Re: Rhyme generator

[Brovnik]

I just had some fun playing with this.

A couple of comments :

1. The code to read in the dictionary is a bit inefficient. The folloing in Dictionary.pm speeds things up by removing surplus split/join code :

   our %dict;
   our %reversedict;
   
   sub read_dict
   {
     while (<DATA>)
     {
        chomp;
        if (/^([^\s]+)\s\s(.*)$/)
        {
           my ($key,$value) = ($1,$2);
           $dict{$key} = $value;
           $reversedict{$value} = $key;
        }
      }
      return;
   }
   

   This speeds up the initial read from 10+ seconds to 6.5 seconds on my machine.
2. I also played with using Storable to speed up subsequent accesses.

   my $dict = "/tmp/Lingua::EN::Rhyme::Dictionary::DICT";
   my $reversedict = "/tmp/Lingua::EN::Rhyme::Dictionary::REVDICT";
   my $loaded = retrieve_dict();
   
   sub store_dict
   {
      read_dict();
      store(\%dict,$dict);
      store(\%reversedict,$reversedict);
   }
   
   sub retrieve_dict
   {
      my $ref = retrieve("DICT");
      if ($ref)
      {
         %dict = %$ref;
         $ref = retrieve("REVDICT");
         %reversedict = %$ref;
      }
      else
      {
         # Read in normal way and store
         read_dict();
         store_dict();
      }
      return 1;
   }
   

   This uses Storable to store and retrieve the dictionaries. It saves some time, but not a lot and leves the extra files around.

   I also started playing with storing this in a mysql DB, but haven't got that clean yet (it really needs a rewrite of all of Rhyme.pl.

Hope that is of help.
--
Brovnik