Mini Shell

#!/usr/bin/sh
#
# smistrip --
#
#	Extract MIB and PIB modules from text files, like RFCs or I-Ds.
#
# Copyright (c) 1999 Frank Strauss, Technical University of Braunschweig.
#
# See the file "COPYING" for information on usage and redistribution
# of this file, and for a DISCLAIMER OF ALL WARRANTIES.
#
# $Id: smistrip.in 3434 2006-04-07 07:02:51Z strauss $
#
# NOTE, that this script relies on awk (tested with GNU awk) and getopts
# (shell builtin like in bash or standalone).
#

AWK=/usr/bin/awk
GETOPTS=getopts
VERSION=0.4.8


do_version () {
    echo "smistrip $VERSION"
}



do_usage () {
    echo "Usage: smistrip [-Vhn] [-i dir] [-d dir] [-m module] file1 [file2 [...]]"
    echo "-V         show version and license information"
    echo "-h         show usage information"
    echo "-n         do not write module files"
    echo "-i dir     try to read files from directory dir"
    echo "-d dir     write module to directory dir"
    echo "-m module  strip only the specified module"
    echo "file1 ...  input files to parse (RFCs, I-Ds, ...)"
}



do_strip () {
    if [ "$indir" ] ; then
	FILE="$indir/$1"
    else
	FILE="$1"
    fi
    if [ ! -f "$FILE" -a -f "$FILE.gz" ] ; then
	FILE="$FILE.gz"
    fi
    echo "$FILE" | grep -q '\.gz$'
    if [ $? -ne 0 ] ; then
    	CMD=cat
    else
    	CMD=zcat
    fi

    $CMD "$FILE" | \
    tr -d '\015' | \
    grep -i -v '^[ ]*Internet[ \-]Draft' | \
    $AWK -vtest="$test" -vdir="$dir" -vsingle="$single" '

    # start of module
    /^[ \t]*[A-Za-z0-9-]* *(PIB-)?DEFINITIONS *(::=)? *(BEGIN)? *$/ {
	module = $1
	skip = 9
	skipped = -1
	macro = 0
	n = 0
    }

    # process each line
    {
	# at the end of a page we go back one line (which is expected to
	# be a separator line), and start the counter skipped to skip the
	# next few lines.
	if ($0 ~ /\[[pP]age [iv0-9]*\] */) {
            # some drafts do not use that separator line. so keep it if
            # there are non-blank characters.
            if (!(line[n] ~ /^[ \t]*$/)) { print "WARNING: the line ::"line[n]":: should be a separator before a page break. It was kept. " ; n-- }
	    skipped = 0
	}

	# if we are skipping...
	if (skipped >= 0) {
	    skipped++

	    # if we have skipped enough lines to the top of the next page...
	    if (skipped >= skip) {
		skipped = -1
	    } else {
    
	    	# finish skipping, if we find a non-empty line, but not before
	    	# we have skipped four lines. remember the miminum of lines
	    	# we have ever skipped to keep empty lines in a modules that
	    	# appear near the top of a page.
	    	if ((skipped >= 4) && ($0 ~ /[^ \t]/)) {
		    if (skipped < skip) { skip = skipped }
		    skipped = -1
	    	}   
	    }
	}

	# so, if we are not skipping and inside a module, remember the line.
        if ((skipped == -1) && (length(module) > 0)) {
	    line[n++] = $0
	}
    }

    # remember when we enter a macro definition
    /^[ \t]*[A-Za-z0-9-]* *MACRO *::=/ {
	macro = 1
    }

    # end of module
    /^[ \t]*END[ \t]*$/ {
	if (macro == 0) {
	    if ((length(single) == 0) || (single == module)) {
		strip = 99
		for (i=0 ; i < n ; i++) {
		    # find the minimum column that contains non-blank characters
		    # in order to cut a blank prefix off. Ignore lines that only
                    # contain white spaces.
                    if (!(line[i] ~ /^[ \t]*$/)) {
        	        p = match(line[i], "[^ ]")
		        if ((p < strip) && (length(line[i]) > p)) { strip = p }
                    }
		}
    
		if (test != "1") {
		    if (dir) {
		       f = dir"/"module
		    } else {
		       f = module
		    }
		    for (i=0 ; i < n ; i++) {
			print substr(line[i], strip) >f
		    }
		}
    
		print module ": " n " lines."
	    }
	    module = ""
	} else {
	    macro = 0
	}
    }
    '
}



while $GETOPTS Vhnm:i:d: c ; do
    case $c in
	n)	test=1
		;;
	m)	single=$OPTARG
		;;
	i)	indir=$OPTARG
		;;
	d)	dir=$OPTARG
		;;
	h)	do_usage
		exit 0
		;;
	V)	do_version
		exit 0
		;;
	*)	do_usage
		exit 1
		;;
    esac
done

shift `expr $OPTIND - 1`



if [ $# -eq 0 ] ; then
    do_strip -
else 
    for f in "$@" ; do
	do_strip "$f"
    done
fi

exit 0