Change reader server useragent and port

This commit is contained in:
Tanner Collin 2019-08-23 08:21:25 +00:00
parent c0607b3fb6
commit 771c3987ec

View File

@ -1,16 +1,16 @@
const express = require('express');
const app = express();
const port = 3000;
const port = 33843;
const request = require('request');
const JSDOM = require('jsdom').JSDOM;
const createDOMPurify = require('dompurify');
const Readability = require('readability');
app.use(express.urlencoded());
app.use(express.urlencoded({ extended: true }));
app.get('/', (req, res) => {
res.send('POST a URL to urlencoded \'url\' to parse.');
res.send('<form method="POST" accept-charset="UTF-8"><input name="url"><button type="submit">SUBMIT</button></form>');
});
const requestCallback = (url, res) => (error, response, body) => {
@ -24,7 +24,11 @@ const requestCallback = (url, res) => (error, response, body) => {
const reader = new Readability(cleanDoc.window.document);
const article = reader.parse();
res.send(article);
if (article && article.content) {
res.send(article.content);
} else {
res.sendStatus(404);
}
} else {
console.log('Response error:', error ? error.toString() : response.statusCode);
res.sendStatus(response ? response.statusCode : 404);
@ -35,7 +39,8 @@ app.post('/', (req, res) => {
const url = req.body.url;
const requestOptions = {
url: url,
headers: {'User-Agent': 'Googlebot/2.1 (+http://www.google.com/bot.html)'},
//headers: {'User-Agent': 'Googlebot/2.1 (+http://www.google.com/bot.html)'},
headers: {'User-Agent': 'Twitterbot/1.0'},
};
console.log('Parse request for:', url);