Consider the following script, which attempts to read a fixed number of bytes from a file containing Unicode using bash’s read function, and then counts how many bytes and characters are read.

MWE.sh

#!/bin/bash

char_count=$1

while true; do
  read -r -n $char_count output
  echo -n "$output" | wc -c
  echo -n "$output" | wc -m
  echo "${#output}"
  echo
  exit
done < MWE.txt

MWE.txt

main · square

Suppose I invoke it like this:

for x in $(seq 1 10 ); do ./MWE.sh $x; done

This produces the following output:

1
1
1

2
2
2

3
3
3

4
4
4

4
4
4

7
6
6

7
6
6

9
8
8

10
9
9

11
10
10

If bash is attempting to read a particular number of characters, then I would expect discontinuities in the byte count, but a consistently increasing character count.

However, we observe discontinuities in both byte and character count.
For example, when we attempt to read 5 characters, we obtain only 4 characters.

Why does this happen, and is there a way to make read -n <count> behave in a more consistent way?

I am running the following bash version:

bash --version
GNU bash, version 5.0.17(1)-release (x86_64-pc-linux-gnu)