Lobashov_Ivan_PIBD-21_IP/Lab2/node_modules/steno
2024-01-10 22:39:31 +04:00
..
.npmignore lab 2-5 2024-01-10 22:39:31 +04:00
.travis.yml lab 2-5 2024-01-10 22:39:31 +04:00
index.js lab 2-5 2024-01-10 22:39:31 +04:00
LICENSE lab 2-5 2024-01-10 22:39:31 +04:00
package.json lab 2-5 2024-01-10 22:39:31 +04:00
README.md lab 2-5 2024-01-10 22:39:31 +04:00

steno

Simple file writer with atomic writing and race condition prevention.

Can be used as a drop-in replacement to fs.writeFile().

Built on graceful-fs and used in lowdb.

Install

npm install steno --save

Usage

const steno = require('steno')

steno.writeFile('file.json', data, err => {
  if (err) throw err
})

The problem it solves

Without steno

Let's say you have a server and want to save data to disk:

var data = { counter: 0 }

server.post('/', (req, res) => {
  // Increment counter
  ++data.counter

  // Save data asynchronously
  fs.writeFile('data.json', JSON.stringify(data), err => {
    if (err) throw err
    res.end()
  })
})

Now if you have many requests, for example 1000, there's a risk that you end up with:

// In your server
data.counter === 1000

// In data.json
data.counter === 865 // ... or any other value

Why? Because, fs.write doesn't guarantee that the call order will be kept. Also, if the server is killed while data.json is being written, the file can get corrupted.

With steno

server.post('/increment', (req, res) => {
  ++data.counter

  steno.writeFile('data.json', JSON.stringify(data), err => {
    if (err) throw err
    res.end()
  })
})

With steno you'll always have the same data in your server and file. And in case of a crash, file integrity will be preserved.

if needed, you can also use steno.writeFileSync() which offers atomic writing too.

Important: works only in a single instance of Node.

License

MIT - Typicode