How to batch convert legacy files to UTF8?

Question

I think the ideal solution for me is a Nautilus Script that performs encoding conversion on selected files.

The following script reads the encoding of a selected file and performs utf8 conversion if it's not utf8, but I couldn't figure out how to make it work on multiple files:

CHARSET="$(file -bi "$1"|awk -F "=" '{print $2}')"
if [ "$CHARSET" != utf-8 ]; then
iconv -f "$CHARSET" -t utf8 "$1" -c -o "$1.utf8"
fi

Answer 1

Let's say your script is named convert-to-utf-8.sh. Here's how you would make it work across multiple files:

for filename in file1 file2 file3 ; do ./convert-to-utf-8.sh "$filename" ; done

You could incorporate that for-loop in the script itself, like this:

for filename in "$@"; do
    CHARSET="$(file -bi "$filename"|awk -F "=" '{print $2}')"
    if [ "$CHARSET" != utf-8 ]; then
        iconv -f "$CHARSET" -t utf8 "$filename" -c -o "$filename.utf8"
    fi
done

You could then run the script with multiple filenames like this:

./convert-to-utf-8.sh file1 file2 file3

Answer 2

Here's a nice and simple Nautilus Script that I've managed to write thanks to Flimm as seen in the answer above:

#!/bin/sh
#Nautilus Script to determine encoding of selected file(s) and convert to utf8 if necessary, and then replace Turkish characters
#
for filename in $@; do
    CHARSET="$(file -bi "$filename"|awk -F "=" '{print $2}')"
    if [ "$CHARSET" != utf-8 ]; then
        iconv -f "$CHARSET" -t utf8 "$filename" -c -o "$filename.utf8" &&
        mv -f "$filename.utf8" "$filename"
    fi
        cat "$filename" | sed -e 's/Ð/Ğ/g' -e 's/ð/ğ/g' -e 's/Þ/Ş/g' -e 's/þ/ş/g' > "$filename.rep" &&
    mv -f "$filename.rep" "$filename"   
done

It also converts certain special characters (which can be modified as needed) after the utf8 conversion.