share
Unix & Linuxsorting out files to different directories for a main directory
[0] [3] stormctr2
[2020-04-17 12:01:57]
[ bash command-line find ]
[ https://unix.stackexchange.com/questions/580674/sorting-out-files-to-different-directories-for-a-main-directory ]

I have numerous files I am trying to move from a main directory in to separate directories which have multiple sub-directories inside.

Example file naming conventions 1SA_1BA_1W.out, 2SA_3BA_3W.out, 4SA_2BA_5W.out.... Is there a way given the number in from of each SA, BA, W or other items I have to move all at one time to a given location. The ranges of SA would be 1-10, BA 1-4 ect... Currently I am using a script command such as........

find . -type f -name '1A_1HNO3_W.out' | xargs -I {} cp '{}' /location/A_HNO3
find . -type f -name '2SA_BA_W.out' | xargs -I {} cp '{}' /location/SA_BA_W 
find . -type f -name '3SA_BA_W.out' | xargs -I {} cp '{}' /location/SA_BA_W 
find . -type f -name '4SA_BA_W.out' | xargs -I {} cp '{}' /location/SA_BA_W 
find . -type f -name '1SA_DEA_W.out' | xargs -I {} cp '{}' /location/SA_DEA_W 
find . -type f -name '2SA_DEA_W.out' | xargs -I {} cp '{}' /location/SA_DEA_W 
find . -type f -name '3SA_DEA_W.out' | xargs -I {} cp '{}' /location/SA_DEA_W 
find . -type f -name '4SA_DEA_*W.out' | xargs -I {} cp '{}' /location/SA_DEA_W

have over 1700 files with various identification abbreviations, such as SA BA A HNO3 W MA EAM TMA DMA DEA ect... I need to be able to get pull these files and transfer them into an easy database structure. The "location" would be the name of the main parent directory and folders like SA_DEA_W are sub directories inside the parent directory.

[0] [2020-04-17 15:01:56] Ashot Terteryan 
mv *SA_BA_W.out  /location/SA_BA_W;

mv *SA_DEA_W.out /location/SA_DEA_W

and so on.


1
[0] [2020-04-17 17:19:55] Stalin Vignesh Kumar

You can create directories one time using below command:

Create sub-directories 

ls *.out | sed -Ee 's/[0-9]*([A-Za-z]+)_([0-9]*[A-Za-z]+[0-9]*)_W.out$/\1_\2_W/g' | awk ' { system("mkdir -p /location/" $0) }'

OR

find . -regextype awk -regex './([0-9]*[A-Za-z])+_([0-9]*[A-Za-z]+[0-9]*)_W.out' | sed -Ee 's/[0-9]*([A-Za-z]+)_([0-9]*[A-Za-z]+[0-9]*)_W.out$/\1_\2_W/g' | awk '{ system("mkdir -p /location/" $0) }'

OR

find . -regextype awk -regex './([0-9]*[A-Za-z])+_([0-9]*[A-Za-z]+[0-9]*)_W.out' | sed -Ee 's/[0-9]*([A-Za-z]+)_([0-9]*[A-Za-z]+[0-9]*)_W.out$/\1_\2_W/g' | xargs -I {} mkdir -p /location/{}

YOU CAN DO BOTH operations using below for loop

To Create sub directories and Move the files to the respective directories :

for i in $(ls *.out)
do
DIR_NAME=$(echo $i | sed -Ee 's/[0-9]*([A-Za-z]+)_([0-9]*[A-Za-z]+[0-9]*)_W.out$/\1_\2_W/g')
test -d /location/$DIR_NAME || mkdir /location/$DIR_NAME
mv $i $DIR_NAME
done

2
[0] [2020-04-17 19:56:36] bu5hman

You could try

find . -iname "*.out" -exec sh -c '
     f=${1##*/};
     f=${f%.out};
     d=./location/$f/;
     [ ! -d $f ] && mkdir -p $d;
     mv $1 $d'  sh {} \;

Provided you have no spaces in your file paths

Walkthrough

Find all of your .out files, and for each one -exec .... \;

find . -iname "*.out" -exec sh -c '...script...` sh {} \;

The construct

sh -c `...script......` sh {}

Takes each file name as {} and passes it back into the script as $1 where we

Remove the directory path to the last / and store is as f

         f=${1##*/};

Remove the .out suffix

         f=${f%.out};

Calculate the directory you want and store it as d

         d=./location/$f/;

Create the directory if it doesn't exist, together with any parents -p

         [ ! -d $d ] && mkdir -p $d;

Do the move

         mv $1 $d'  sh {} \;

3