diff-lcs 1.1.1

data/lib/diff/lcs/array.rb

@@ -0,0 +1,21 @@
+#! /usr/env/bin ruby
+#--
+# Copyright 2004 Austin Ziegler <diff-lcs@halostatue.ca>
+#   adapted from:
+#     Algorithm::Diff (Perl) by Ned Konz <perl@bike-nomad.com>
+#     Smalltalk by Mario I. Wolczko <mario@wolczko.com>
+#   implements McIlroy-Hunt diff algorithm
+#
+# This program is free software. It may be redistributed and/or modified under
+# the terms of the GPL version 2 (or later), the Perl Artistic licence, or the
+# Ruby licence.
+# 
+# $Id: array.rb,v 1.3 2004/08/08 20:33:09 austin Exp $
+#++
+# Includes Diff::LCS into the Array built-in class.
+
+require 'diff/lcs'
+
+class Array
+  include Diff::LCS
+end

data/lib/diff/lcs/block.rb

@@ -0,0 +1,51 @@
+#! /usr/env/bin ruby
+#--
+# Copyright 2004 Austin Ziegler <diff-lcs@halostatue.ca>
+#   adapted from:
+#     Algorithm::Diff (Perl) by Ned Konz <perl@bike-nomad.com>
+#     Smalltalk by Mario I. Wolczko <mario@wolczko.com>
+#   implements McIlroy-Hunt diff algorithm
+#
+# This program is free software. It may be redistributed and/or modified under
+# the terms of the GPL version 2 (or later), the Perl Artistic licence, or the
+# Ruby licence.
+# 
+# $Id: block.rb,v 1.3 2004/08/08 20:33:09 austin Exp $
+#++
+# Contains Diff::LCS::Block for bin/ldiff.
+
+  # A block is an operation removing, adding, or changing a group of items.
+  # Basically, this is just a list of changes, where each change adds or
+  # deletes a single item. Used by bin/ldiff.
+class Diff::LCS::Block
+  attr_reader :changes, :insert, :remove
+
+  def initialize(chunk)
+    @changes = []
+    @insert = []
+    @remove = []
+
+    chunk.each do |item|
+      @changes << item
+      @remove << item if item.deleting?
+      @insert << item if item.adding?
+    end
+  end
+
+  def diff_size
+    @insert.size - @remove.size
+  end
+
+  def op
+    case [@remove.empty?, @insert.empty?]
+    when [false, false]
+      '!'
+    when [false, true]
+      '-'
+    when [true, false]
+      '+'
+    else # [true, true]
+      '^'
+    end
+  end
+end

data/lib/diff/lcs/callbacks.rb

@@ -0,0 +1,322 @@
+#! /usr/env/bin ruby
+#--
+# Copyright 2004 Austin Ziegler <diff-lcs@halostatue.ca>
+#   adapted from:
+#     Algorithm::Diff (Perl) by Ned Konz <perl@bike-nomad.com>
+#     Smalltalk by Mario I. Wolczko <mario@wolczko.com>
+#   implements McIlroy-Hunt diff algorithm
+#
+# This program is free software. It may be redistributed and/or modified under
+# the terms of the GPL version 2 (or later), the Perl Artistic licence, or the
+# Ruby licence.
+# 
+# $Id: callbacks.rb,v 1.4 2004/09/14 18:51:26 austin Exp $
+#++
+# Contains definitions for all default callback objects.
+
+require 'diff/lcs/change'
+
+module Diff::LCS
+    # This callback object implements the default set of callback events, which
+    # only returns the event itself. Note that #finished_a and #finished_b are
+    # not implemented -- I haven't yet figured out where they would be useful.
+    #
+    # Note that this is intended to be called as is, e.g.,
+    #
+    #     Diff::LCS.LCS(seq1, seq2, Diff::LCS::DefaultCallbacks)
+  class DefaultCallbacks
+    class << self
+        # Called when two items match.
+      def match(event)
+        event
+      end
+        # Called when the old value is discarded in favour of the new value.
+      def discard_a(event)
+        event
+      end
+        # Called when the new value is discarded in favour of the old value.
+      def discard_b(event)
+        event
+      end
+        # Called when both the old and new values have changed.
+      def change(event)
+        event
+      end
+
+      private :new
+    end
+  end
+
+    # An alias for DefaultCallbacks that is used in Diff::LCS#traverse_sequences.
+    #
+    #     Diff::LCS.LCS(seq1, seq2, Diff::LCS::SequenceCallbacks)
+  SequenceCallbacks = DefaultCallbacks
+    # An alias for DefaultCallbacks that is used in Diff::LCS#traverse_balanced.
+    #
+    #     Diff::LCS.LCS(seq1, seq2, Diff::LCS::BalancedCallbacks)
+  BalancedCallbacks = DefaultCallbacks
+end
+
+  # This will produce a compound array of simple diff change objects. Each
+  # element in the #diffs array is a +hunk+ or +hunk+ array, where each
+  # element in each +hunk+ array is a single Change object representing the
+  # addition or removal of a single element from one of the two tested
+  # sequences. The +hunk+ provides the full context for the changes.
+  #
+  #     diffs = Diff::LCS.diff(seq1, seq2)
+  #       # This example shows a simplified array format.
+  #       # [ [ [ '-',  0, 'a' ] ],   # 1
+  #       #   [ [ '+',  2, 'd' ] ],   # 2
+  #       #   [ [ '-',  4, 'h' ],     # 3
+  #       #     [ '+',  4, 'f' ] ],
+  #       #   [ [ '+',  6, 'k' ] ],   # 4
+  #       #   [ [ '-',  8, 'n' ],     # 5
+  #       #     [ '-',  9, 'p' ],
+  #       #     [ '+',  9, 'r' ],
+  #       #     [ '+', 10, 's' ],
+  #       #     [ '+', 11, 't' ] ] ]
+  #
+  # There are five hunks here. The first hunk says that the +a+ at position 0
+  # of the first sequence should be deleted (<tt>'-'</tt>). The second hunk
+  # says that the +d+ at position 2 of the second sequence should be inserted
+  # (<tt>'+'</tt>). The third hunk says that the +h+ at position 4 of the
+  # first sequence should be removed and replaced with the +f+ from position 4
+  # of the second sequence. The other two hunks are described similarly.
+  #
+  # === Use
+  # This callback object must be initialised and is used by the Diff::LCS#diff
+  # method.
+  #
+  #     cbo = Diff::LCS::DiffCallbacks.new
+  #     Diff::LCS.LCS(seq1, seq2, cbo)
+  #     cbo.finish
+  #
+  # Note that the call to #finish is absolutely necessary, or the last set of
+  # changes will not be visible. Alternatively, can be used as:
+  #
+  #     cbo = Diff::LCS::DiffCallbacks.new { |tcbo| Diff::LCS.LCS(seq1, seq2, tcbo) }
+  #
+  # The necessary #finish call will be made.
+  #
+  # === Simplified Array Format
+  # The simplified array format used in the example above can be obtained
+  # with:
+  #
+  #     require 'pp'
+  #     pp diffs.map { |e| e.map { |f| f.to_a } }
+class Diff::LCS::DiffCallbacks
+    # Returns the difference set collected during the diff process.
+  attr_reader :diffs
+
+  def initialize # :yields self:
+    @hunk = []
+    @diffs = []
+
+    if block_given?
+      begin
+        yield self
+      ensure
+        self.finish
+      end
+    end
+  end
+
+    # Finalizes the diff process. If an unprocessed hunk still exists, then it
+    # is appended to the diff list.
+  def finish
+    add_nonempty_hunk
+  end
+
+  def match(event)
+    add_nonempty_hunk
+  end
+
+  def discard_a(event)
+    @hunk << Diff::LCS::Change.new('-', event.old_position, event.old_element)
+  end
+
+  def discard_b(event)
+    @hunk << Diff::LCS::Change.new('+', event.new_position, event.new_element)
+  end
+
+private
+  def add_nonempty_hunk
+    @diffs << @hunk unless @hunk.empty?
+    @hunk = []
+  end
+end
+
+  # This will produce a compound array of contextual diff change objects. Each
+  # element in the #diffs array is a "hunk" array, where each element in each
+  # "hunk" array is a single change. Each change is a Diff::LCS::ContextChange
+  # that contains both the old index and new index values for the change. The
+  # "hunk" provides the full context for the changes. Both old and new objects
+  # will be presented for changed objects. +nil+ will be substituted for a
+  # discarded object.
+  #
+  #     seq1 = %w(a b c e h j l m n p)
+  #     seq2 = %w(b c d e f j k l m r s t)
+  #
+  #     diffs = Diff::LCS.diff(seq1, seq2, Diff::LCS::ContextDiffCallbacks)
+  #       # This example shows a simplified array format.
+  #       # [ [ [ '-', [  0, 'a' ], [  0, nil ] ] ],   # 1
+  #       #   [ [ '+', [  3, nil ], [  2, 'd' ] ] ],   # 2
+  #       #   [ [ '-', [  4, 'h' ], [  4, nil ] ],     # 3
+  #       #     [ '+', [  5, nil ], [  4, 'f' ] ] ],
+  #       #   [ [ '+', [  6, nil ], [  6, 'k' ] ] ],   # 4
+  #       #   [ [ '-', [  8, 'n' ], [  9, nil ] ],     # 5
+  #       #     [ '+', [  9, nil ], [  9, 'r' ] ],
+  #       #     [ '-', [  9, 'p' ], [ 10, nil ] ],
+  #       #     [ '+', [ 10, nil ], [ 10, 's' ] ],
+  #       #     [ '+', [ 10, nil ], [ 11, 't' ] ] ] ]
+  #
+  # The five hunks shown are comprised of individual changes; if there is a
+  # related set of changes, they are still shown individually.
+  #
+  # This callback can also be used with Diff::LCS#sdiff, which will produce
+  # results like:
+  #
+  #     diffs = Diff::LCS.sdiff(seq1, seq2, Diff::LCS::ContextCallbacks)
+  #       # This example shows a simplified array format.
+  #       # [ [ [ "-", [  0, "a" ], [  0, nil ] ] ],  # 1
+  #       #   [ [ "+", [  3, nil ], [  2, "d" ] ] ],  # 2
+  #       #   [ [ "!", [  4, "h" ], [  4, "f" ] ] ],  # 3
+  #       #   [ [ "+", [  6, nil ], [  6, "k" ] ] ],  # 4
+  #       #   [ [ "!", [  8, "n" ], [  9, "r" ] ],    # 5
+  #       #     [ "!", [  9, "p" ], [ 10, "s" ] ],
+  #       #     [ "+", [ 10, nil ], [ 11, "t" ] ] ] ]
+  #
+  # The five hunks are still present, but are significantly shorter in total
+  # presentation, because changed items are shown as changes ("!") instead of
+  # potentially "mismatched" pairs of additions and deletions.
+  #
+  # The result of this operation is similar to that of
+  # Diff::LCS::SDiffCallbacks. They may be compared as:
+  #
+  #     s = Diff::LCS.sdiff(seq1, seq2).reject { |e| e.action == "=" }
+  #     c = Diff::LCS.sdiff(seq1, seq2, Diff::LCS::ContextDiffCallbacks).flatten
+  #
+  #     s == c # -> true
+  #
+  # === Use
+  # This callback object must be initialised and can be used by the
+  # Diff::LCS#diff or Diff::LCS#sdiff methods.
+  #
+  #     cbo = Diff::LCS::ContextDiffCallbacks.new
+  #     Diff::LCS.LCS(seq1, seq2, cbo)
+  #     cbo.finish
+  #
+  # Note that the call to #finish is absolutely necessary, or the last set of
+  # changes will not be visible. Alternatively, can be used as:
+  #
+  #     cbo = Diff::LCS::ContextDiffCallbacks.new { |tcbo| Diff::LCS.LCS(seq1, seq2, tcbo) }
+  #
+  # The necessary #finish call will be made.
+  #
+  # === Simplified Array Format
+  # The simplified array format used in the example above can be obtained
+  # with:
+  #
+  #     require 'pp'
+  #     pp diffs.map { |e| e.map { |f| f.to_a } }
+class Diff::LCS::ContextDiffCallbacks < Diff::LCS::DiffCallbacks
+  def discard_a(event)
+    @hunk << Diff::LCS::ContextChange.simplify(event)
+  end
+
+  def discard_b(event)
+    @hunk << Diff::LCS::ContextChange.simplify(event)
+  end
+
+  def change(event)
+    @hunk << Diff::LCS::ContextChange.simplify(event)
+  end
+end
+
+  # This will produce a simple array of diff change objects. Each element in
+  # the #diffs array is a single ContextChange. In the set of #diffs provided
+  # by SDiffCallbacks, both old and new objects will be presented for both
+  # changed <strong>and unchanged</strong> objects. +nil+ will be substituted
+  # for a discarded object.
+  #
+  # The diffset produced by this callback, when provided to Diff::LCS#sdiff,
+  # will compute and display the necessary components to show two sequences
+  # and their minimized differences side by side, just like the Unix utility
+  # +sdiff+.
+  # 
+  #     same             same
+  #     before     |     after
+  #     old        <     -
+  #     -          >     new
+  #
+  #     seq1 = %w(a b c e h j l m n p)
+  #     seq2 = %w(b c d e f j k l m r s t)
+  #
+  #     diffs = Diff::LCS.sdiff(seq1, seq2)
+  #       # This example shows a simplified array format.
+  #       # [ [ "-", [  0, "a"], [  0, nil ] ],
+  #       #   [ "=", [  1, "b"], [  0, "b" ] ],
+  #       #   [ "=", [  2, "c"], [  1, "c" ] ],
+  #       #   [ "+", [  3, nil], [  2, "d" ] ],
+  #       #   [ "=", [  3, "e"], [  3, "e" ] ],
+  #       #   [ "!", [  4, "h"], [  4, "f" ] ],
+  #       #   [ "=", [  5, "j"], [  5, "j" ] ],
+  #       #   [ "+", [  6, nil], [  6, "k" ] ],
+  #       #   [ "=", [  6, "l"], [  7, "l" ] ],
+  #       #   [ "=", [  7, "m"], [  8, "m" ] ],
+  #       #   [ "!", [  8, "n"], [  9, "r" ] ],
+  #       #   [ "!", [  9, "p"], [ 10, "s" ] ],
+  #       #   [ "+", [ 10, nil], [ 11, "t" ] ] ]
+  #
+  # The result of this operation is similar to that of
+  # Diff::LCS::ContextDiffCallbacks. They may be compared as:
+  #
+  #     s = Diff::LCS.sdiff(seq1, seq2).reject { |e| e.action == "=" }
+  #     c = Diff::LCS.sdiff(seq1, seq2, Diff::LCS::ContextDiffCallbacks).flatten
+  #
+  #     s == c # -> true
+  #
+  # === Use
+  # This callback object must be initialised and is used by the Diff::LCS#sdiff
+  # method.
+  #
+  #     cbo = Diff::LCS::SDiffCallbacks.new
+  #     Diff::LCS.LCS(seq1, seq2, cbo)
+  #
+  # As with the other initialisable callback objects, Diff::LCS::SDiffCallbacks
+  # can be initialised with a block. As there is no "fininishing" to be done,
+  # this has no effect on the state of the object.
+  #
+  #     cbo = Diff::LCS::SDiffCallbacks.new { |tcbo| Diff::LCS.LCS(seq1, seq2, tcbo) }
+  #
+  # === Simplified Array Format
+  # The simplified array format used in the example above can be obtained
+  # with:
+  #
+  #     require 'pp'
+  #     pp diffs.map { |e| e.to_a }
+class Diff::LCS::SDiffCallbacks
+    # Returns the difference set collected during the diff process.
+  attr_reader :diffs
+
+  def initialize #:yields self:
+    @diffs = []
+    yield self if block_given?
+  end
+
+  def match(event)
+    @diffs << Diff::LCS::ContextChange.simplify(event)
+  end
+
+  def discard_a(event)
+    @diffs << Diff::LCS::ContextChange.simplify(event)
+  end
+
+  def discard_b(event)
+    @diffs << Diff::LCS::ContextChange.simplify(event)
+  end
+
+  def change(event)
+    @diffs << Diff::LCS::ContextChange.simplify(event)
+  end
+end

data/lib/diff/lcs/change.rb

@@ -0,0 +1,169 @@
+#! /usr/env/bin ruby
+#--
+# Copyright 2004 Austin Ziegler <diff-lcs@halostatue.ca>
+#   adapted from:
+#     Algorithm::Diff (Perl) by Ned Konz <perl@bike-nomad.com>
+#     Smalltalk by Mario I. Wolczko <mario@wolczko.com>
+#   implements McIlroy-Hunt diff algorithm
+#
+# This program is free software. It may be redistributed and/or modified under
+# the terms of the GPL version 2 (or later), the Perl Artistic licence, or the
+# Ruby licence.
+# 
+# $Id: change.rb,v 1.4 2004/08/08 20:33:09 austin Exp $
+#++
+# Provides Diff::LCS::Change and Diff::LCS::ContextChange.
+
+  # Centralises the change test code in Diff::LCS::Change and
+  # Diff::LCS::ContextChange, since it's the same for both classes.
+module Diff::LCS::ChangeTypeTests
+  def deleting?
+    @action == '-'
+  end
+
+  def adding?
+    @action == '+'
+  end
+
+  def unchanged?
+    @action == '='
+  end
+
+  def changed?
+    @changed == '!'
+  end
+
+  def finished_a?
+    @changed == '>'
+  end
+
+  def finished_b?
+    @changed == '<'
+  end
+end
+
+  # Represents a simplistic (non-contextual) change. Represents the removal or
+  # addition of an element from either the old or the new sequenced enumerable.
+class Diff::LCS::Change
+    # Returns the action this Change represents. Can be '+' (#adding?), '-'
+    # (#deleting?), '=' (#unchanged?), # or '!' (#changed?). When created by
+    # Diff::LCS#diff or Diff::LCS#sdiff, it may also be '>' (#finished_a?) or
+    # '<' (#finished_b?).
+  attr_reader :action
+  attr_reader :position
+  attr_reader :element
+
+  include Comparable
+  def ==(other)
+    (self.action == other.action) and
+    (self.position == other.position) and
+    (self.element == other.element)
+  end
+
+  def <=>(other)
+    r = self.action <=> other.action
+    r = self.position <=> other.position if r.zero?
+    r = self.element <=> other.element if r.zero?
+    r
+  end
+
+  def initialize(action, position, element)
+    @action = action
+    @position = position
+    @element = element
+  end
+
+    # Creates a Change from an array produced by Change#to_a.
+  def to_a
+    [@action, @position, @element]
+  end
+
+  def self.from_a(arr)
+    Diff::LCS::Change.new(arr[0], arr[1], arr[2])
+  end
+
+  include Diff::LCS::ChangeTypeTests
+end
+
+  # Represents a contextual change. Contains the position and values of the
+  # elements in the old and the new sequenced enumerables as well as the action
+  # taken.
+class Diff::LCS::ContextChange
+    # Returns the action this Change represents. Can be '+' (#adding?), '-'
+    # (#deleting?), '=' (#unchanged?), # or '!' (#changed?). When
+    # created by Diff::LCS#diff or Diff::LCS#sdiff, it may also be '>'
+    # (#finished_a?) or '<' (#finished_b?).
+  attr_reader :action
+  attr_reader :old_position
+  attr_reader :old_element
+  attr_reader :new_position
+  attr_reader :new_element
+
+  include Comparable
+
+  def ==(other)
+    (@action == other.action) and
+    (@old_position == other.old_position) and
+    (@new_position == other.new_position) and
+    (@old_element == other.old_element) and
+    (@new_element == other.new_element)
+  end
+
+  def inspect(*args)
+    %Q(#<#{self.class.name}:#{__id__} @action=#{action} positions=#{old_position},#{new_position} elements=#{old_element.inspect},#{new_element.inspect}>)
+  end
+
+  def <=>(other)
+    r = @action <=> other.action
+    r = @old_position <=> other.old_position if r.zero?
+    r = @new_position <=> other.new_position if r.zero?
+    r = @old_element <=> other.old_element if r.zero?
+    r = @new_element <=> other.new_element if r.zero?
+    r
+  end
+
+  def initialize(action, old_position, old_element, new_position, new_element)
+    @action = action
+    @old_position = old_position
+    @old_element = old_element
+    @new_position = new_position
+    @new_element = new_element
+  end
+
+  def to_a
+    [@action, [@old_position, @old_element], [@new_position, @new_element]]
+  end
+
+    # Creates a ContextChange from an array produced by ContextChange#to_a.
+  def self.from_a(arr)
+    if arr.size == 5
+      Diff::LCS::ContextChange.new(arr[0], arr[1], arr[2], arr[3], arr[4])
+    else
+      Diff::LCS::ContextChange.new(arr[0], arr[1][0], arr[1][1], arr[2][0],
+                                   arr[2][1])
+    end
+  end
+
+    # Simplifies a context change for use in some diff callbacks. '<' actions
+    # are converted to '-' and '>' actions are converted to '+'. 
+  def self.simplify(event)
+    ea = event.to_a
+
+    case ea[0]
+    when '-'
+      ea[2][1] = nil
+    when '<'
+      ea[0] = '-'
+      ea[2][1] = nil
+    when '+'
+      ea[1][1] = nil
+    when '>'
+      ea[0] = '+'
+      ea[1][1] = nil
+    end
+
+    Diff::LCS::ContextChange.from_a(ea)
+  end
+
+  include Diff::LCS::ChangeTypeTests
+end