用bash组合多组文本文件

Question

I have a folder of text files that are labelled something like this:

0filename1
1filename1
2filename1
....
0filename2
1filename2
2filename2
....

et cetera. What I want to do is take all the files that end in filename1 and combine them all into a file named filename1, and similarly for filename2 and all other files. Normally I would do something like this

cat [0123456789]*filename1 > filename1

and just repeat the command for every different file name I have. However, I want to be able to automate this. The exact form of the file names change regularly, so it's not as simple as just writing a script that will do the above command for filename1, filename2, etc. The length of the file names do stay constant though, so I suspect the right way to automate this would be for a script to take every file that has the same last n characters in the filename and copy them into a file with the name of these n characters. I'm not sure how to do this though - any suggestions?

Answer 1

Sounds pretty simple, just need to filter the files to get the 'base' strings.

for $base in $( ls | cut -b 1-8 | sort -u ); do
    cat [0123456789]*$base > $base
done

where 1-8 is the characters you intend to keep , so <consistent length of filenames> - <N last characters that vary instead of 8 .

Answer 2

Bit more complex solution that handles files with whitespace characters, with multi-digit numbers or flexible filename length:

#!/usr/bin/env bash

shopt -s extglob nullglob

files=(+([0-9])?*)
(( ${#files[@]} )) || exit 1

while IFS= read -rd '' filename; do
    printf '%s\0' +([0-9])"$filename" | sort -zn | xargs -0 cat > "$filename"
done < <(printf '%s\0' "${files[@]##+([0-9])}" | sort -zu)

Answer 3

#!/bin/bash

str="filename"
for i in {1..2}
do
    cat {?,??}"${str}${i}" > "${str}${i}"
done

Script uses Bash Expansion {..} and character wildcard ? to expand the available filenames. If you have 0filename1 to 9filename1 , then use a single ? and use ?? for 10filename1-99filename1 .

Example:

$ cat 0filename1 
011
$ cat 1filename1 
111
$ cat 2filename1 
211
$ cat 10filename1 
1011
$ cat 0filename2
022
$ cat 1filename2
122
$ cat 2filename2
222
$ cat 10filename2
1022

Output of the above script would be:

$ cat filename1
011
111
211
1011

$ cat filename2
022
122
222
1022