#!/bin/bash
#
# Perform set operations in the UNIX shell!

PROGNAME=${0##*/}
VERSION="0.1.0"

usage()
{
  cat <<EO
Usage: $PROGNAME <cmd> <args>
Perform set operations in the UNIX shell!
Commands:
EO
cat <<EO | column -s\& -t
  help                          & Shows this help
  is-member <kwd> <set1> ... <setn>  & Tests whether the line set1 is present in the file set2
  equals <set1> <set2>          & Tests whether the unique lines in file set1 are the same as the unique lines in set2
  count <set1> <set2> ... <setn> & Counts the number of unique lines in files set1, set2, ..., setn combined
  is-subset <base> <set1> ... <setn> & Tests whether all lines in file base are present in files set1, ..., setn combined
  union <set1> <set2> ... <setn> & Displays all unique lines that are present in files set1, set2, ..., setn combined
  inter <set1> <set2>          & Displays all unique lines that are common to files set1 and set2
  minus <set1> <set2>          & Displays all unique lines in file set1 that are not present in file set2
  sym-diff <set1> <set2> ... <setn> & Displays all unique lines that are present in either files set1 or set2 ... or setn, but not in all of them
  product <set1> <set2>          & Displays the cartesian product of the unique lines from files set1 and set2
  is-disjoint <set1> <set2>          & Tests whether there are any lines common to both files set1 and set2
  is-empty <set1> <set2> ... <setn> & Tests whether there are any lines in files set1, set2, ..., setn combined
  min <set1> <set2> ... <setn> & Displays the lexicographic minimum from among the lines in files set1, set2, ..., setn combined
  max <set1> <set2> ... <setn> & Displays the lexicographic maximum from among the lines in files set1, set2, ..., setn combined
EO
}

version() {
  cat <<EO
v$VERSION
EO
}

log_error_and_exit() {
  echo "setop: $1: $2" 1>&2; exit 1
}

is_not_empty() {
  head -1 "$@" | wc -l | awk {'print $1'}
}

is_empty() {
  head -1 "$@" | wc -l | awk {'print ($1 == 0 ? 1 : 0)'}
}

is_member() {
  kwd="$1"
  shift
  grep "$kwd" "$@" | is_not_empty
}

equals() {
  diff <(sort -u "$1") <(sort -u "$2") | is_empty
}

count() {
  sort -u "$@" | wc -l | awk {'print $1'}
}

is_subset() {
  base="$1"
  shift
  comm -23 <(sort -u "$base") <(sort -u "$@") | is_empty
}

union() {
  sort -u "$@"
}

inter() {
  comm -12 <(sort -u "$1") <(sort -u "$2")
}

minus() {
  comm -23 <(sort -u "$1") <(sort -u "$2")
}

sym_diff() {
  sort "$@" | uniq -u
}

product() {
  awk 'NR==FNR { a[$0]; next } { for (i in a) print i, $0 }' <(sort -u "$1") <(sort -u "$2")
}

is_disjoint() {
  inter "$1" "$2" | is_empty
}

min() {
  sort "$@" | head -1
}

max() {
  sort "$@" | tail -1
}

CMD="$1"
case $CMD in
  -h|--help|help)
    usage; exit 0
    ;;
  -v|--version|version)
    version; exit 0
    ;;
  is-member)
    kwd="$2"
    shift 2
    is_member "$kwd" "$@"
    ;;
  equals)
    equals "$2" "$3"
    ;;
  count)
    shift; count "$@"
    ;;
  is-subset)
    base="$2"
    shift 2
    is_subset "$base" "$@"
    ;;
  union)
    shift; union "$@"
    ;;
  inter)
    inter "$2" "$3"
    ;;
  minus)
    minus "$2" "$3"
    ;;
  sym-diff)
    shift; sym_diff "$@"
    ;;
  product)
    product "$2" "$3"
    ;;
  is-disjoint)
    is_disjoint "$2" "$3"
    ;;
  is-empty)
    shift; is_empty "$@"
    ;;
  min)
    shift; min "$@"
    ;;
  max)
    shift; max "$@"
    ;;
  *)
    log_error_and_exit "$CMD" "Unknown command"
    ;;
esac