Replacing the duplicate uuids across multiple files - bash

Question

I am trying to replace the duplicate UUIDs from multiple files in a directory. Even the same file can have duplicate UUIDs.
I am using Unix utilities to solve this.
Till now I have used grep, cut, sort and uniq to find all the duplicate UUIDs across the folder and store it in a file (say duplicate_uuids)
Then I tried sed to replace the UUIDs by looping through the file.
filename="$1"
re="*.java"
while read line; do
uuid=$(uuidgen)
sed -i'.original' -e "s/$line/$uuid/g" *.java
done < "$filename"
As you would expect, I ended up replacing all the duplicate UUIDs with new UUID but still, it is duplicated throughout the file!
Is there any sed trick that can work for me?

Answer 1

There are a bunch of ways this can likely be done. Taking a multi-command approach using a function might give you greater flexibility if you want to customize things later, for example:
#!/bin/bash
checkdupes() {
files="$*"
for f in $files; do
filename="$f"
printf "Searching File: %s\n" "${filename}"
while read -r line; do
arr=( $(grep -n "${line}" "${filename}" | awk 'BEGIN { FS = ":" } ; {print $1" "}') )
for i in "${arr[#]:1}"; do
sed -i '' ''"${i}"'s/'"${line}"'/'"$(uuidgen)"'/g' "${filename}"
printf "Replaced UUID [%s] at line %s, first found on line %s\n" "${line}" "${i}" "${arr[0]}"
done
done< <( sort "${filename}" | uniq -d )
done
}
checkdupes /path/to/*.java
So what this series of commands does is to first sort the duplicates (if any) in whatever file you choose. It takes those duplicates and uses grep and awk to create an array of line numbers which each duplicate is found. Looping through the array (while skipping the first value) will allow the duplicates to be replaced by a new UUID and then re-saving the file.
Using a duplicate list file:
If you want to use a file with a list of dupes to search other files and replace the UUID in each of them that match it's just a matter of changing two lines:
Replace:
for i in "${arr[#]:1}"; do
With:
for i in "${arr[#]}"; do
Replace:
done< <( sort "${filename}" | uniq -d )
With:
done< <( cat /path/to/dupes_list )
NOTE: If you don't want to overwrite the file, then remove sed -i '' at the beginning of the command.

Answer 2

This worked for me:
#!/bin/bash
duplicate_uuid=$1
# store file names in array
find . -name "*.java" > file_names
IFS=$'\n' read -d '' -r -a file_list < file_names
# store file duplicate uuids from file to array
IFS=$'\n' read -d '' -r -a dup_uuids < $duplicate_uuid
# loop through all files
for file in "${file_list[#]}"
do
echo "$file"
# Loop through all repeated uuids
for old_uuid in "${dup_uuids[#]}"
do
START=1
# Get the number of times uuid present in this file
END=$(grep -c $old_uuid $file)
if (( $END > 0 )) ; then
echo " Replacing $old_uuid"
fi
# Loop through them one by one and change the uuid
for (( c=$START; c<=$END; c++ ))
do
uuid=$(uuidgen)
echo " [$c of $END] with $uuid"
sed -i '.original' -e "1,/$old_uuid/s/$old_uuid/$uuid/" $file
done
done
rm $file.original
done
rm file_names