你可以使用“rev”来反转元素，然后用前两个元素排序：

#!/bin/bash

filenames="filename1_20230517_065806.txt filename2_20230517_055806.txt filename_3_20230517_125806.txt file_name_4_20230517_063806.txt"

# Convert the filenames string into an array
IFS=' ' read -r -a files_array <<< "$filenames"

# Sort the array based on the timestamp with reversed elements
sorted_array=($(for file in "${files_array[@]}"; do
  timestamp=$(echo "$file" | rev | cut -d_ -f1,2 | rev)
  echo "$timestamp $file"
done | sort | cut -d' ' -f2))

for file in "${sorted_array[@]}"; do
  echo "$file"
done

输出看起来与预期相同：

filename2_20230517_055806.txt
file_name_4_20230517_063806.txt
filename1_20230517_065806.txt
filename_3_20230517_125806.txt

我认为您需要awk的帮助，使用临时增加数据的“技巧”，通过在每行添加所需的排序键。我还在下面的排序键中添加了日期（$（NF-1））。
for i in $(ls *.txt | awk -F_ ‘{print $(NF-1) $NF “\t” $0}’ | sort | cut -f 2-) do...

$ find . -type f -name "*.txt"|awk -F_ '{a[$NF]=$0} END{n=asorti(a,b); for(i=1; i<=n; i++) print a[b[i]]}'
./filename2_20230517_055806.txt
./file_name_4_20230517_063806.txt
./filename1_20230517_065806.txt
./filename_3_20230517_125806.txt