tools/license-detector

   1 #! /usr/bin/ksh
   2 #
   3 #
   4 # This file and its contents are supplied under the terms of the
   5 # Common Development and Distribution License ("CDDL"), version 1.0.
   6 # You may only use this file in accordance with the terms of version
   7 # 1.0 of the CDDL.
   8 #
   9 # A full copy of the text of the CDDL should have accompanied this
  10 # source.  A copy of the CDDL is also available via the Internet at
  11 # http://www.illumos.org/license/CDDL.
  12 #
  13
  14 #
  15 # Copyright 2022 Marcel Telka
  16 #
  17
  18
  19 function usage
  20 {
  21         [[ -n "$1" ]] && printf "ERROR: %s\n\n" "$1" >&2
  22         printf "Usage: license-detector [-d] [-l LICENSE] LICENSE_FILE\n" >&2
  23         [[ -n "$1" ]] && exit 1
  24         exit 0
  25 }
  26
  27
  28 DEBUG=0
  29 LICENSE="*"
  30 while getopts ":hdl:" OPT ; do
  31         case "$OPT" in
  32         "?"|"h")        usage ;;
  33         "d")            DEBUG=1 ;;
  34         "l")            LICENSE="*.$OPTARG" ;;
  35         esac
  36 done
  37 shift $((OPTIND - 1))
  38
  39 (($# == 0)) && usage
  40 (($# > 1)) && usage "Too many arguments"
  41
  42 LICENSE_FILE="$1"
  43 [[ -e "$LICENSE_FILE" ]] || usage "$LICENSE_FILE not found"
  44 [[ -d "$LICENSE_FILE" ]] && usage "$LICENSE_FILE is directory"
  45 [[ -r "$LICENSE_FILE" ]] || usage "Unable to read $LICENSE_FILE"
  46
  47 WS_TOOLS=$(dirname $0)
  48
  49
  50 F="$LICENSE_FILE"
  51
  52 if grep -q -i "Artistic License" "$F" ; then
  53         if ! grep -q -i "Artistic License.*2" "$F" ; then
  54                 D=
  55                 grep -q "7\. C subroutines" "$F" && grep -q "10\. THIS PACKAGE IS PROVIDED" "$F" && D="Artistic-1.0-Perl"
  56                 grep -q "7\. C or perl subroutines" "$F" && grep -q "10\. THIS PACKAGE IS PROVIDED" "$F" && D="Artistic-1.0-cl8"
  57                 grep -q "7\. C or perl subroutines" "$F" && grep -q "9\. THIS PACKAGE IS PROVIDED" "$F" && D="Artistic-1.0"
  58         else
  59                 D="Artistic-2.0"
  60         fi
  61         [[ -n "$L" && -n "$D" ]] && L="$L OR " ; L="$L$D"
  62 fi
  63
  64 if grep -A 1 "GNU GENERAL PUBLIC LICENSE" "$F" | grep -q "Version 1, February 1989" ; then
  65         D="GPL-1.0-only"
  66         grep -A 2 "GNU General Public License as published by the" "$F" | grep -q "or (at your option) any" && D="GPL-1.0-or-later"
  67         [[ -n "$L" ]] && L="$L OR " ; L="$L$D"
  68 fi
  69
  70 TMPFILE=$(mktemp -q)
  71 [[ -z "$TMPFILE" ]] && printf "ERROR: Temporary file creation failed\n" >&2 && exit 1
  72
  73 typeset -A matched
  74 for l in "$WS_TOOLS"/licenses/$LICENSE ; do
  75         [[ -f "$l" ]] || continue
  76         # skip filters
  77         [[ "$l" != "${l%.filter}" ]] && continue
  78
  79         # extract license identifier
  80         license_id="${l##*/}"
  81         license_id="${license_id#header.}"
  82         license_id="${license_id#license.}"
  83         # sanity check, this should never happen
  84         [[ -z "$license_id" ]] && continue
  85
  86         # make sure we do not match one license twice
  87         [[ -n "$matched[$license_id]" ]] && continue
  88
  89         cat <<#EOF > "$TMPFILE"
  90                 dos2unix -ascii \\
  91                         | tr -d '\\014' \\
  92                         | LC_ALL=C sed -E -e 's/^[[:space:]]+\$//g' \\
  93                         | awk '/^#/{next}/^\$/{\$0="\n"}1' ORS=' ' \\
  94                         | LC_ALL=C sed -E -e 's/[[:space:]]+/ /g' -e 's/^ //' -e 's/ \$//' -e '/^\$/d' \\
  95         EOF
  96         # Remove some reStructuredText markup
  97         if [[ "${F%.rst}" != "$F" ]] ; then
  98                 cat <<#EOF >> "$TMPFILE"
  99                                 | sed -e '/^\*\*\$/d' -e 's/^\*\*\([^*]\)/\1/' -e 's/\([^*]\)\*\*\$/\1/' -e 's/\([^*]\)\*\*\([^*]\)/\1\2/g' \\
 100                 EOF
 101         fi
 102         # Punctuation - Quotes
 103         printf '\t| sed -e %cs/“/"/g%c -e %cs/”/"/g%c \\\n' "'" "'" "'" "'" >> "$TMPFILE"
 104         # Apply filter if any
 105         [[ -x "$l.filter" ]] && printf '\t| LC_ALL=C %s \\\n' "$l.filter" >> "$TMPFILE"
 106         cat <<#EOF >> "$TMPFILE"
 107                         | LC_ALL=C tr '[:upper:]' '[:lower:]' \\
 108                         | sed -e 's|http://|https://|g' \
 109                         | tr ' ' '\\n' | fmt
 110         EOF
 111
 112         REDIRECT="/dev/null"
 113
 114         if ((DEBUG)) ; then
 115                 REDIRECT="/dev/stdout"
 116                 printf "[DBG] TEMPLATE %s\n" "${l##*/}"
 117                 . "$TMPFILE" < "$l"
 118                 printf "[DBG] FILE\n"
 119                 . "$TMPFILE" < "$F"
 120                 printf "[DBG] DIFFS\n"
 121         fi
 122
 123         diff -i <(. "$TMPFILE" < "$l") <(. "$TMPFILE" < "$F") > "$REDIRECT" || continue
 124
 125         matched[$license_id]="$l"
 126
 127         [[ -n "$L" ]] && L="$L OR "
 128         L="$L$license_id"
 129 done
 130
 131 rm -f "$TMPFILE"
 132
 133 [[ -n "$L" ]] && printf "%s\n" "$L"