Adding upstream version 1:2.3.19.1+dfsg1.upstream/1%2.3.19.1+dfsg1upstream

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

1 files changed, 34 insertions, 0 deletions

diff --git a/src/lib-fts/word-properties.pl b/src/lib-fts/word-properties.pl
new file mode 100644
index 0000000..c600d14
--- /dev/null
+++ b/src/lib-fts/word-properties.pl
@@ -0,0 +1,34 @@
+#!/usr/bin/env perl
+use strict;
+use warnings;
+
+my @categories;
+my $which = shift(@ARGV);
+if ($which eq 'boundaries') {
+    @categories = qw(CR LF Newline Extend Regional_Indicator Format Katakana Hebrew_Letter ALetter
+		    Single_Quote Double_Quote MidNumLet MidLetter MidNum Numeric ExtendNumLet);
+} elsif ($which eq 'breaks') {
+    @categories = qw(White_Space Dash Quotation_Mark Terminal_Punctuation STerm Pattern_White_Space);
+} else {
+    die "specify 'boundaries' or 'breaks'";
+}
+
+my $catregexp=join('|', @categories);
+my %catlists = map { $_ => []; } (@categories);
+
+while(<>) {
+    next if (m/^#/ or m/^\s*$/);
+    push(@{$catlists{$3}}, defined($2) ? (hex($1)..hex($2)) : hex($1))
+	if (m/([[:xdigit:]]+)(?:\.\.([[:xdigit:]]+))?\s+; ($catregexp) #/)
+}
+
+print "/* This file is automatically generated by word-properties.pl from $ARGV */\n";
+foreach(@categories) {
+    my $arref=$catlists{$_};
+    print "static const uint32_t ${_}[]= {\n";
+    while(scalar(@$arref)) {
+	print("\t", join(", ", map { sprintf("0x%05X", $_); } splice(@$arref, 0, 8)));
+	print(scalar(@$arref) ? ", \n" : "\n");
+    }
+    print("};\n");
+}