Make project a Perl package
Change-Id: I9cf3c99a1b849d939ace9c9a3b265267e45ab21c
diff --git a/.gitignore b/.gitignore
index aec92ac..9ca8f68 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1,6 +1,5 @@
log
blib
-script*
MYMETA*
Makefile
pm_to_blib
@@ -16,3 +15,4 @@
npm-debug.log
package-lock.json
_Inline
+!.gitlab-ci.yml
diff --git a/Makefile.PL b/Makefile.PL
new file mode 100644
index 0000000..bd46e23
--- /dev/null
+++ b/Makefile.PL
@@ -0,0 +1,40 @@
+#!/usr/bin/env perl
+use v5.16;
+use strict;
+use warnings;
+use ExtUtils::MakeMaker;
+
+WriteMakefile(
+ NAME => 'derekovecs::server',
+ AUTHOR => 'Marc Kupietz',
+ ABSTRACT => 'Visualization of paradigmatic and syntagmatic word relations.',
+ VERSION_FROM => 'script/derekovecs-server',
+ LICENSE => 'apache',
+ BUILD_REQUIRES => {},
+ PREREQ_PM => {
+ 'Inline::C' => '0.82',
+ 'Getopt::Std' => 0,
+ 'Getopt::Long' => 0,
+ 'Log::Any' => '1.701',
+ 'Encode' => "3.17",
+ 'Mojolicious' => '9.23',
+ 'Mojolicious::Plugin::Util::RandomString' => '0.09',
+ 'Mojolicious::Plugin::CHI' => '0.20',
+ 'Mojolicious::Plugin::ClientIP' => '0.02',
+ 'Mojolicious::Plugin::AutoSecrets' => '0.006',
+ 'Mojolicious::Plugin::RequestBase' => '0.3',
+ 'Mojolicious::Plugin::RemoteAddr' => 0,
+ 'Mojolicious::Plugin::Log::Access' => '0.041',
+ 'Mojolicious::Plugin::Piwik' => '1.00',
+ 'Mojo::JSON' => 0,
+
+ # Required for older perl bundles
+ 'List::Util' => 1.45,
+
+ # Currently on GitHub only (github.com/akron)
+ 'Mojolicious::Plugin::Localize' => '0.21'
+
+ },
+ MIN_PERL_VERSION => '5.016',
+ EXE_FILES => ['./script/derekovecs-server']
+);
diff --git a/lib/IDS/DeReKoVecs/Read.pm b/lib/IDS/DeReKoVecs/Read.pm
new file mode 100644
index 0000000..d76ddf0
--- /dev/null
+++ b/lib/IDS/DeReKoVecs/Read.pm
@@ -0,0 +1,93 @@
+use utf8;
+package IDS::DeReKoVecs::Read;
+use strict;
+use warnings;
+use Config;
+
+my $src_file = undef;
+
+our $mergedEnd=0;
+our %cache;
+our %cccache; # classic collocator cache
+our %spcache; # similar profile cache
+our $opt_p = 5676;
+our $opt_C;
+
+BEGIN {
+ $src_file = __FILE__;
+ $src_file =~ s/Read.pm/derekovecs-server.c/;
+}
+
+use Inline C => "$src_file" => CLEAN_AFTER_BUILD => 0, BUILD_NOISY => 1, ccflags => $Config{ccflags} . "-Wall -fno-rtti -O4 -I/usr/local/kl/include", libs => "-L/usr/local/kl/lib64 -l:libcollocatordb.so.1.3.0";
+#use Inline C => Config => BUILD_NOISY => 1, CFLAGS => $Config{cflags}." -O4 -mtune k9";
+#use Inline C => Config => CLEAN_AFTER_BUILD => 0, ccflags => $Config{ccflags}." -Ofast -march k8 -mtune k8 ";
+
+use Mojo::JSON qw(decode_json encode_json to_json);
+use Exporter qw(import);
+
+our @EXPORT = qw(init_net load_sprofiles getCollocationAssociation getClassicCollocatorsCached getClassicCollocatorsCached getBiggestMergedDifferences filter_garbage get_neighbours);
+
+
+sub getCollocationAssociation {
+ my ($c, $word, $collocate) = @_;
+ return getCollocationScores($word, $collocate)
+}
+
+sub getClassicCollocatorsCached {
+ my ($c, $word) = @_;
+ my $s2 = "";
+ if($word > $mergedEnd) {
+ $word-=$mergedEnd;
+ }
+
+ if($opt_p >= 5000 && $opt_p < 5600) { # German non-reference
+ open PIPE, "GET http://corpora.ids-mannheim.de/openlab/derekovecs/getClassicCollocators?w=$word |";
+ }
+ if($opt_C || !$cccache{$word}) {
+ $c->app->log->info("Getting classic collocates of $word.");
+ $cccache{$word} = getClassicCollocators($word);
+ $cccache{$word} =~ s/:(-?)(nan|inf)/:"${1}${2}"/g;
+ } else {
+ $c->app->log->info("Getting classic collocates for $word from cache.");
+ }
+ if($opt_p >= 5000 && $opt_p < 5600) { # German non-reference
+ while(<PIPE>) {
+ $s2 .= $_;
+ }
+ close(PIPE);
+ }
+
+ if(length($s2) > 2000) {
+ my $d1 = decode_json($cccache{$word});
+ my $d2 = decode_json($s2);
+ my %d2ld;
+ my $minLd = 14;
+ foreach my $i (@{$d2->{collocates}}) {
+ $d2ld{$i->{word}}=$i->{ld};
+ $minLd=$i->{ld} if($i->{ld} < $minLd);
+ }
+ foreach my $i (@{$d1->{collocates}}) {
+ my $w = $i->{word};
+ $i->{delta} = $i->{ld} - (defined $d2ld{$w} ? $d2ld{$w} : $minLd-0.1);
+ }
+ return(encode_json($d1));
+ } else {
+ my $d1 = decode_json($cccache{$word});
+ foreach my $i (@{$d1->{collocates}}) {
+ $i->{delta} = 0;
+ }
+ return(encode_json($d1));
+ }
+}
+
+sub getSimilarProfilesCached {
+ my ($c, $word) = @_;
+ if(!$spcache{$word}) {
+ $spcache{$word} = getSimilarProfiles($word);
+ } else {
+ $c->app->log->info("Getting similar profiles for $word from cache:");
+ }
+ return $spcache{$word};
+}
+
+return 1;
diff --git a/derekovecs-server.c b/lib/IDS/DeReKoVecs/derekovecs-server.c
similarity index 100%
rename from derekovecs-server.c
rename to lib/IDS/DeReKoVecs/derekovecs-server.c
diff --git a/derekovecs-server.pl b/script/derekovecs-server
similarity index 81%
rename from derekovecs-server.pl
rename to script/derekovecs-server
index 7df39d0..8f71355 100755
--- a/derekovecs-server.pl
+++ b/script/derekovecs-server
@@ -1,7 +1,7 @@
#!/usr/local/bin/perl
-use Inline C => "./derekovecs-server.c" => CLEAN_AFTER_BUILD => 0, BUILD_NOISY => 1, ccflags => $Config{ccflags} . "-Wall -fno-rtti -O4 -I/usr/local/kl/include", libs => "-L/usr/local/kl/lib64 -l:libcollocatordb.so.1.3.0";
-#use Inline C => Config => BUILD_NOISY => 1, CFLAGS => $Config{cflags}." -O4 -mtune k9";
-#use Inline C => Config => CLEAN_AFTER_BUILD => 0, ccflags => $Config{ccflags}." -Ofast -march k8 -mtune k8 ";
+our $VERSION = '0.90';
+
+use IDS::DeReKoVecs::Read qw(init_net load_sprofiles getCollocationAssociation getClassicCollocatorsCached getClassicCollocatorsCached getBiggestMergedDifferences filter_garbage get_neighbours);
use Mojolicious::Lite;
use Mojo::JSON qw(decode_json encode_json to_json);
use base 'Mojolicious::Plugin';
@@ -11,7 +11,7 @@
#use Mojo::Server::Daemon;
use Cwd;
-my $mojo_config = $ENV{MOJO_CONFIG} // 'derekovecs-server.conf';
+my $mojo_config = $ENV{MOJO_CONFIG} // '../derekovecs-server.conf';
plugin Config => {file => $mojo_config};
my $DEFAULT_VECS = app->config->{w2v}->{vecs} // "../models/dereko-2021-i.vecs";
@@ -40,26 +40,27 @@
dict => {
_ => sub { $_->locale },
},
- resources => ['derekovecs-server.dict']
+ resources => ['../derekovecs-server.dict']
};
our $opt_i = 0; # latin1-input?
our $opt_l = undef;
-our $opt_p = 5676;
our $opt_m;
our $opt_M;
our $opt_n = $DEFAULT_NET;
our $opt_d;
our $opt_D;
our $opt_G = 1;
+
+our $mergedEnd=0;
+our %cache;
+our %cccache; # classic collocator cache
+our %spcache; # similar profile cache
+our $opt_p = 5676;
our $opt_C;
my %marked;
my $title="";
my $training_args="";
-my $mergedEnd=0;
-my %cache;
-my %cccache; # classic collocator cache
-my %spcache; # similar profile cache
getopts('d:D:Gil:p:m:n:M:C');
@@ -148,67 +149,6 @@
}
}
-sub getCollocationAssociation {
- my ($c, $word, $collocate) = @_;
- return getCollocationScores($word, $collocate)
-}
-
-sub getClassicCollocatorsCached {
- my ($c, $word) = @_;
- my $s2 = "";
- if($word > $mergedEnd) {
- $word-=$mergedEnd;
- }
-
- if($opt_p >= 5000 && $opt_p < 5600) { # German non-reference
- open PIPE, "GET http://corpora.ids-mannheim.de/openlab/derekovecs/getClassicCollocators?w=$word |";
- }
- if($opt_C || !$cccache{$word}) {
- $c->app->log->info("Getting classic collocates of $word.");
- $cccache{$word} = getClassicCollocators($word);
- $cccache{$word} =~ s/:(-?)(nan|inf)/:"${1}${2}"/g;
- } else {
- $c->app->log->info("Getting classic collocates for $word from cache.");
- }
- if($opt_p >= 5000 && $opt_p < 5600) { # German non-reference
- while(<PIPE>) {
- $s2 .= $_;
- }
- close(PIPE);
- }
-
- if(length($s2) > 2000) {
- my $d1 = decode_json($cccache{$word});
- my $d2 = decode_json($s2);
- my %d2ld;
- my $minLd = 14;
- foreach my $i (@{$d2->{collocates}}) {
- $d2ld{$i->{word}}=$i->{ld};
- $minLd=$i->{ld} if($i->{ld} < $minLd);
- }
- foreach my $i (@{$d1->{collocates}}) {
- my $w = $i->{word};
- $i->{delta} = $i->{ld} - (defined $d2ld{$w} ? $d2ld{$w} : $minLd-0.1);
- }
- return(encode_json($d1));
- } else {
- my $d1 = decode_json($cccache{$word});
- foreach my $i (@{$d1->{collocates}}) {
- $i->{delta} = 0;
- }
- return(encode_json($d1));
- }
-}
-
-sub getSimilarProfilesCached {
- my ($c, $word) = @_;
- if(!$spcache{$word}) {
- $spcache{$word} = getSimilarProfiles($word);
- } else {
- $c->app->log->info("Getting similar profiles for $word from cache:");
- }
- return $spcache{$word};
-}
post '/derekovecs/getVecsByRanks' => sub {
my $self = shift;
@@ -432,6 +372,7 @@
}
);
+app->renderer->paths([app->home->rel_file('../templates')]);
app->start;
#$daemon->run;
# app->start;