I am trying to copy a file (let's call it list.txt) that is in several places and give it a new name. Some folders should be ignored (e.g. those containing _old)

Assume the following structure:

/data/folder1/subfolder/user/list.txt
/data/folder2/subfolder/user/list.txt
/data/folder2/subfolder_old/user/list.txt
/data/folder3/anothername/other/list.txt
/data/folder3/anothername/ignorethisfolder/list.txt

With following command I get the exact files listed that I expect:

ll /data/*/{subfolder,anothername}/{user,other}/*.txt 2>/dev/null

That's what I want to use for copying and the following:

cp /data/*/{subfolder,anothername}/{user,other}/*.txt /data/*/{subfolder, anothername}/{user,other}/*.txt_backup 2>/dev/null

Unfortunately, this does not deliver the desired result. What should the command for the copying process be?

CodePudding user response：

Unfortunately the cp command doesn't work that way, you need to copy each file independently:

for f in /data/*/{subfolder,anothername}/{user,other}/*.txt
do
    cp "$f" "${f}_backup"
done

CodePudding user response：

You could use the find command:

find /data -not -path "*/*_old/*" -name "list.txt" -type f -exec cp {} {}_backup \;

Explanation

1. find /data traverse your /data directory
2. -not -path "*/*_old/*" exclude any paths with a directory ending in _old
3. -name "list.txt" select files/directories with the name list.txt
4. -type f only select files
5. -exec cp {} {}_backup \; execute cp on all remaining matching paths the {} is replaced with the match path and the ; is needed to end the exec statement

note You can add your 2>/dev/null if you still get errors on some results.