Shell脚本从目录复制并粘贴随机文件

Question

Title pretty much says it i'm looking to add a line to a script i'm working on that would copy a random file from a directory say ~/Desktop/old and paste it into another folder say ~/Desktop/new. I only want to move one file to the new folder each time the script is run i googled around and only found solutions to echo a random file but couldn't figure out how to copy a random one thank you for any help with this problem

Answer 1

You should not parse the output of 'ls': http://mywiki.wooledge.org/ParsingLs

terse version:

files=(src/*)
mv "${files[$RANDOM % ${#files[@]}]}" dest/

This code will move a random file found within a 'src/' subdirectory to a dest/ subdirectory.

files=(src/*)                    #creates an array of all the files within src/ */
filecount="${#files[@]}"         #determines the length of the array
randomid=$((RANDOM % filecount)) #uses $RANDOM to choose a random number between 0 and $filecount
filetomove="${files[$randomid]}" #the random file wich we'll move
mv "$filetomove" dest/           #does the actual moving

Answer 2

Well if you can echo it just pass the result to cp using xargs. If you could provide the code to generate the random filename it would be helpful.

Answer 3

This demo script shows how you can select a random file from a directory, and should be a good start.

#!/bin/bash

# Set up test data.

rm -rf tmpdata ; mkdir tmpdata
touch tmpdata/fileA tmpdata/fileB tmpdata/fileC tmpdata/fileD tmpdata/fileE

# From and To directories

fromdir=./tmpdata
todir=./tmpdata2

# Get a list of the files to a temporary file.

ls -1 ${fromdir} >/tmp/filelist.$$

# Select a number from 1 to n where n is the line count of that file.
# Then use head and tail to get the line.

filenum=$(expr $RANDOM % $(cat /tmp/filelist.$$ | wc -l) + 1)
file=$(head -${filenum} /tmp/filelist.$$ | tail -1)

# DEBUG stuff.

cat /tmp/filelist.$$ | sed 's/^/DEBUG file: /'
echo "DEBUG nmbr: ${filenum}"

echo "'cp ${fromdir}/${file} ${todir}'"

# Remove temporary file.

rm -f /tmp/filelist.$$

And some sample output:

pax$ ./cprnd.sh
DEBUG file: fileA
DEBUG file: fileB
DEBUG file: fileC
DEBUG file: fileD
DEBUG file: fileE
DEBUG nmbr: 3
'cp ./tmpdata/fileC ./tmpdata2'

pax$ ./cprnd.sh
DEBUG file: fileA
DEBUG file: fileB
DEBUG file: fileC
DEBUG file: fileD
DEBUG file: fileE
DEBUG nmbr: 1
'cp ./tmpdata/fileA ./tmpdata2'

pax$ ./cprnd.sh
DEBUG file: fileA
DEBUG file: fileB
DEBUG file: fileC
DEBUG file: fileD
DEBUG file: fileE
DEBUG nmbr: 5
'cp ./tmpdata/fileE ./tmpdata2'

The "magic" lies in these two lines:

filenum=$(expr $RANDOM % $(cat /tmp/filelist.$$ | wc -l) + 1)
file=$(head -${filenum} /tmp/filelist.$$ | tail -1)

The first uses wc to get the line count (number of files). It then gives you the remainder when dividing a random number by this value so that you end up with 0..n-1 and, by adding 1, you get 1..n . Let's assume it gives you 10 for a fifty-line file.

The next line uses head to get the first ten lines, then pipes that through tail to get the last line of that set (ie, the tenth line from the file).

Answer 4

Ruby (1.9+)

require 'fileutils'
files=[]
Dir["*"].each { |file| test(?f,file) && files << file }
FileUtils.cp(files[ rand(files.size) ] , File.join("/tmp")  )

Answer 5

这些答案大多数都可以在我的计算机上从终端运行时使用，但是它们都不在android终端上运行，所以我最终用Java编写了它，谢谢所有帮助