script/generate-mapfile-from-header.sh

   1 #!/bin/sh
   2 ###########################################################################
   3 #
   4 #  Window Maker window manager
   5 #
   6 #  Copyright (c) 2014 Christophe CURIS
   7 #  Copyright (c) 2014-2015 Window Maker Team
   8 #
   9 #  This program is free software; you can redistribute it and/or modify
  10 #  it under the terms of the GNU General Public License as published by
  11 #  the Free Software Foundation; either version 2 of the License, or
  12 #  (at your option) any later version.
  13 #
  14 #  This program is distributed in the hope that it will be useful,
  15 #  but WITHOUT ANY WARRANTY; without even the implied warranty of
  16 #  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  17 #  GNU General Public License for more details.
  18 #
  19 #  You should have received a copy of the GNU General Public License along
  20 #  with this program; if not, write to the Free Software Foundation, Inc.,
  21 #  51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
  22 #
  23 ###########################################################################
  24 #
  25 # generate-mapfile-from-header.sh:
  26 #   from a list of C header files, extract the name of the functions and
  27 #   global variables that are considered public for a library, and generate
  28 #   a 'version script' for ld with that list, so that all other symbols
  29 #   will be considered local.
  30 #
  31 # The goal is that the symbol that are not part of the public API should
  32 # not be visible to the user because it can be a source of problems (from
  33 # name clash to evolutivity issues).
  34 #
  35 ###########################################################################
  36 #
  37 # Please note that this script is writen in sh+awk on purpose: this script
  38 # is gonna be run on the machine of the person who is trying to compile
  39 # WindowMaker, and as such we cannot be sure to find any scripting language
  40 # in a known version and that works (python/ruby/tcl/perl/php/you-name-it).
  41 #
  42 # So for portability, we stick to the same sh+awk constraint as Autotools
  43 # to limit the problem, see for example:
  44 #   http://www.gnu.org/savannah-checkouts/gnu/autoconf/manual/autoconf-2.69/html_node/Portable-Shell.html
  45 #
  46 ###########################################################################
  47
  48 # Report an error on stderr and exit with status 1 to tell make could not work
  49 arg_error() {
  50     echo "$0: $@" >&2
  51     exit 1
  52 }
  53
  54 # print help and exit with success status
  55 print_help() {
  56     echo "$0: generate a script for ld to keep only the symbols from C header"
  57     echo "Usage: $0 [options...] input_file[s]"
  58     echo "valid options are:"
  59     echo "  -l         : add symbol's definition line number for debug"
  60     echo "  -n name    : name to use for the library"
  61     echo "  -v version : set the library version for ld"
  62     exit 0
  63 }
  64
  65 # Extract command line arguments
  66 while [ $# -gt 0 ]; do
  67     case $1 in
  68         -l)
  69             debug_info='        /* " symbol_type " at " FILENAME ":" NR " */'
  70           ;;
  71
  72         -n)
  73             shift
  74             echo "$1" | grep '^[A-Za-z][A-Za-z_0-9]*$' > /dev/null || arg_error "invalid name \"$1\" for a library"
  75             libname="$1"
  76           ;;
  77
  78         -v)
  79             shift
  80             version="$1"
  81             echo "$version" | grep -E '^[1-9][0-9]*(:[0-9]+(:[0-9]+)?)?$' > /dev/null || \
  82                 arg_error "version \"$1\" is not valid"
  83           ;;
  84
  85         -h|-help|--help) print_help ;;
  86         -*) arg_error "unknow option '$1'" ;;
  87
  88         *)
  89             [ -r "$1" ] || arg_error "source file \"$1\" is not readable"
  90             input_files="$input_files $1"
  91           ;;
  92     esac
  93     shift
  94 done
  95
  96 # Check consistency of command-line
  97 [ "x$input_files" = "x" ] && arg_error "no source file given"
  98
  99 # Automatic values
 100 if [ -z "$libname" ]; then
 101     # build the library name from the first header name, remove path stuff, extension and invalid characters
 102     libname="`echo "$input_files" | sed -e 's,^ ,,;s, .*$,,;s,^.*/\([^/]*\)$,\1,;s,\.[^.]*$,,;s,[^A-Za-z_0-9],_,g;s,^_*,,' `"
 103 fi
 104
 105 # Parse the input file and find the function declarations; extract the names
 106 # and place them in the 'global' section so that ld will keep the symbols;
 107 # generate the rest of the script so that other symbols will not be kept.
 108 awk '
 109 BEGIN {
 110   # Version number here uses only 1 number from the full version used in libtool
 111   libversion="'"$version"'";
 112   if (split(libversion, subversions, ":") > 1) {
 113     # Calculate [CURRENT - AGE], the goal is that the number will not
 114     # change when functions are added to the API, but it will be incremented
 115     # when functions are removed or argument change (which breaks compat)
 116     libversion = subversions[1] - subversions[3];
 117   }
 118
 119   print "/* Generated version-script for ld */";
 120   print "";
 121   print "'"$libname"'" libversion;
 122   print "{";
 123   print "  global:";
 124 }
 125
 126 /^#[ \t]*define/ {
 127   # Ignore macro definition because their content could be mis-interpreted
 128   while (/\\$/) {
 129     if (getline == 0) { break; }
 130   }
 131   next;
 132 }
 133
 134 /\/\*/ {
 135   # Remove comments to avoid mis-detection
 136   pos = index($0, "/*");
 137   comment = substr($0, pos + 2);
 138   $0 = substr($0, 1, pos - 1);
 139   while (1) {
 140     pos = index(comment, "*/");
 141     if (pos > 0) { break; }
 142     getline comment;
 143   }
 144   $0 = $0 substr(comment, pos + 2);
 145 }
 146
 147 /extern[ \t].*;/ {
 148   line = $0;
 149
 150   # Remove everything from the ";"
 151   sub(/;.*$/, "", line);
 152
 153   # Check if it is a global variable
 154   nb = split(line, words, /[\t *]/);
 155
 156   valid = 1;
 157   for (i = 1; i < nb; i++) {
 158     # If the word looks valid, do not abort
 159     if (words[i] ~ /^[A-Za-z_0-9]*$/) { continue; }
 160
 161     # Treat the case were the variable is an array, and there was space(s) after the name
 162     if (words[i] ~ /^\[/) { nb = i - 1; break; }
 163
 164     # If we are here, the word is not valid; we stop processing the line here but we do
 165     # not use "next" so the line may match function prototype handler below
 166     valid = 0;
 167     break;
 168   }
 169
 170   if (valid) {
 171     # Remove array size, if any
 172     sub(/\[.*$/, "", words[nb]);
 173
 174     if (words[nb] ~ /^[A-Za-z][A-Za-z_0-9]*$/) {
 175       symbol_type = "variable";
 176       print "    " words[nb] ";'"$debug_info"'";
 177       next;
 178     }
 179   }
 180 }
 181
 182 /^[ \t]*typedef/ {
 183   # Skip type definition because they could be mis-interpreted when it
 184   # defines a prototype for a callback function
 185   nb_braces = 0;
 186   while (1) {
 187     # Count the number of brace pairs to skip the content of the definition
 188     nb = split($0, dummy_array, /\{/);
 189     nb_braces = nb_braces + nb;
 190
 191     nb = split($0, dummy_array, /\}/);
 192     nb_braces = nb_braces - nb;
 193
 194     # Stop when we have out count of pair of braces and there is the final ";"
 195     if ((nb_braces == 0) && (dummy_array[nb] ~ /;/)) { break; }
 196
 197     if (getline == 0) {
 198       break;
 199     }
 200   }
 201   next;
 202 }
 203
 204 /[\t ]\**[A-Z_a-z][A-Z_a-z0-9]*[\t ]*\(/ {
 205   # Get everything until the end of the definition, to avoid mis-interpreting
 206   # arguments and attributes on next lines
 207   while (1) {
 208     pos = index($0, ";");
 209     if (pos != 0) { break; }
 210
 211     getline nxt;
 212     $0 = $0 nxt;
 213   }
 214
 215   # Remove the argument list
 216   sub(/[ \t]*\(.*$/, "");
 217
 218   # Trim possible indentation at beginning of line
 219   sub(/^[\t ]*/, "");
 220
 221   # If it does not start by a keyword, it is probably not a valid function
 222   if (!/^[A-Z_a-z]/) { next; }
 223
 224   nb = split($0, words, /[\t *]/);
 225   for (i = 1; i < nb; i++) {
 226     # Do not consider "static" because it is used for functions to be inlined
 227     if (words[i] == "static") { next; }
 228
 229     # Allow empty keyword, that were just "*"
 230     if (words[i] == "") { continue; }
 231
 232     # If it does not match, it is unlikely to be a function
 233     if (words[i] !~ /^[A-Z_a-z][A-Z_a-z0-9]*$/) { next; }
 234   }
 235
 236   # If we arrived so far, everything looks valid and the last argument of
 237   # the array contains the name of the function
 238   symbol_type = "function";
 239   print "    " words[nb] ";'"$debug_info"'";
 240 }
 241
 242 END {
 243   print "";
 244   print "  local:";
 245   print "    *;";
 246   print "};";
 247 }
 248 ' $input_files