AI-MicroStructure

 view release on metacpan or  search on metacpan

bin/leo  view on Meta::CPAN

re="${re//oe/$onetwo}"
re="${re//ss/$onetwo}"

# have to use perl for grepping because of umlauts

q=$(lynx -dump -nolist  'http://dict.leo.org/'$from$to'?lp='$from$to'&lang='$from'&searchLoc=0&cmpType=strict&sectHdr=on&spellToler=on&search='$search'&relink=on' | perl -n -e "print if /$re/i")
#perl -MHTML::Entities  -le ' print \$q,encode_entities(\$q),\n';

#echo "$q";

perl  -MEncode -le "print decode('utf-8','$q');";

bin/micro-dict  view on Meta::CPAN

stop=$(perl -MAI::MicroStructure::WordBlacklist -E  "my \$s=AI::MicroStructure::WordBlacklist::getStopWords('de'); my @s = keys %\$s; print join('|',@s);")
IFS=$'\n';

$cmd $1 |   tr A-Z a-z |                # Convert to lowercase.
        tr ' ' '_' |             # New: change spaces to newlines.
       #tr -cd '\012[a-z][0-9]' |   #  Get rid of everything
                                    #+ non-alphanumeric (in orig. script).
        tr -c '\012a-z'  '\012' |   #  Rather than deleting non-alpha
        egrep -v '^#' |              # Delete lines starting with hashmark.
        egrep -v "^[ ]*([A-Za-z][A-Za-z]|[A-Za-z])$" | egrep -v "^$" | egrep -v -i "^ (denkbarer|ganze|bez|ver�ffentlichtes|uns�gliches|ungew�hnliche|vollstaendig|erstem|Inf.|titel|unsaeglichem|beforehand|denkbares|yours|contains|gedurft|seithe...


 stop=$(perl -MAI::MicroStructure::WordBlacklist -E  "my \$s=AI::MicroStructure::WordBlacklist::getStopWords('de'); my @s = keys %\$s; print join('|',@s);")

 cat /tmp/micro-dict.tmp | sort -n | egrep -v "^.*.[\ ].*.[1-9][\:][\ ][\ ]($stop)";


 #if [ !  "$(echo  "$stop" | egrep -i zzzzzzzzzzzz)" ]; then  echo cool; fi

bin/micro-relation-test  view on Meta::CPAN

             compress_threshold => 100_000,
             compress_ratio => 0.9,
             compress_methods => [ \&IO::Compress::Gzip::gzip,
                                   \&IO::Uncompress::Gunzip::gunzip ],
             max_failures => 3,
             failure_timeout => 2,
             ketama_points => 150,
             nowait => 1,
             hash_namespace => 1,
             serialize_methods => [ \&Storable::freeze, \&Storable::thaw ],
             utf8 => ($^V ge v5.8.1 ? 1 : 0),
             max_size => 512 * 1024,
         });





sub  decruft  {
  my($file)  =  @_;
  my($cruftSet)  =  q{%§&|#[^+*(]),'";};

bin/micro-sense  view on Meta::CPAN

#print Dumper [split("sense",`wn $line  -coorn`)];
push @{$data->{'search'}},grep{/[a-z|A-Z]/}split("\n|,",lc `micro-wnet $line words`) unless(!$words);
#push @{$data->{'coordinate'}},map{$_=trim($_); @_=grep(/[A-Z]|[0-9]|[a-z]/,map{$_=trim($_)}split("\n|=>",$_))}split(/Sense/,`wn $line -coorn`) unless(!$words);
#shift @{$data->{'coordinate'}} &&  shift @{$data->{'search'}} for(0..1);
#@{$data->{'search'}};
END{

 $data = {"rows"=>$data,"senses"=>[sort keys %{$data->{'senses'}}]};


   my $utf8_encoded_json_text = encode_json($data);

  if($debug == 1) {
    print Dumper $data,$wn;
  }


  print $utf8_encoded_json_text;

}

bin/micro-wiki  view on Meta::CPAN

#!/usr/bin/perl -w
use utf8;
use File::Basename;
use Data::Printer;
use Data::Dumper;
use Parallel::Iterator qw( iterate );
use Env qw/PWD/;
use JSON::XS;
use HTML::Strip;
use AI::MicroStructure::Util;
use WWW::Wikipedia;
use LWP::UserAgent;
use HTML::SimpleLinkExtor;
use URI::Escape qw( uri_unescape );
our $e = HTML::SimpleLinkExtor->new;

binmode STDOUT, ':utf8';
binmode STDERR, ':utf8';


my $state  = AI::MicroStructure::Util::config();

my @CWD    = $state->{cwd};
our $config = $state->{cfg};


die("require a argument") unless($ARGV[0]);

bin/remote.pl  view on Meta::CPAN

    $str;
}

sub tr_accent {
    my $str = shift;
    $str =~ tr{ÀÁÂÃÄÅÇÈÉÊËÌÍÎÏÑÒÓÔÕÖØÙÚÛÜÝàáâãäåçèéêëìíîïñòóôõöøùúûüýÿ}
              {AAAAAACEEEEIIIINOOOOOOUUUUYaaaaaaceeeeiiiinoooooouuuuyy};
    return $str;
}

my %utf2asc = (
    "\xc3\x89" => 'E',
    "\xc3\xa0" => 'a',
    "\xc3\xa1" => 'a',
    "\xc3\xa9" => 'e',
    "\xc3\xaf" => 'i',
    "\xc3\xad" => 'i',
    "\xc3\xb6" => 'o',
    "\xc3\xb8" => 'o',
    "\xc5\xa0" => 'S',
    "\x{0160}" => 'S',
    # for pokemons
    "\x{0101}"     => 'a',
    "\x{012b}"     => 'i',
    "\x{014d}"     => 'o',
    "\x{016b}"     => 'u',
    "\xe2\x99\x80" => 'female',
    "\xe2\x99\x82" => 'male',
    "\x{2640}"     => 'female',
    "\x{2642}"     => 'male',
);
my $utf_re = qr/(@{[join( '|', sort keys %utf2asc )]})/;

sub tr_utf8_basic {
    my $str = shift;
    $str =~ s/$utf_re/$utf2asc{$1}/go;
    return $str;
}

1;

package main;

use Data::Printer;
my $ps = AI::MicroStructure::KnowHow->new();
my @go = $ps->remote_Knowledge();

lib/AI/MicroStructure/RemoteList.pm  view on Meta::CPAN

    $str;
}

sub tr_accent {
    my $str = shift;
    $str =~ tr{ÀÁÂÃÄÅÇÈÉÊËÌÍÎÏÑÒÓÔÕÖØÙÚÛÜÝàáâãäåçèéêëìíîïñòóôõöøùúûüýÿ}
              {AAAAAACEEEEIIIINOOOOOOUUUUYaaaaaaceeeeiiiinoooooouuuuyy};
    return $str;
}

my %utf2asc = (
    "æ"        => 'ae',
    "Æ"        => 'AE',
    "\xc5\xa0" => 'S',
    "\x{0160}" => 'S',
    # for pokemons
    "\x{0101}"     => 'a',
    "\x{012b}"     => 'i',
    "\x{014d}"     => 'o',
    "\x{016b}"     => 'u',
    "\xe2\x99\x80" => 'female',
    "\xe2\x99\x82" => 'male',
    "\x{2640}"     => 'female',
    "\x{2642}"     => 'male',
);
my $utf_re = qr/(@{[join( '|', sort keys %utf2asc )]})/;

sub tr_utf8_basic {
    my $str = shift;
    $str =~ s/$utf_re/$utf2asc{$1}/go;
    return $str;
}

1;

__END__

=head1 NAME

AI::MicroStructure::RemoteList - Retrieval of a remote source for a structure

 view all matches for this distribution
 view release on metacpan -  search on metacpan