0

I have a txt file with around 5000 lines in it, I would like to print a random line from that file. I have tried using readline but cant figure out how to choose which line to print using integers. What would be the fastest way to do this without using much memory?

Here is my code

const readline = require("readline");
const fs = require("fs");

const file = readline.createInterface({
    input: fs.createReadStream('file.txt'),
    output: process.stdout,
    terminal: false
});

file.on('line', (line) => {
    console.log(line);
})
L4z
  • 11
  • 1
  • Can you please clarify, does "without using much memory" mean that it's unacceptable to read the entire contents of the file into memory? – Wyck May 03 '22 at 13:49

2 Answers2

1

The straightforward solution would be:

  1. read file
  2. split by \n, save to array
  3. get number of lines (== length of array)
  4. generate random number between 0 and number of lines
  5. return array[generated_number]

But it might use quite a bit of memory if your file is really large.

so without reading the entire file, off the top of my head - Maybe seek to a random position, then read until you hit \n, and then again read until \n, then split by \n and pick the middle part of the array?

Note- this will break if the last line isn't followed by a new line. (if the random position is in the last two lines)

Yarin_007
  • 1,449
  • 1
  • 10
  • 17
-1

two approaches here:

  1. read all file, choose random number between 1 and 5000 and return that line (not good for general solution)
  2. pre-define random line and return it when getting to that line.
const readline = require("readline");
const fs = require("fs");

const file = readline.createInterface({
    input: fs.createReadStream('file.txt'),
    output: process.stdout,
    terminal: false
});
const radnomLineNumber = getRandomInt(5000);
let i = 0;
file.on('line', (line) => {
   if (randomLineNumber === i) {
      console.log(line);
      return;
   }
   i++;
});
benjamin
  • 263
  • 1
  • 7
  • 1
    Of course *around 5000 lines* is not necessarily *exactly 5000 lines*, so it may print nothing, or may never return lines beyond the 5000th. So your approach should stipulate, _"If the number of lines in the text file is known ahead of time..."_, rather than saying the rather vague, _"not good for general solution"_. – Wyck May 03 '22 at 13:48