Name generation using Markov chains

names.rb

# example list of names generated from http://listofrandomnames.com/

# tweakable values
PREFIX = 3
MAX_LENGTH = 30
MIN_WORDS = 2
MAX_WORDS = 3

# create dictionary from corpus of names
dict = Hash.new { |h,k| h[k] = [] }
DATA.readlines.each do |name|
  letters = "^#{name.strip.downcase}$".chars
  while letters.length > PREFIX do
    dict[letters.first(PREFIX).join] << letters[PREFIX]
    letters.shift
  end
end

# generate the name
name = ''
while name[-1] != '$'
  if dict.include?(name[-PREFIX..-1])
    name += dict[name[-PREFIX..-1]].sample
  else
    name = dict.keys.select { |k| k[0] == '^' }.sample
  end

  if name.length == MAX_LENGTH + 1
    name += '$' unless name[-1] == '$'

  elsif name.split.length < MIN_WORDS and name[-1] == '$'
    name[-1] = ' '

  elsif name.split.length == MAX_WORDS and name[-1] == ' '
    name[-1] = '$'
  end
end

puts name[1...-1].split.map(&:capitalize).join(' ')

__END__
Moon Rockefeller
Renata Pasha
Grant Gin
Echo Pinkett
Nakesha Hey
Onita Troxel
Alesha Macias
Marlene Marlin
Dana Olds
Roosevelt Standish
Ursula Bast
Mallie Holdaway
Quinton Marceau
Mariana Daye
King Lewis
Kesha Edmonson
Magen Criger
Ricarda Woelfel
Adalberto Cressey
Gidget Restrepo
Mistie Penaflor
Serafina Neidert
Altha Doggett
Ashly Iman
Christiane Sae
Madalene Costigan
Gabrielle Linen
Milda Thornsberry
Viola Sessions
Laureen Pagano
Pearly Nunnery
Obdulia Seman
Rolando Soderman
Cassaundra Rowser
Jimmy Hazelton
Dusty Newbill
Cliff Reel
Emma Tessman
Gwenda Hadnott
Gregorio Eller
Kevin Eargle
Hye Hardee
Iliana Betterton
Carmelita Mccallion
Martine Cashen
Clorinda Hajek
Vennie Moffatt
Armida Goudy
Mckenzie Shake
Lois Lacey