CODEWITHSUNDEEP
linuxbashshellawksed
Marco Falzone
Marco Falzone

Reputation: 139

Linux batch copy files into directories based on filename pattern

I have a list of almost 500 pdf files with the following filename structure:

XXXX-YYYY-MM-DD.pdf

where XXXX is a variable lenght numeric code (1 to 4 digits) always delimitated by "-", for example:

51-2016-08-22.pdf
776-2016-08-22.pdf
3881-2016-08-22.pdf
4-2016-08-22.pdf
2860-2016-08-22.pdf

The goal is to copy each file into its own directory, naming the directories like the pattern (ie: file 776-2016-08-22.pdf goes to directory 776). How can I use awk or sed to delimitate the variable lenght field?

Here's my code:

for f in *.pdf

    do
        FOLDERNAME=`echo $f| awk (awk or sed missing code here)`
        mkdir /my/dir/structure/$FOLDERNAME
        cp $f /my/dir/structure/$FOLDERNAME/
    done

Thanks for your support.

Upvotes: 2

Views: 1476

Answers (2)

acsrujan
acsrujan

Reputation: 1049

As rightly pointed out by ed-morton, This is NOT recommended solution as it fails in many cases. Please follow https://stackoverflow.com/a/39089589/3834860

Keeping this answer for reference.

awk -F '-' to specify delimiter and '{print $1}' for first element before delimiter.

for f in *.pdf

do
    FOLDERNAME=`echo $f| awk -F '-' '{print $1}'`
    mkdir /my/dir/structure/$FOLDERNAME
    cp $f /my/dir/structure/$FOLDERNAME/
done

Upvotes: 1

anubhava
anubhava

Reputation: 785611

You can use:

for f in *.pdf; do
    d="${f%%-*}"
    mkdir -p "$d" && cp "$f" "$d"
done

Upvotes: 5

Related Questions