Ubuntu – Bash Script: Batch file rename to remove/change all characters illegal under a windows filesystem

bashbatch-renamecommand linescripts

I need an assist here, got a project going where we need to modify file & folder names within a specific directory in order to remove or replace any and all characters which would be considered illegal under Windows, with a few additional characters.
Leave final extnsion

ILLEGAL CHARACTERS:

    < (less than)                         MAKE -
    > (greater than)                      MAKE -
    : (colon)                             MAKE -
    " (double quote)                      REMOVE
    / (forward slash)                     MAKE -
    \ (backslash)                         MAKE -
    | (vertical bar or pipe)              MAKE -
    ? (question mark)                     REMOVE
    * (asterisk)                          REMOVE


ADDITIONAL CHARACTERS:

    REMOVE '
    REMOVE %
    REMOVE #
    REMOVE ^
    REMOVE !@#$%^&()=+][{};,`~

Any character with "MAKE -" next to it needs to be made a – (hyphen) in place of the character, for example a file named John\Sykes.doc needs to become John-Sykes.doc.
If the character is flagged as "REMOVE" then it should be removed, and nothing be put in it's place, for example a file named John?Sykes.doc would become JohnSykes.doc

This is a rather important step in a very large project we have going in our organisation, and I haven't been able to create a script thus far. I know how to make simple versions, that can only change file names on a case by case basis but can't do it en mass. For example, I can create a script to remove only @ from file names, but I don't know how to include folders/directories nor do I know how to include all the other characters in this one script.

Ideally the script would run the first time and produce a file using echo with all the changes that need to be made including the full path and then the changed path. The script would then either need to be modified slightly to use mv instead, or it could refer to the file created using echo and perform all the necessary changes using it…

Any and all help would be appreciated as this is really rather important and I lack the skills at this point in time to make this myself…

A final note, I can't use the #rename command. It breaks our IT policy.

EDIT

I'm unable to use any version of rename, the head of my IT dept. has forbade it. I'm sure he has his reasons, but I can't say that I'm privy to them. The rename command is not available on our servers at all.

This means I have to use pure bash to perform the operation.

There aren't any additional requirements. Basically I need to be able to run a script to output all the changes that will be made to a file, before anything is changed. This will allow us to review the results and make sure there are no stuff-ups. After that, we want to run (the same script or a different one) it to actually make the changes.

All I need is stated above, so anything marked "MAKE -" needs to become a hyphen and anything marked "REMOVE" needs to just go away.

If there's anything else, please ask. I'll happily give further comment.

Best Answer

Create a script rename1.sh:

#! /bin/bash
old=$1
new=${old//[\''"?*%#^!@$&()=+[]{};,`~']}
new=${new//['<>:\|']/-}                   # I removed /, see below.
if [[ -e $new ]] ; then
    if [[ $old != $new ]] ; then
        echo Cannot rename "$old" to "$new" - target already exists.
    fi
else
    mv "$old" "$new"
fi

It renames one file according to the given rules using parameter expansion. To rename a subtree, run it with find:

find /path -exec rename1.sh {} \;

Are you sure you want to replace / with -? It would put all the files into the same directory.

Related Solutions

Ubuntu – How to batch rename files (images) based on CSV file

This should work for you:

sed 's/"//g' files.csv | while IFS=, read orig new; do mv "$orig" "$new"; done

Explanation:

sed 's/"//g' files.csv : remove the quotes
IFS=, : split the input on ,
while read orig new; do ... done : This will read each input line, split it on the value of $IFS (here a comma) and save the 1st field as $orig and the rest as $new.
mv "$orig" "$new" : this will rename the files as requested.

If your file only contains file names (like orig.jpg) and no paths (not /home/take2/orig.jpg or similar), the command above will only affect files in your current directory. So, you need to open a terminal, cd to the target directory and run it there.

Test first:

To test this, you can do a dry run first by printing the commands that will be run without actually executing them:

sed 's/"//g' files.csv | while IFS=, read orig new; do echo mv "$orig" "$new"; done

Ubuntu – Mass renaming files with special format

I wrote a small bash script to do the job.

#!/bin/bash

# Variables
extension='mobi'
report='report.log'    

if [ -f $report ];
then
    rm -rf $report
fi

echo $'renaming files . . .\n'

for filename in *.$extension
do
    temp=$(echo $filename | tr '_' ' ' | tr '%20' ' ' | tr '[]' '()' | tr -s ' ' | sed 's/\.[^.]*$//' )

    part1=$(echo $temp | cut -f1 -d-)
    part2=$(echo $temp | cut -f2 -d-)
    new_filename=$(echo "${part2#?} - ${part1%?}.$extension")
    echo $(mv -v "$filename" "$new_filename") | tee -a $report
    if [ $? -ne 0 ]; then
        echo $'\n\nScript FAILED'
        exit 1
    fi
done

echo $'\n\nScript SUCCESSFUL'
exit 0

Create a .bsh file inside the directory in which your .mobi files are stored and paste the above code :

Open a Terminal with Ctrl + Alt + T and navigate to the directory in which your script is stored :

cd /path/to/directory/

To change the permissions of the script :

chmod +x <filename>.bsh

To execute the script :

bash <filename>.bsh

If for some reason the mv fails you will receive an error 'Script FAILED', otherwise you'll get 'Script SUCCESSFUL'.

Explanation

tr '_' ' ' replace underscores with whitespace

tr '%20' ' ' replace %20 with whitespace

tr '[]' '()' replace square brackets with parentheses

tr -s ' ' replace multiple spaces with one

sed 's/\.[^.]*$//' extracts only the name of the file without the extension

${string#?} remove first character of a string

${string%?} remove last character of a string

mv -v this will show in the terminal which files are being moved. -v stands for verbose.

tee -a changelog.log By default tee command reads from standard input, and writes to standard output and files. -a stands for append.

if [ $? -ne 0 ] checks if the previous command was successful.

report.log stores the previous and latest name of each file.

Note

The script assumes that the filename contains only one dash -

Best Answer

Related Solutions

Ubuntu – How to batch rename files (images) based on CSV file

Ubuntu – Mass renaming files with special format

Explanation

Related Question