Support non-word-tokens (fixes #5)
Change-Id: I6867745afd7c0fb865722bcd62a0724aaa9a6ccb
diff --git a/t/script/non_word_tokens.t b/t/script/non_word_tokens.t
new file mode 100644
index 0000000..0eb9e14
--- /dev/null
+++ b/t/script/non_word_tokens.t
@@ -0,0 +1,95 @@
+#/usr/bin/env perl
+use strict;
+use warnings;
+use File::Basename 'dirname';
+use File::Spec::Functions qw/catdir catfile/;
+use File::Temp qw/ :POSIX /;
+use Mojo::File;
+use Mojo::JSON qw/decode_json/;
+use IO::Uncompress::Gunzip;
+use Test::More;
+use Test::Output;
+use Data::Dumper;
+use utf8;
+
+my $f = dirname(__FILE__);
+my $script = catfile($f, '..', '..', 'script', 'korapxml2krill');
+
+my $input = catdir($f, '..', 'corpus', 'WPD', '00001');
+ok(-d $input, 'Input directory found');
+
+my $output = tmpnam();
+my $cache = tmpnam();
+
+ok(!-f $output, 'Output does not exist');
+
+my $call = join(
+ ' ',
+ 'perl', $script,
+ '--input' => $input,
+ '--output' => $output,
+ '--cache' => $cache,
+ '-t' => 'OpenNLP#tokens',
+ '-l' => 'INFO'
+);
+
+# Test without compression
+stderr_like(
+ sub {
+ system($call);
+ },
+ qr!The code took!,
+ $call
+);
+
+ok(-f $output, 'Output does exist');
+ok((my $file = Mojo::File->new($output)->slurp), 'Slurp data');
+ok((my $json = decode_json $file), 'decode json');
+is($json->{textSigle}, 'WPD/AAA/00001', 'text sigle');
+is($json->{title}, 'A', 'Title');
+is($json->{data}->{tokenSource}, 'opennlp#tokens', 'Title');
+is($json->{data}->{foundries}, 'base base/paragraphs base/sentences connexor connexor/morpho connexor/phrase connexor/sentences connexor/syntax corenlp corenlp/sentences dereko dereko/structure mate mate/dependency mate/morpho opennlp opennlp/morpho opennlp/sentences treetagger treetagger/morpho treetagger/sentences xip xip/constituency xip/morpho xip/sentences', 'Foundries');
+my $stream = $json->{data}->{stream};
+my $token = $stream->[12];
+is($token->[17], 's:Vokal', 'Token');
+$token = $stream->[13];
+is($token->[23], 's:Der', 'Token');
+
+
+$call = join(
+ ' ',
+ 'perl', $script,
+ '--input' => $input,
+ '--output' => $output,
+ '--cache' => $cache,
+ '-t' => 'OpenNLP#tokens',
+ '-l' => 'INFO',
+ '-w' => '',
+ '-nwt' => ''
+);
+
+# Test without compression
+stderr_like(
+ sub {
+ system($call);
+ },
+ qr!The code took!,
+ $call
+);
+
+ok(-f $output, 'Output does exist');
+ok(($file = Mojo::File->new($output)->slurp), 'Slurp data');
+ok(($json = decode_json $file), 'decode json');
+$stream = $json->{data}->{stream};
+$token = $stream->[12];
+is($token->[17], 's:Vokal', 'Token');
+$token = $stream->[13];
+is($token->[7], 's:.', 'Token');
+is($token->[11], 'xip/p:PUNCT', 'Token');
+$token = $stream->[14];
+is($token->[23], 's:Der', 'Token');
+
+
+done_testing;
+
+__END__
diff --git a/t/tokenization.t b/t/tokenization.t
new file mode 100644
index 0000000..4c4ddaa
--- /dev/null
+++ b/t/tokenization.t
@@ -0,0 +1,65 @@
+#!/usr/bin/env perl
+use strict;
+use warnings;
+use utf8;
+use Test::More;
+use JSON::XS;
+
+use File::Basename 'dirname';
+use File::Spec::Functions 'catdir';
+
+sub _t2h {
+ my $string = shift;
+ $string =~ s/^\[\(\d+?-\d+?\)(.+?)\]$/$1/;
+ my %hash = ();
+ foreach (split(qr!\|!, $string)) {
+ $hash{$_} = 1;
+ };
+ return \%hash;
+};
+
+use_ok('KorAP::XML::Krill');
+
+my $path = catdir(dirname(__FILE__), 'corpus/WPD/00001');
+ok(my $doc = KorAP::XML::Krill->new( path => $path ), 'Load Korap::Document');
+like($doc->path, qr!\Q$path\E/$!, 'Path');
+ok($doc->parse, 'Parse document');
+is($doc->text_sigle, 'WPD/AAA/00001', 'ID');
+
+
+# Get tokens
+use_ok('KorAP::XML::Tokenizer');
+
+# Get tokenization
+ok(my $tokens = KorAP::XML::Tokenizer->new(
+ path => $doc->path,
+ doc => $doc,
+ foundry => 'OpenNLP',
+ layer => 'Tokens',
+ name => 'tokens'
+), 'New Tokenizer');
+ok($tokens->parse, 'Parse');
+
+like($tokens->stream->pos(12)->to_string, qr/s:Vokal/);
+like($tokens->stream->pos(13)->to_string, qr/s:Der/);
+
+
+# Get tokenization with non word tokens
+ok($tokens = KorAP::XML::Tokenizer->new(
+ path => $doc->path,
+ doc => $doc,
+ foundry => 'OpenNLP',
+ layer => 'Tokens',
+ name => 'tokens',
+ non_word_tokens => 1
+), 'New Tokenizer');
+ok($tokens->parse, 'Parse');
+
+like($tokens->stream->pos(12)->to_string, qr/s:Vokal/);
+like($tokens->stream->pos(13)->to_string, qr/s:\./);
+like($tokens->stream->pos(14)->to_string, qr/s:Der/);
+
+
+done_testing;
+
+__END__