Get ony by one the results from find/locate

Question

How can I get 1 by 1 the results of a command like

# locate file.txt
/home/alex/file.txt
/backup/file.txt

What I want to do is to make an interactive script where to ask: Which of the following files would you like to copy/recover? [1],[2]:

Example:

filetorecover.txt
_________________
   file.txt
   file1.txt
   ...

./recover filestorecover.txt /home/alex/data/
file.txt could not be located at /home/alex/data, but it was locate here:
[1] /home/alex/files/
[2] /backup/file.txt

• filestorecover => is the file that contains file to look for
• /home/alex/files/ => is the main folder where to look for and if the file is not located there, then do a locate file.txt

So, mainly I am looking for a way of grabbing those X results from find or locate one by one for the late use as explained above.

UPDATE

Is it possible to see the most common path for find/locate's result? I mean if I have 6 files located at /backup/ and 2 at /home/alex/files/ I would like to set, or ask if they what to get all the files from the most common folder, /backup/!

This is meant when I have more than X files (ex: 10), because sometimes I can get up to hundreds of results and I can't go 1 by 1!

Answer 1

I'd use awk and read.

# list of files in temp file
choices=$(mktemp -t choices.XXXXX)
find . -name "file.txt") > $choices

# choices with number prefix
cat $choices | awk '{print "[" NR "]" $0}'

# prompt
echo "Choose one"
read num

# find choice
choice=$( cat $choices | awk "{if (NR==$num){print \$0}}" )
echo "You choose $choice"

Answer 2

If you are using bash, you can use the built-in arrays like:

# Get all the files
files=`find . -name "file.txt"`
# Put them into an array
declare -a afiles
afiles=( ${files} )

# Output filenames
for (( i = 0; i < ${#afiles[@]}; i += 1 ));
do
    printf "[%d] %s\n" ${i} ${afiles[$i]}
done

(The declare is not neccessary but IMHO it is a good comment saying that afiles is an array.)

Using arrays has the advantage that you can access the filename after the user entered some number with ${afiles[${number}]}.

Answer 3

For your update

#paths in order
find . -name '*.txt' | awk '{gsub(/\/[^/]*$/,"");path[$0]+=1} END {for (i in path) print path[i]" "i}' | sort -rg | cut -d ' ' -f 2

Use what has already been posted for the rest