DZone Snippets: dictionary code

Ruby dictionary username generation

elliottcable (elliott cable) — Thu, 13 Sep 2007 13:57:59 GMT

Generate a new random name from dictionary words.



DICT_PATH = '/usr/share/dict/words'

DICT_SIZE = 234936



def self.generated_name words = 2, length = 23

  name = 'a'*(length+1)

  while name.length > length

    name = (1..words).map{%x[sed -n '#{rand(DICT_SIZE)} {p;q;}' '#{DICT_PATH}'].chomp.capitalize}.join

  end

end

BloominSimple: Ultra-easy, pure Ruby Bloom filter library

peter (Peter Cooperx) — Mon, 02 Jul 2007 02:14:09 GMT

Requires BitField.



#        NAME: BloominSimple

#      AUTHOR: Peter Cooper

#     LICENSE: MIT ( http://www.opensource.org/licenses/mit-license.php )

#   COPYRIGHT: (c) 2007 Peter Cooper

# DESCRIPTION: Very basic, pure Ruby Bloom filter. Uses my BitField, pure Ruby

#              bit field library (http://snippets.dzone.com/posts/show/4234).

#              Supports custom hashing (default is 3).

#

#              Create a Bloom filter that uses default hashing with 1Mbit wide bitfield

#                bf = BloominSimple.new(1_000_000)

#              

#              Add items to it

#                File.open('/usr/share/dict/words').each { |a| bf.add(a) }

#

#              Check for existence of items in the filter

#                bf.includes?("people")     # => true

#                bf.includes?("kwyjibo")    # => false

#

#              Add better hashing (c'est easy!)

#                require 'digest/sha1'

#                b = BloominSimple.new(1_000_000) do |item|

#                  Digest::SHA1.digest(item.downcase.strip).unpack("VVVV")

#                end

#

#              More

#                %w{wonderful ball stereo jester flag shshshshsh nooooooo newyorkcity}.each do |a|

#                  puts "#{sprintf("%15s", a)}: #{b.includes?(a)}"

#                end

#

#                 #  =>   wonderful: true

#                 #  =>        ball: true

#                 #  =>      stereo: true

#                 #  =>      jester: true

#                 #  =>        flag: true

#                 #  =>  shshshshsh: false

#                 #  =>    nooooooo: false

#                 #  => newyorkcity: false



require 'benchmark'

require 'bitfield'



class BloominSimple

  attr_reader :bitfield, :hasher

  

  def initialize(bitsize, &block)

    @bitfield = BitField.new(bitsize)

    @size = bitsize

    @hasher = block || lambda do |word|

      word = word.downcase.strip

      [h1 = word.sum, h2 = word.hash, h2 + h1 ** 3]

    end

  end

  

  def add(item)

    @hasher[item].each { |hi| @bitfield[hi % @size] = 1 }

  end

  

  def includes?(item)

    @hasher[item].each { |hi| return false unless @bitfield[hi % @size] == 1 } and true

  end

end

Shell script to look up words at dictionary.com from the command line

madphilosopher (Darren Paul Griffith) — Wed, 09 Nov 2005 05:54:19 GMT

The following will look up a word on dictionary.com from the command line, where $1 is the word you want to look up, entered in as a parameter on the command line. (Lynx is a text-based browser that handles the html-to-text conversion for us.)



#!/bin/sh 

lynx -dump -nolist -pseudo_inlines                \

  'http://dictionary.reference.com/search?q='$1'&r=67'  \

  | tail -n +13 | less -r

Counting characters in string

korakot (Korakot Chaovavanich) — Tue, 26 Jul 2005 18:39:46 GMT



>>> s = 'a;jfkd;aflhakfhaskfjalghlakfhfnkjafyksd'

>>> cnt = {}

>>> for c in s:

	cnt[c] = cnt.get(c,0) + 1



>>> print cnt

{'a': 7, 'd': 2, 'g': 1, 'f': 7, 'h': 4, 'k': 6, 'j': 3, 'l': 3, 'n': 1, 's': 2, 'y': 1, ';': 2}

This can be used to count any distribution.
Note the use of dict.get(key,default) to set to 0
if the key is not avaiable.
If this were perl, I would just do a
cnt[c] += 1
But python will give an error instead of returning 0.
It's not too bad, though.