recdiff

#!/bin/bash
#
# This file is part of ShellUtils. Copyright © 2011 Christophe Labouisse.
# 
# ShellUtils is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
# 
# ShellUtils is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
# 
# You should have received a copy of the GNU General Public License
# along with ShellUtils.  If not, see <http://www.gnu.org/licenses/>.

function CopyOrLink ()
{
    local source=$1
    local dest=$2

    if [ -f $source ]
    then
	ln -s $(realpath $source) $dest
    else
	cp -r $source $dest
    fi
}

function extractTARBZ2 ()
{
    extractTAR "$1" "$2" "j"
}

function extractTGZ ()
{
    extractTAR "$1" "$2" "z"
}

function extractTAR ()
{
    local file=$1
    local destDir=$2
    local compressFlag=$3

    tar -x${compressFlag}f $file -C $destDir
}

function extract7ZIP ()
{
    local file=$1
    local destDir=$2

    7z x -o$destDir -y >/dev/null
}

function extractZIP ()
{
    local file=$1
    local destDir=$2

    unzip -q $file -d $destDir
}

function extractDEB ()
{
    local file=$1
    local destDir=$2

    mkdir $destDir/DEBIAN
    dpkg --control $file $destDir/DEBIAN
    dpkg --extract $file $destDir
}

function extractArchive ()
{
    local archive=$1
    local recursive=$2
    local keepArchive=$3

    local destDir

    if [ -d $archive ]
    then
	destDir=$archive
	echo "$archive is a directory, skipping extraction"
    else
	destDir=${archive}-dir
	echo "Extracting $archive to $destDir"

	if [ "$keepArchive" = "yes" ]
	then
	    rm -rf $destDir
	else
	    if [ -e $destDir ]
	    then
		echo "Cannot extract archive $destDir already exists"
		exit 1
	    fi 
	fi

	mkdir $destDir

	# Find out the extraction program. Zip will be used for both
	# zip and jar and 7zip will be used for both rar and 7z
	# archives.
	local format=$(echo $archive | awk '
	/\.tar\.gz$/ { print "TGZ"; }
	/\.tgz$/ { print "TGZ"; }
	/\.tar$/ { print "TAR"; }
	/\.zip$/ { print "ZIP"; }
	/\.jar$/ { print "ZIP"; }
	/\.war$/ { print "ZIP"; }
	/\.tar\.bz$/ { print "TARBZ2"; }
	/\.7z$/ { print "7ZIP"; }
	/\.rar$/ { print "7ZIP"; }
	/\.deb$/ { print "DEB"; }
	')

	if [ -z "$format" ]
	then
	    echo "Cannot find an extractor for $archive"
	    return
	fi

	extract$format $archive $destDir

	if [ "$keepArchive" != "yes" ]
	then
	    rm -f $archive
	fi
    fi

    if [ "$recursive" = "yes" ] 
    then
	local file
	find $destDir -name '*.tar.gz' -o -name '*.jar' -o -name '*.zip' -o -name '*.tgz' -o -name '*.tar' -o -name '*.tar.bz2' -o -name '*.7z' -o -name -o -name '*.deb' | while read file
	do
	    extractArchive $file yes
	done
    fi
}

function Usage()
{
    cat <<EOF
Usage:
    $MYNAME [-l] [-c] [-w workdir] [-d diffcommand] archive1 archive2

	-l: do not recurse into archives
	-c: clear the work directory before starting
	-w workdir: specify the work directory (defaults to recdiff-workdir)
	-d diffcommand: specify the diff command to be used at the end.
                        This command will take the expanded directory
			names as arguments. The default is:
			$DEFAULT_DIFF_COMMAND

EOF
    exit 10
}

MYNAME=$(basename $0)
OPTS=$(getopt -o lcw:d: -- "$@")

RECURSIVE=yes

DEFAULT_DIFF_COMMAND="diff -qr -x MYCOM.DSA -x MYCOM.SF -x '*.html'"
DIFFCMD="$DEFAULT_DIFF_COMMAND"

if [ $? != 0 ]
then
    Usage
fi

eval set -- "$OPTS"

while true
do
    case "$1" in
        -l) RECURSIVE=no ; shift ;;
        -w) WORKDIR="$2" ; shift 2 ;;
        -c) CLEAR_WORK_DIR=yes ; shift ;;
	-d) DIFFCMD="$2" ; shift 2;;
        --) shift ; break ;;
        *) echo "Internal error!" ; exit 1 ;;
    esac
done

if [ -z "$2" ]
then
    Usage
fi

FILE1=$1
FILE2=$2
WORKDIR=${WORKDIR-recdiff-workdir}

if [ -e $WORKDIR ]
then
    if [ "$CLEAR_WORK_DIR" = "yes" ]
    then
	echo "Clearing $WORKDIR"
	rm -rf $WORKDIR
    else
	echo "$WORKDIR already exists, remove it before running recdiff"
	exit -1
    fi
fi

mkdir -p $WORKDIR

echo "Copying files to $WORKDIR"
CopyOrLink $FILE1 $WORKDIR
CopyOrLink $FILE2 $WORKDIR

cd $WORKDIR

FILE1=$(basename $FILE1)
FILE2=$(basename $FILE2)

extractArchive $FILE1 $RECURSIVE
extractArchive $FILE2 $RECURSIVE

if [ -d $FILE1 ]
then
    DIR1=$FILE1
else
    DIR1=${FILE1}-dir
fi

if [ -d $FILE2 ]
then
    DIR2=$FILE2
else
    DIR2=${FILE2}-dir
fi

$DIFFCMD $DIR1 $DIR2 | tee diff.txt

echo "Done, results in $WORKDIR/diff.txt"