telegraf-middleware-console-time

Debug your Telegram middlewares

When testing your bot it's nice to know what's coming in and how long it did take to process this request.

This library is not meant for usage in production. It's meant for debugging purposes.

In production the individual request wouldn't be as interesting as the general kind of request. For example not taking a look onto a specific '/start' command but onto all /start messages. This is kind of the opposite of what this library tries to achieve: Helping with understanding individual requests.

Install

1npm install telegraf-middleware-console-time

Usage

Test your implementation

1import {generateUpdateMiddleware} from 'telegraf-middleware-console-time';
2
3const bot = new Bot(…);
4
5// Other middlewares unrelated to your code
6bot.use(…);
7
8// Check your implementation for each telegram update (only when not in production)
9if (process.env['NODE_ENV'] !== 'production') {
10    bot.use(generateUpdateMiddleware());
11}
12
13// Your implementation
14bot.command('start', …);

This will output something like this:

12020-03-31T14:32:36.974Z 490af message Edgar 6 /start: 926.247ms
22020-03-31T14:32:57.750Z 490ag message Edgar 6 /start: 914.764ms
32020-03-31T14:33:01.188Z 490ah message Edgar 5 /stop: 302.666ms
42020-03-31T14:46:11.385Z 490ai message Edgar 6 /start: 892.452ms

The 490af is the update_id.

The number before the commands is the total length of the content. This is helpful when considering max length for stuff like callback data.

The content itself is shortened in order to prevent log spamming.

Test a middleware

When you create your own middleware or assume slow timings of another middleware you can use these middlewares to create a timing profile

1import {generateBeforeMiddleware, generateAfterMiddleware} from 'telegraf-middleware-console-time';
2
3const bot = new Bot(…);
4
5// Use BeforeMiddleware before loading the to be tested middleware
6bot.use(generateBeforeMiddleware('foo'))
7
8// Middleware to be tested
9bot.use(…)
10
11// Use AfterMiddleware after loading the to be tested middleware (with the same label)
12bot.use(generateAfterMiddleware('foo'))
13
14// Other middlewares / implementations (they will take the 'inner' amount of time when used)
15bot.use(…)
16bot.on(…, …)

This will output something like this:

1490ai foo before: 304.185ms
2490ai foo inner: 83.122ms
3490ai foo after: 501.028ms
4490ai foo total: 891.849ms

This indicates the checked middleware alone took 800ms and isnt as performant as maybe needed.