Why `wc -c` always count 1 more character?

One way is to tr to delete newlines, then you can count the characters.

Standard behavior:

echo HELLO | wc -m
# result: 6
echo -n HELLO | wc -m
# result: 5

To show the count of newline characters found:

echo HELLO | wc -l
# result: 1
echo -n HELLO | wc -l
# result: 0

Strip the newline character and count characters:

echo HELLO | tr -d '\n' | wc -m
# result: 5

Strip the newline character (and possible returns with \r) and count characters for an input file:

tr -d '\n\r' < input.txt | wc -m

Because newlines are characters too. Tell your text editor to not add one at the end of the file. No, I don't know how.

I've been using a similar suggestion to the-wabbit's for my calculations.

as a workaround, you could count newlines with wc -l and substract them from the count of wc -c.

function num_chars () {
    # echo -e tells echo to honor certain sequences like \n
    chars=$(echo -e "${1}" | wc -c)
    lines=$(echo -e "${1}" | wc -l)
    real_chars=$(echo "$chars - $lines" | bc)
    echo "$real_chars"
}

num_chars "hello Dolly"
11 #Result
num_chars "hello
dolly"
11 #Result
num_chars "hello \nDolly"
11 #Result