如何在Bash中使用子字符串对字符串数组进行排序

Question

I am writing my first unix bash script and I find it really difficult because I've never used linux before. It is homework from university. I tried a lot of stuff and I've been looking for hours but I simply can't find the right solution. Can someone give me some directions of what to use to sort strings from array which look like this:

room_100056.dat 4
room_8973.dat 2
room_7764.dat 1
room_2092.dat 20

to this:

room_7764.dat 1
room_8973.dat 2
room_10056.dat 4
room_2092.dat 20 

When sorting the initial array, I am creating a new array which contains the sorted data. The problem is that in the sorted array, the data is printed like this:

room_100056.dat 
room_8973.dat
room_7764.dat 
room_2092.dat
1
2
4
20

Code:

file=$1/*.dat

count_visits=()

for room in $file
do
   visits=$(grep "<Visit>" $room | wc -l)
   filename=$(basename $room)
   count_visits+=($filename$'\t'$visits)
done

sorted_visits=($(echo ${count_visits[@]} | tr " " "\n" | sort -g))
printf '%s\n' "${sorted_visits[@]}"
exit

Answer 1

告诉sort在数字上使用第二列：

printf '%s\n' "${count_visits[@]}" | sort -k2n