Find and remove spaces in files and folders, then find patterns in files, then copy the files found

Question

First I will describe the task I have:

• I have several disks
• I have a list of phrases (words) in a patterns.txt file that I use as search patterns
• I must to review the disks in search of files containing these patterns
• Copy found files to a separate location on another disk
• Erase the source disk

I came up with such a solution:

1. Find folders and files by pattern file and save list them to the file:

   grep -ril -f /home/user/patterns.txt /media/user/source-disk | tee /home/user/listfiles.txt

2. Find pdf files (because grep are skipping pdfs) by patterns file and attach them to the file:

   pdfgrep -ril -f /home/user/patterns.txt /media/user/source-disk | tee -a /home/user/listfiles.txt

3. Copy the files found to the external disk2:

   cp -i --preserve=all -v `cat /home/user/listfiles.txt` /home/user/disk2

4. Use ShredOS to wipe source-disk

Now I encountered a problem:

In the listfiles.txt results some folders and files contain spaces = that means that cp reports a mistake (which seems correct)

I found help here and added one more point to work, and now looks like this:

1. Find folders and files containing space and replace it with underscores:

   find /media/user/source-disk -iname "* *" | while read file; do mv "$file" ${file// /_}; done

2. Find folders and files by pattern file and save list them to the file:

   grep -ril -f /home/user/patterns.txt /media/user/source-disk | tee /home/user/listfiles.txt

3. Find pdf files (because grep are skipping pdfs) by patterns file and attach them to the file:

   pdfgrep -ril -f /home/user/patterns.txt /media/user/source-disk | tee -a /home/user/listfiles.txt

4. Copy the files found to the external disk2:

   cp -i --preserve=all -v `cat /home/user/listfiles.txt` /home/user/disk2

5. Use ShredOS to wipe source-disk

My questions:

1. Does it all look correct?
2. Did I forget to do something? (After erasing, it will not be possible to repeat and fix my stupidity)
3. Is it better to use the script from here instead of find -iname "* *" | while...?

Script:

#!/bin/bash # set -o xtrace # uncomment for debugging declare weirdchars=" &\'" function normalise_and_rename() { declare -a list=("${!1}") for fileordir in "${list[@]}"; do newname="${fileordir//[${weirdchars}]/_}" [[ ! -a "$newname" ]] && \ mv "$fileordir" "$newname" || \ echo "Skipping existing file, $newname." done } declare -a dirs files while IFS= read -r -d '' dir; do dirs+=("$dir") done < <(find -type d -print0 | sort -z) normalise_and_rename dirs[@] while IFS= read -r -d '' file; do files+=("$file") done < <(find -type f -print0 | sort -z) normalise_and_rename files[@] 

• advantage: can find other strange characters except spaces
• Disadvantage: I can't make the script work on any selected disk (for example, USB). It works only on my main start disk

4. Can anyone help convert this script so that they work on the disk indicated by me?
5. Can I do all the work in one command without problems using &&?

find /media/user/source-disk -iname "* *" | while read file; do mv "$file" ${file// /_}; done && grep -ril -f /home/user/patterns.txt /media/user/source-disk | tee /home/user/listfiles.txt && pdfgrep -ril -f /home/user/patterns.txt /media/user/source-disk | tee -a /home/user/listfiles.txt && cp -i --preserve=all -v `cat /home/user/listfiles.txt` /home/user/disk2

Answer 1

find -type f -exec grep -iqf ../patterns.txt {} "+" -exec cp --parents -t ../target "{}" "+" 

The switch --parents in the cp command creates the parents directories if needed.

For testing, I just used ../target and ../patterns.txt - you'll have to adjust that for your needs.

To find with -iname "* " is generally unneeded, since " *" does not contain a characters, which upper- or lowerness could be ignored, but maybe this doesn't cost you anything on the other side.

Find already creates an iterator, so no need to use a while-loop (if you happen to use GNU-find, which I suspect (default Linux find?)). But more important: Find knows whether a space is part of a filename, or a separator between such.

I would suggest to test the command first and only wipe or shred after carefully controlling whether it works as intended.

I didn't include a call to pdfgrep, because it isn't installed on my system, but to keep it simple, you may just use 2 variants of the command, with -iname "*.pdf" ... pdfgrep ... and -not -iname "*.pdf" ... grep ....

I tested with standard Linux commands and subdir and files with and without blanks.