src/common/unicode/generate-unicode_east_asian_fw_table.pl

   1 #!/usr/bin/perl
   2 #
   3 # Generate a sorted list of non-overlapping intervals of East Asian Wide (W)
   4 # and East Asian Fullwidth (F) characters, using Unicode data files as input.
   5 # Pass EastAsianWidth.txt as argument.  The output is on stdout.
   6 #
   7 # Copyright (c) 2019-2022, PostgreSQL Global Development Group
   8
   9 use strict;
  10 use warnings;
  11
  12 my $range_start = undef;
  13 my ($first, $last);
  14 my $prev_last;
  15
  16 print
  17   "/* generated by src/common/unicode/generate-unicode_east_asian_fw_table.pl, do not edit */\n\n";
  18
  19 print "static const struct mbinterval east_asian_fw[] = {\n";
  20
  21 foreach my $line (<ARGV>)
  22 {
  23         chomp $line;
  24         $line =~ s/\s*#.*$//;
  25         next if $line eq '';
  26         my ($codepoint, $width) = split ';', $line;
  27
  28         if ($codepoint =~ /\.\./)
  29         {
  30                 ($first, $last) = split /\.\./, $codepoint;
  31         }
  32         else
  33         {
  34                 $first = $last = $codepoint;
  35         }
  36
  37         ($first, $last) = map(hex, ($first, $last));
  38
  39         if ($width eq 'F' || $width eq 'W')
  40         {
  41                 # fullwidth/wide characters
  42                 if (!defined($range_start))
  43                 {
  44                         # save for start of range if one hasn't been started yet
  45                         $range_start = $first;
  46                 }
  47                 elsif ($first != $prev_last + 1)
  48                 {
  49                         # ranges aren't contiguous; emit the last and start a new one
  50                         printf "\t{0x%04X, 0x%04X},\n", $range_start, $prev_last;
  51                         $range_start = $first;
  52                 }
  53         }
  54         else
  55         {
  56                 # not wide characters, print out previous range if any
  57                 if (defined($range_start))
  58                 {
  59                         printf "\t{0x%04X, 0x%04X},\n", $range_start, $prev_last;
  60                         $range_start = undef;
  61                 }
  62         }
  63 }
  64 continue
  65 {
  66         $prev_last = $last;
  67 }
  68
  69 # don't forget any ranges at the very end of the database (though there are none
  70 # as of Unicode 13.0)
  71 if (defined($range_start))
  72 {
  73         printf "\t{0x%04X, 0x%04X},\n", $range_start, $prev_last;
  74 }
  75
  76 print "};\n";