0

I have a list of files stored in txt file.

./test7
./test4
./test1
./test5
./test6
./test10
./test8
./test2
./test9
./test3

I want to run a command on all those files but I want to sleep 1 second after each two files are processed, eg:

cp ./test1 test1-backup
cp ./test2 test2-backup
sleep 1
cp ./test3 test3-backup
cp ./test4 test4-backup
sleep 1
...
cp ./test9 test9-backup
cp ./test10 test10-backup

Is there a way to achieve this by bash script? I would like to parametrize the amount of commands executed by 1 iteration (calling sleep 1). Also another problem is that real list of files has hundreds thousands of lines.

Improve this question
2
  • 1
    I think you are better off using tar and gzip to archive and backup these files. Commented Apr 13, 2015 at 16:02
  • If you are able to execute commands remotely, can you not tar/zip those files/directories and back them up? as glenn mentioned, this means you'd have to wait days until the script could possibly complete which is not a good idea especially over a remote connection. If you do decide to go that route, you might want to consider doing it in batches. Commented Apr 13, 2015 at 16:12

2 Answers 2

Reset to default
3

I assume you're not describing your real-world scenario: not only do you want to copy hundreds of thousands of files, but you want to sleep for hundreds of thousands of seconds... wtf?!?

Anyway:

while IFS= read -r file1
      IFS= read -r file2
do
    cp "$file1" "${file1##*/}-backup"
    cp "$file2" "${file2##*/}-backup"
    sleep 1
done < inputFile
Improve this answer
6
  • It is a real world scenario. Iam performing commands on remote filesystem which is very fragile to those operations, eg. copying or removing. Thanks for your answer. Commented Apr 13, 2015 at 16:02
  • @glenn jackman, you just beat me to the answer ;) Commented Apr 13, 2015 at 16:11
  • 2
    +1 but i'd also recommend to do this directly on the remote system using its crontab (if any) or another kind of robust session (screen?). not through a ssh session (or, if through ssh, via a "screen" session, if available on the remote host). ie, anything that will survive a disconnection Commented Apr 13, 2015 at 16:16
  • Cool! +1 One question: why note "$file-backup"? Commented Apr 13, 2015 at 17:41
  • According to the question, the destination file should not have a directory in the filename. Commented Apr 13, 2015 at 20:09
1

Sorry: I have not noticed the "txt file" with the "hundreds thousands of lines". This is just a naif solution...

for a in test*; do ls -l $a; if [[ $((i++ % 2)) != 0 ]]; then sleep 1; fi; done

Update: Explanation and (partially update to txt file with filenames).

...reformating:

for a in `cat file.txt`
do 
   cp "$a" "$a-backup"              ## REPLACE THIS LINE 
   if [[ $((i++ % 2)) != 0 ]] then
       sleep 1
   fi
done
  • this one-liner means; for a in files process the file $a and sleep if it is a even line.

In order to see if the line is even, we are counting it (i++) and seeing if i % 2 is 0

  • in bash $(( ...exp )) calculates ...exp as numerical expression
  • this way $((i++ % 2)) != 0 will be true for even iterations (and in this case we sleep 1)
  • in bash [[ ...exp ]] calculates ...exp as a boolean expression

Note: this solution worked with a file.txt containing 100_000 files but it will fail for very big file.txt

Improve this answer
2
  • Hi! This answer would probably be much more helpful if you explained how it is meant to work, and what placeholders are supposed to be replaced. Commented Apr 13, 2015 at 15:46
  • Do not use a for-loop to iterate over the lines of a file: as soon as you encounter a filename with a space in it, $a will not contain the whole filename (unless you alter $IFS) -- further reading: mywiki.wooledge.org/… Commented Apr 13, 2015 at 20:13

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.