editalign 1.0.0 → 1.1.1

data/README

@@ -43,6 +43,8 @@ arrays.  See the EditAlign::DijkstraSearch class for details.
 = History
 
 * 1-0-0 ... First version
+* 1-1-0 ... Added EditAlign::VERSION
+* 1-1-1 ... Improved command-line handling in align-strings and stress-test
 
 = See Also
 
@@ -55,11 +57,11 @@ strategy and optimizations to the one implemented here.
 
 = Copyright
 
-Copyright 2006, William Patrick McNeill
+Copyright 2006-2009, William Patrick McNeill
 
 This program is distributed under the GNU General Public License.
 
 = Author
 
-W.P. McNeill mailto:billmcn@u.washington.edu
+W.P. McNeill mailto:billmcn@gmail.com
 

data/examples/align-strings

@@ -1,7 +1,7 @@
-#!/bin/env ruby
+#!/usr/bin/env ruby
 
 #--
-# Copyright 2006 William Patrick McNeill
+# Copyright 2006-2009 William Patrick McNeill
 #
 # This file is part of Editalign.
 #
@@ -24,49 +24,55 @@
 # Print the character alignment between two strings passed in on the
 # command line.
 
-require 'getoptlong'
+require 'optparse'
 require 'editalign'
 
 class ExhaustiveSellersAlignment < EditAlign::SellersAlignment
   include EditAlign::ExhaustiveSearch
 end
 
-# Process command line options.
-opts = GetoptLong.new(["--match", "-m", GetoptLong::REQUIRED_ARGUMENT],
-                      ["--nomatch", "-n", GetoptLong::REQUIRED_ARGUMENT],
-                      ["--insert", "-i",GetoptLong::REQUIRED_ARGUMENT],
-                      ["--delete", "-d",  GetoptLong::REQUIRED_ARGUMENT],
-                      ["--exhaustive", "-e", GetoptLong::NO_ARGUMENT]
-                      )
-
 match = 0
-mismatch = 1
+substitute = 1
 insert = 1
 delete = 1
 exhaustive = false
-opts.each do |opt, arg|
-  case opt
-  when "--match"
-    match = arg.to_f
-  when "--nomatch"
-    mismatch = arg.to_f
-  when "--insert"
-    insert = arg.to_f
-  when "--delete"
-    delete = arg.to_f
-  when "--exhaustive"
-    exhaustive = true
+OptionParser.new do |opts|
+  opts.banner =<<-EOTEXT
+#{File.basename(__FILE__)} [OPTION] source dest
+
+Perform an edit distance alignment between the source and dest strings.
+
+The options may be used to specify match, substitution, insertion, and
+deletion costs.
+EOTEXT
+  opts.on("-m", "--match COST", Float, "Match cost (default 0)") do |value|
+    match = value
   end
-end
+  opts.on("-s", "--substitution COST", Float, "Substitution cost (default 1)") do |value|
+    substitute = value
+  end
+  opts.on("-i", "--insertion COST", Float, "Insertion cost (default 1)") do |value|
+    insert = value
+  end
+  opts.on("-d", "--deletion COST", Float, "Deletion cost (default 1)") do |value|
+    delete = value
+  end
+  opts.on("-e", "--exhaustive", "Do exhaustive search for comparison (default false)") do |value|
+    exhaustive = value
+  end
+end.parse!
 
-source = ARGV[0]
-dest = ARGV[1]
+if not ARGV.length == 2
+  puts "Incorrect number of arguments."
+  exit(0)
+end
+source, dest = ARGV
 
 # Do alignments and print results.
 alignments = [EditAlign::SellersAlignment]
 alignments << ExhaustiveSellersAlignment if exhaustive
 
 alignments.each do |align_class|
-  a = align_class.new(source, dest, match, mismatch, insert, delete)
+  a = align_class.new(source, dest, match, substitute, insert, delete)
   puts a, a.to_grid
 end

data/examples/stress-test

@@ -1,7 +1,7 @@
-#!/bin/env ruby
+#!/usr/bin/env ruby
 
 #--
-# Copyright 2006 William Patrick McNeill
+# Copyright 2006-2009 William Patrick McNeill
 #
 # This file is part of Editalign.
 #
@@ -21,7 +21,7 @@
 #
 #++
 
-require "getoptlong"
+require "optparse"
 require "editalign"
 
 class PureRubyEditAlign < EditAlign::Alignment
@@ -51,24 +51,31 @@ end
 # * verbose ... prints strings and edit distances
 # * pure-ruby ... uses the pure Ruby priority queue instead of the C extension
 # * int-cost ... uses integer costs
-def parse_command_line
-  opts = GetoptLong.new(["--verbose", "-v", GetoptLong::NO_ARGUMENT],
-                        ["--pure-ruby", "-p", GetoptLong::NO_ARGUMENT],
-                        ["--int-cost", "-i", GetoptLong::NO_ARGUMENT])
+def parse_command_line  
   verbose = false
   pure_ruby = false
   int_cost = false
-  opts.each do |opt, arg|
-    case opt
-    when "--verbose"
-      verbose = true
-    when "--pure-ruby"
-      pure_ruby = true
-    when "--int-cost"
-      int_cost = true
+  OptionParser.new do |opts|
+    opts.banner =<<-EOTEXT
+#{File.basename(__FILE__)} [OPTION] trials length edits
+
+Perform a stress test on the alignment code.
+  EOTEXT
+    opts.on("-i", "--int-cost", "Use integer costs") do |value|
+      int_cost = value
     end
+    opts.on("-p", "--pure-ruby", "Use the pure Ruby priority queue instead of the C extension") do |value|
+      pure_ruby = value
+    end
+    opts.on("-v", "--verbose", "Print strings and edit distances") do |value|
+      verbose = value
+    end
+  end.parse!
+  
+  if not ARGV.length == 3
+    puts "Incorrect number of arguments."
+    exit(0)
   end
-
   trials = Integer(ARGV[0])
   length = Integer(ARGV[1])
   edits = Integer(ARGV[2])

data/lib/editalign.rb

@@ -1,4 +1,4 @@
-# Copyright 2006 William Patrick McNeill
+# Copyright 2006-2009 William Patrick McNeill
 #
 # Editalign is free software; you can redistribute it and/or modify it
 # under the terms of the GNU General Public License as published by
@@ -18,6 +18,8 @@
 # functions.
 module EditAlign
 
+  VERSION = "1.1.1"
+
   # This module employs Dijkstra's algorithm to find the lowest-cost
   # sequence of edit operations that will transform the source array
   # into the destination array.  The alignment grid is treated as a
@@ -86,8 +88,9 @@ module EditAlign
           next unless next_cost < @cost[next_cell]
           @cost[next_cell] = next_cost
           @backtrace[next_cell] = cell
-          agenda[next_cell] = priority_factory(next_cost, next_cell) \
           unless next_cost >= @cost[@end]
+            agenda[next_cell] = priority_factory(next_cost, next_cell)
+          end
         end
       end
     end
@@ -207,32 +210,31 @@ module EditAlign
 
 
   # The Alignment class is given a source and destination array at
-  # construction time.  It does a dynamic programming alignment
-  # between them and makes the results of that alignment available
-  # through instance methods.
+  # construction time.  It does a dynamic programming alignment between them
+  # and makes the results of that alignment available through instance
+  # methods.
   #
-  # If there are multiple alignments with equal edit distances
-  # Alignment will find one of them.  Which one is undefined.
+  # If there are multiple alignments with equal edit distances Alignment will
+  # find one of them.  Which one is undefined.
   #
-  # Alignment works by constructing a matrix with dimensions equal to
-  # the length of the source and destination arrays.  Moving
-  # horizontally and vertically in the matrix represents insertion and
-  # deletion operations, respectively, while moving diagonally
-  # represents substitution.  Each cell of the matrix contains the
-  # minimum cost it takes to reach that cell.  The algorithm fills in
-  # cells in the matrix until it reaches the furthest corner.
+  # Alignment works by constructing a matrix with dimensions equal to the
+  # length of the source and destination arrays.  Moving horizontally and
+  # vertically in the matrix represents insertion and deletion operations,
+  # respectively, while moving diagonally represents substitution.  Each cell
+  # of the matrix contains the minimum cost it takes to reach that cell.  The
+  # algorithm fills in cells in the matrix until it reaches the furthest
+  # corner.
   #
-  # The search is done using Dijkstra's algorithm as implemented in
-  # the DijkstraSearch.  A different search algorithm may be specified
-  # by including a mixin that redefines the
-  # #find_lowest_cost_alignment function.
+  # The search is done using Dijkstra's algorithm as implemented in the
+  # DijkstraSearch.  A different search algorithm may be specified by
+  # including a mixin that redefines the #find_lowest_cost_alignment function.
   #
-  # This class uses Levenshtein weighting scheme. Levenshtein assigns
-  # a cost of 1 to insertions and deletions.  It assigns a cost of 1
-  # to substitutions when the items are different and 0 when they are
-  # the same.  Different weighting schemes may be specified by
-  # overloading the #insert, #delete, and #substitute functions.  The
-  # costs must be non-negative numbers.
+  # This class uses the Levenshtein weighting scheme. Levenshtein assigns a
+  # cost of 1 to insertions and deletions.  It assigns a cost of 1 to
+  # substitutions when the items are different and 0 when they are the same. 
+  # Different weighting schemes may be specified by overloading the #insert,
+  # #delete, and #substitute functions.  The costs must be non-negative
+  # numbers.
   class Alignment
     include DijkstraSearch
 
@@ -245,6 +247,8 @@ module EditAlign
     # Optionally either <em>source</em> or <em>dest</em> may be
     # strings, in which they will be treated as arrays of characters.
     def initialize(source, dest)
+      @path = nil
+      
       # Convert strings into arrays.
       source = source.unpack('U*').collect {|c| c.chr} if source.class == String
       dest = dest.unpack('U*').collect {|c| c.chr} if dest.class == String
@@ -314,7 +318,7 @@ module EditAlign
     def edit_sequence # :yields: cell, {:substitute, :insert, :delete}
       # The first time this function is called, walk backwards through
       # the backtrace to create the @path instance variable.
-      if not @path
+      if @path.nil?
         @path = [@end]
         while cell = @backtrace[@path[0]]
           @path.unshift(cell)
@@ -400,7 +404,7 @@ module EditAlign
     #    => true
     #    irb(main):002:0> a = EditAlign::Alignment.new('captained', 'caspian')
     #    => <Alignment: 5>
-    #    irb(main):003:0> puts a
+    #    irb(main):003:0> puts a.to_grid
     #                 -         c         a         p         t         a         i         n         e         d  
     #    -         [0.00]     1.00      2.00      3.00      4.00      5.00        *         *         *         *  
     #    c          1.00     [0.00]     1.00      2.00      3.00      4.00      5.00        *         *         *  

data/test/test_editalign.rb

@@ -1,7 +1,7 @@
-#!/bin/env ruby
+#!/usr/bin/env ruby
 
 #--
-# Copyright 2006 William Patrick McNeill
+# Copyright 2006-2009 William Patrick McNeill
 #
 # This file is part of Editalign.
 #

metadata

@@ -1,16 +1,16 @@
-!ruby/object:Gem::Specification 
-rubygems_version: 0.8.11
+--- !ruby/object:Gem::Specification 
+rubygems_version: 0.9.2
 specification_version: 1
 name: editalign
 version: !ruby/object:Gem::Version 
-  version: 1.0.0
-date: 2006-05-30 00:00:00 -07:00
-summary: Edit alignments between arrays
+  version: 1.1.1
+date: 2009-04-22 00:00:00 -07:00
+summary: EditAlign calculates dynamic programming alignments between arrays
 require_paths: 
 - lib
-email: billmcn@u.washington.edu
-homepage: http://staff.washington.edu/billmcn/index.shtml
-rubyforge_project: 
+email: billmcn@gmail.com
+homepage: http://editalign.rubyforge.org/
+rubyforge_project: editalign
 description: This module performs edit alignments between arrays.  It returns alignments and edit distances.
 autorequire: 
 default_executable: 
@@ -25,6 +25,7 @@ required_ruby_version: !ruby/object:Gem::Version::Requirement
 platform: ruby
 signing_key: 
 cert_chain: 
+post_install_message: 
 authors: 
 - W.P. McNeill
 files: 
@@ -36,12 +37,12 @@ files:
 test_files: 
 - test/test_editalign.rb
 rdoc_options: 
-- --title
-- EditAlign -- Ruby Edit Alignment
-- --main
-- README
-- --line-numbers
-- --inline-source
+- - --title
+  - EditAlign -- Edit Alignment
+  - --main
+  - README
+  - --line-numbers
+  - --inline-source
 extra_rdoc_files: 
 - README
 executables: []
@@ -50,13 +51,5 @@ extensions: []
 
 requirements: []
 
-dependencies: 
-- !ruby/object:Gem::Dependency 
-  name: PriorityQueue
-  version_requirement: 
-  version_requirements: !ruby/object:Gem::Version::Requirement 
-    requirements: 
-    - - ">"
-      - !ruby/object:Gem::Version 
-        version: 0.0.0
-    version: 
+dependencies: []
+