Working script, Would like someone to look at for me

Shell Programming and Scripting
1

First off I have to say, it is because of this forum and countless Google searches I have this script working at all. So thank you for that.

I have created a web based open source project and I am trying to automate some of the tasks when I am releasing a new version. I do most of my development on my Windows box and have a test site running on a local Linux test server.

I have created a script and if I can get a few optimization pointers on it I would appreciate it very much.

This script does exactly what I want it to and I want to know, is there a better way?

#!/bin/bash

CONVERT_FROM=ISO-8859-1
CONVERT_TO=UTF-8

# move into our working directory
cd /var/www/localhost/htdocs/dist/

# remove execute bit
find . -type f -name "*.*" -exec chmod 664 {} \;
echo "Files have been modified to remove the execute bit"

# if line endings in our file are windows CR change to unix
find . -type f -name "*.*" -exec dos2unix -q -o {} \;
echo "Dos 2 Unix new line conversion has completed"

# delete any previous utf-8 files so we can write new ones
find . -type f \( -iname "*utf8*" \) -exec rm -f {} \;
echo "Old UTF-8 files have been deleted"

# get all files in our current working directory and make sure they are UTF-8 charset
find . -type f \( ! -iname "*.ttf" ! -iname "*.gif" ! -iname "*.jpg" ! -iname "*.jpeg" ! -iname "*.png" \) -exec iconv -f $CONVERT_FROM -t $CONVERT_TO -s -o {}.utf8 {} \;
echo "New UTF-8 files have been created"

# delete unconverted files so we can move the new files back later
find . -type f \( ! -iname "*.utf8" ! -iname "*.ttf" ! -iname "*.gif" ! -iname "*.jpg" ! -iname "*.jpeg" ! -iname "*.png" \) -exec rm {} \;
echo "Removal of original unconverted files is complete"

# rename all files converted to utf-8 with utf8 extension
mmv -m ";*.utf8" "#1#2"
# deal with .htaccess
mmv -m ";.*.utf8" "#1.#2"
echo "Dist Prep has completed"

This is the first step in the rest of my automation scheme. After this runs then I will work on some automated diffs and tar functions.

Thanks for any advice.
Best Regards,
Brandon

2

did you mean 'mv' instead of 'mmv'

3

don't have to use that many find commands. you can use it just one time, then pipe whatever files it find to while read loop

find . -type f -name "*.*" | while read FILE
do
 chmod 664 $FILE
 dos2unix $FILE $FILE
 case $FILE in
    *utf8* )) rm -f $FILE    
 esac
 #continue

done
4

ghostdog74, thanks for that it worked perfectly after I removed the extra ). That is much more elegant and succinct then what I had.

find . -type f -name "*.*" | while read FILE
do
  chmod 664 $FILE
  dos2unix $FILE $FILE
  case $FILE in
  *utf8* ) rm -f $FILE    
 esac
#continue

done

matrixmadhan, Yes I was referring to mmv not mv. mmv was something I stumbled on while looking to do recursive file renames because I was having issues getting mv, cp or rename to work.

Thanks for the quick responses.
Best Regards,
Brandon

5

mmv?

I have not heard about, is it for multiple-mv's ? Which distribution is it?

6

I am running it on Gentoo and just emerged it

emerge mmv

According to the link below you can get it from any other repo I have had a hard time finding just a source package for it though.

Linux Blog mmv � move multiple files by wildcard patterns

Best Regards,
Brandon