotDirectory)){
fs.mkdirSync(screenshotDirectory, { recursive: true });
}
if (!fs.existsSync(pdfDirectory)){
fs.mkdirSync(pdfDirectory, { recursive: true });
}
module.exports = async (link, id) => {
const browser = await puppeteer.launch();
const page = await browser.newPage();
await PuppeteerBlocker.fromPrebuiltAdsAndTracking(fetch).then((blocker) => {
blocker.enableBlockingInPage(page);
});
await page.goto(link, { waitUntil: 'load', timeout: 0 });
const title = await page.title();
await page.screenshot({ path: screenshotDirectory + id + '.png', fullPage: true});
await page.pdf({ path: pdfDirectory + id + '.pdf', format: 'a4' });
await browser.close();
return title;
}
+989 -1822
View File
File diff suppressed because it is too large Load Diff
+4 -3
View File
@@ -9,12 +9,13 @@
"author": "",
"license": "MIT",
"dependencies": {
"cheerio": "^1.0.0-rc.10",
"@cliqz/adblocker-puppeteer": "^1.23.8",
"cors": "^2.8.5",
"cross-fetch": "^3.1.5",
"express": "^4.17.3",
"mongodb": "^4.5.0",
"phantom": "^6.3.0",
"url-parse": "^1.5.10"
"puppeteer": "^14.1.1",
"uuid": "^8.3.2"
},
"devDependencies": {
"nodemon": "^2.0.15"
+23 -33
View File
@@ -1,58 +1,48 @@
const express = require('express');
const app = express();
const { MongoClient, ObjectId } = require('mongodb');
const request = require('request');
const cheerio = require('cheerio');
const URL = require('url-parse');
const { MongoClient } = require('mongodb');
const cors = require('cors');
const config = require('../src/config.json')
const config = require('../src/config.json');
const getData = require('./modules/getData.js')
const port = config.server.port;
const port = config.api.port;
const uri = config.server.mongodb_full_address;
const database = config.server.database_name;
const collection = config.server.collection_name;
const URI = config.api.mongodb_URI;
const database = config.api.database_name;
const collection = config.api.collection_name;
const client = new MongoClient(uri);
const client = new MongoClient(URI);
app.use(cors());
app.use(express.json());
app.get('/get', async (req, res) => {
app.get('/api', async (req, res) => {
const data = await getDoc();
res.send(data);
});
app.post('/post', (req, res) => {
let title;
app.post('/api', async (req, res) => {
const pageToVisit = req.body.link;
request(pageToVisit, (error, response, body) => {
try {
if(response.statusCode === 200) {
// Parse the document body
const $ = cheerio.load(body);
req.body.title = $('title').text();
insertDoc(req.body);
} else {
req.body.title = null;
insertDoc(req.body);
}
} catch (error) {
console.log(error);
req.body.title = null;
insertDoc(req.body);
}
});
try {
const dataResult = await getData(pageToVisit, req.body._id);
req.body.title = dataResult;
insertDoc(req.body);
} catch (err) {
console.log(err);
insertDoc(req.body);
} finally {
res.send('DONE!');
}
});
app.delete('/delete', async (req, res) => {
app.delete('/api', async (req, res) => {
const id = req.body.id.toString();
await deleteDoc(id);
res.send(`Bookmark with ObjectId "${id}" deleted.`);
res.send(`Bookmark with _id:${id} deleted.`);
});
async function insertDoc(doc) {
@@ -85,7 +75,7 @@ async function deleteDoc(doc) {
try {
const db = client.db(database);
const list = db.collection(collection);
const result = await list.deleteOne({"_id": ObjectId(doc)});
const result = await list.deleteOne({"_id": doc});
return result;
}