source: other-projects/gti/remove_extra_lines.sh@ 30722

#! /bin/bash

cp $1 $1.bak

CLEANVERSION_NAME="$1.clean"

# http://stackoverflow.com/questions/2613800/how-to-convert-dos-windows-newline-crlf-to-unix-newline-n-in-bash-script
#sed $'s/\r$//' $1 > $CLEANVERSION_NAME
#mv $CLEANVERSION_NAME $1

# echo $1

# Removing lines that will become completely empty after sed operation
# http://soft.zoneo.net/Linux/remove_empty_lines.php
# http://www.grymoire.com/Unix/Sed.html#uh-30
sed '/ *# -- Missing translation/d' $1 > $CLEANVERSION_NAME
mv $CLEANVERSION_NAME $1

# Now we got rid of all the empty lines created by removing "Missing translation" lines
# We remove any lines containing "Missing translation" but also containing other stuff

cat $1 | sed -e 's/# -- Missing translation:.*$//' > $CLEANVERSION_NAME

# Remove comment of form: #Updated 21-Nov-2011 by <name>
# The date can be either 1 or 2 digits. It need not be followed by "by <name>"
#cat $CLEANVERSION_NAME | sed -e 's/# Updated \d\d-[a-zA-Z]{3}-\d{4} by.*$//' > $1
#cat $CLEANVERSION_NAME | sed -e 's/ *# Updated .* (by.*)?$//' > $1
cat $CLEANVERSION_NAME | sed -e 's/ *# Updated .[0-9]*\-...\-.....*$//' > $1

# To test if there were ultimately no relevant changes, work on the temporary *.clean file:
# remove all lines that are merely empty/spaces
# and any lines that are purely comments (though these may start with spaces)
# Then test if the *.clean file is non-empty at the end of it
# double sed operation to remove all comments and then all empty lines:
#   http://stackoverflow.com/questions/22080937/bash-skip-blank-lines-when-iterating-through-file-line-by-line
#cat $1 | sed -e 's/^\s*(#.*)$//' > $CLEANVERSION_NAME
sed -e '/\s*#.*$/d' -e '/^\s*$/d' $1 > $CLEANVERSION_NAME


# if the *.clean file is actually empty, print a message
# http://www.unix.com/shell-programming-and-scripting/21753-check-if-file-empty.html
if [[ ! -s $CLEANVERSION_NAME ]] ; then
    echo "@@@ $1 only has comments and empty lines/spaces."
else


    # remove empty newlines, modifying the file in place.
    # http://stackoverflow.com/questions/16414410/delete-empty-lines-using-sed
    # http://unix.stackexchange.com/questions/76061/can-sed-remove-double-newline-characters
    #sed -i '/^$/d' $1


    # remove empty spaces from lines that are purely empty spaces, but leave
    # the newline intact, as we'll remove it further below.
    sed -e 's/^\s*$//' $1 > $1.tmp

    # If there's multiple new lines (3 or more) make them just 2. So leave
    # one blank line between lines wherever there were many more blank lines
    # http://stackoverflow.com/questions/922449/how-can-i-replace-mutliple-empty-lines-with-a-single-empty-line-in-bash
    # "cat with the -s option causes it to remove repeated empty lines from its output"
    # Beware: on Solaris, the -s option to cat can mean "cat is silent about missing files"
    cat -s $1.tmp > $1
fi ;
rm $CLEANVERSION_NAME


echo "**** cleaned version $1"
echo "**** pre-cleaned original is in $1.bak"