use Data::Dumper;
-our $VERSION = '1.00';
+our $VERSION = '1.01';
-if (0) {
- #TODO: automatic download if not specified on stdin
+# determine input data source
+my $input;
+if (@ARGV) {
+ # manual contents specified (either piped or filename(s) given)
+ $input = \*ARGV;
+}
+else {
+ # automatic download from official website
require LWP::Simple;
- LWP::Simple::get('http://www.ietf.org/rfc/rfc1345.txt');
+ my $contents = LWP::Simple::get('http://www.ietf.org/rfc/rfc1345.txt')
+ or die "Couldn't download RFC-1345 from ietf.org";
+ open $input, '<', \$contents; # emulate file handle
}
# skip everything until a character indented by 1 space (table start)
-do {$_ = <>} until /^\s\S/;
+do {
+ $_ = readline $input;
+ defined or die "Premature input end";
+} until /^\s\S/;
my @t = $_; # add first line (already read, assume it's ok)
# read the rest of the character table
-while ($_ = <>) {
+while ($_ = readline $input) {
# check for table end (chapter 4)
- last if /^4/;
+ last if /^\d/;
# parse table lines (ignore (unindented) page break)
next unless s/^ //;
) for sort keys %di;
print "}\n";
+__END__
+
+=head1 NAME
+
+rfc1345convert - Output digraph data from RFC-1345
+
+=head1 SYNOPSIS
+
+Download and convert the digraph specification from ietf.org:
+
+ rfc1345convert > digraphs.inc.pl
+
+Test by printing the character for DO (should be a dollar sign):
+
+ perl -e'$di = do "digraphs.inc.pl"; print chr $di->{DO}->[0]'
+
+Manual specification of source retrieval:
+
+ rfc1345convert rfc1345.txt
+ curl $url | rfc1345convert -
+
+=head1 DESCRIPTION
+
+Parses the official RFC-1345 document, searching the
+'character mnemonic table' for all digraph definitions.
+If successful, Perl code is output resulting in a hash
+with character data keyed by digraph.
+Any errors and warnings are given at STDERR.
+
+The value can either be a scalar string containing another
+digraph which can be considered identical (usually inverted),
+or an array ref containing at least the resulting character's
+Unicode code point value. If available, the following UCD data
+is appended: character name, category, script, and output string.
+For example:
+
+ +{
+ AE => [198, 'LATIN CAPITAL LETTER AE', 'Lu Xl', 'Latin'],
+ EA => 'AE',
+ }
+
+=head1 AUTHOR
+
+Mischa POSLAWSKY <perl@shiar.org>
+
+=head1 LICENSE
+
+Licensed under the GNU Affero General Public License version 3.
+