diff-lcs 1.3 → 1.5.0

data/lib/diff/lcs/hunk.rb

@@ -1,30 +1,43 @@
-# -*- ruby encoding: utf-8 -*-
+# frozen_string_literal: true
 
 require 'diff/lcs/block'
 
-# A Hunk is a group of Blocks which overlap because of the context
-# surrounding each block. (So if we're not using context, every hunk will
-# contain one block.) Used in the diff program (bin/diff).
+# A Hunk is a group of Blocks which overlap because of the context surrounding
+# each block. (So if we're not using context, every hunk will contain one
+# block.) Used in the diff program (bin/ldiff).
 class Diff::LCS::Hunk
-  # Create a hunk using references to both the old and new data, as well as
-  # the piece of data.
+  OLD_DIFF_OP_ACTION = { '+' => 'a', '-' => 'd', '!' => 'c' }.freeze #:nodoc:
+  ED_DIFF_OP_ACTION = { '+' => 'a', '-' => 'd', '!' => 'c' }.freeze #:nodoc:
+
+  private_constant :OLD_DIFF_OP_ACTION, :ED_DIFF_OP_ACTION if respond_to?(:private_constant)
+
+  # Create a hunk using references to both the old and new data, as well as the
+  # piece of data.
   def initialize(data_old, data_new, piece, flag_context, file_length_difference)
     # At first, a hunk will have just one Block in it
-    @blocks = [ Diff::LCS::Block.new(piece) ]
+    @blocks = [Diff::LCS::Block.new(piece)]
+
+    if @blocks[0].remove.empty? && @blocks[0].insert.empty?
+      fail "Cannot build a hunk from #{piece.inspect}; has no add or remove actions"
+    end
+
     if String.method_defined?(:encoding)
-      @preferred_data_encoding = data_old.fetch(0, data_new.fetch(0,'') ).encoding
+      @preferred_data_encoding = data_old.fetch(0, data_new.fetch(0, '')).encoding
     end
+
     @data_old = data_old
     @data_new = data_new
 
     before = after = file_length_difference
     after += @blocks[0].diff_size
     @file_length_difference = after # The caller must get this manually
+    @max_diff_size = @blocks.map { |e| e.diff_size.abs }.max
+
 
     # Save the start & end of each array. If the array doesn't exist (e.g.,
-    # we're only adding items in this block), then figure out the line
-    # number based on the line number of the other file and the current
-    # difference in file lengths.
+    # we're only adding items in this block), then figure out the line number
+    # based on the line number of the other file and the current difference in
+    # file lengths.
     if @blocks[0].remove.empty?
       a1 = a2 = nil
     else
@@ -54,20 +67,27 @@ class Diff::LCS::Hunk
 
   # Change the "start" and "end" fields to note that context should be added
   # to this hunk.
-  attr_accessor :flag_context
-  undef :flag_context=;
-  def flag_context=(context) #:nodoc:
+  attr_accessor :flag_context # rubocop:disable Layout/EmptyLinesAroundAttributeAccessor
+  undef :flag_context=
+  def flag_context=(context) #:nodoc: # rubocop:disable Lint/DuplicateMethods
     return if context.nil? or context.zero?
 
-    add_start = (context > @start_old) ? @start_old : context
+    add_start = context > @start_old ? @start_old : context
+
     @start_old -= add_start
     @start_new -= add_start
 
-    if (@end_old + context) > @data_old.size
-      add_end = @data_old.size - @end_old
-    else
-      add_end = context
-    end
+    old_size = @data_old.size
+
+    add_end =
+      if (@end_old + context) > old_size
+        old_size - @end_old
+      else
+        context
+      end
+
+    add_end = @max_diff_size if add_end >= old_size
+
     @end_old += add_end
     @end_new += add_end
   end
@@ -76,15 +96,13 @@ class Diff::LCS::Hunk
   # a truthy value so that if there is no overlap, you can know the merge
   # was skipped.
   def merge(hunk)
-    if overlaps?(hunk)
-      @start_old = hunk.start_old
-      @start_new = hunk.start_new
-      blocks.unshift(*hunk.blocks)
-    else
-      nil
-    end
+    return unless overlaps?(hunk)
+
+    @start_old = hunk.start_old
+    @start_new = hunk.start_new
+    blocks.unshift(*hunk.blocks)
   end
-  alias_method :unshift, :merge
+  alias unshift merge
 
   # Determines whether there is an overlap between this hunk and the
   # provided hunk. This will be true if the difference between the two hunks
@@ -95,47 +113,53 @@ class Diff::LCS::Hunk
   end
 
   # Returns a diff string based on a format.
-  def diff(format)
+  def diff(format, last = false)
     case format
     when :old
-      old_diff
+      old_diff(last)
     when :unified
-      unified_diff
+      unified_diff(last)
     when :context
-      context_diff
+      context_diff(last)
     when :ed
       self
     when :reverse_ed, :ed_finish
-      ed_diff(format)
+      ed_diff(format, last)
     else
-      raise "Unknown diff format #{format}."
+      fail "Unknown diff format #{format}."
     end
   end
 
   # Note that an old diff can't have any context. Therefore, we know that
   # there's only one block in the hunk.
-  def old_diff
-    warn "Expecting only one block in an old diff hunk!" if @blocks.size > 1
-    op_act = { "+" => 'a', "-" => 'd', "!" => "c" }
+  def old_diff(_last = false)
+    warn 'Expecting only one block in an old diff hunk!' if @blocks.size > 1
 
     block = @blocks[0]
 
     # Calculate item number range. Old diff range is just like a context
     # diff range, except the ranges are on one line with the action between
     # them.
-    s = encode("#{context_range(:old)}#{op_act[block.op]}#{context_range(:new)}\n")
+    s = encode("#{context_range(:old, ',')}#{OLD_DIFF_OP_ACTION[block.op]}#{context_range(:new, ',')}\n")
     # If removing anything, just print out all the remove lines in the hunk
     # which is just all the remove lines in the block.
-    @data_old[@start_old .. @end_old].each { |e| s << encode("< ") + e + encode("\n") } unless block.remove.empty?
-    s << encode("---\n") if block.op == "!"
-    @data_new[@start_new .. @end_new].each { |e| s << encode("> ") + e + encode("\n") } unless block.insert.empty?
+    unless block.remove.empty?
+      @data_old[@start_old..@end_old].each { |e| s << encode('< ') + e.chomp + encode("\n") }
+    end
+
+    s << encode("---\n") if block.op == '!'
+
+    unless block.insert.empty?
+      @data_new[@start_new..@end_new].each { |e| s << encode('> ') + e.chomp + encode("\n") }
+    end
+
     s
   end
   private :old_diff
 
-  def unified_diff
+  def unified_diff(last = false)
     # Calculate item number range.
-    s = encode("@@ -#{unified_range(:old)} +#{unified_range(:new)} @@\n")
+    s = encode("@@ -#{unified_range(:old, last)} +#{unified_range(:new, last)} @@\n")
 
     # Outlist starts containing the hunk of the old file. Removing an item
     # just means putting a '-' in front of it. Inserting an item requires
@@ -148,7 +172,14 @@ class Diff::LCS::Hunk
     # file -- don't take removed items into account.
     lo, hi, num_added, num_removed = @start_old, @end_old, 0, 0
 
-    outlist = @data_old[lo .. hi].map { |e| e.insert(0, encode(' ')) }
+    outlist = @data_old[lo..hi].map { |e| String.new("#{encode(' ')}#{e.chomp}") }
+
+    last_block = blocks[-1]
+
+    if last
+      old_missing_newline = missing_last_newline?(@data_old)
+      new_missing_newline = missing_last_newline?(@data_new)
+    end
 
     @blocks.each do |block|
       block.remove.each do |item|
@@ -157,66 +188,100 @@ class Diff::LCS::Hunk
         outlist[offset][0, 1] = encode(op)
         num_removed += 1
       end
+
+      if last && block == last_block && old_missing_newline && !new_missing_newline
+        outlist << encode('\')
+        num_removed += 1
+      end
+
       block.insert.each do |item|
         op     = item.action.to_s # +
         offset = item.position - @start_new + num_removed
-        outlist[offset, 0] = encode(op) + @data_new[item.position]
+        outlist[offset, 0] = encode(op) + @data_new[item.position].chomp
         num_added += 1
       end
     end
 
+    outlist << encode('\') if last && new_missing_newline
+
     s << outlist.join(encode("\n"))
+
+    s
   end
   private :unified_diff
 
-  def context_diff
+  def context_diff(last = false)
     s = encode("***************\n")
-    s << encode("*** #{context_range(:old)} ****\n")
-    r = context_range(:new)
+    s << encode("*** #{context_range(:old, ',', last)} ****\n")
+    r = context_range(:new, ',', last)
+
+    if last
+      old_missing_newline = missing_last_newline?(@data_old)
+      new_missing_newline = missing_last_newline?(@data_new)
+    end
 
     # Print out file 1 part for each block in context diff format if there
     # are any blocks that remove items
     lo, hi = @start_old, @end_old
-    removes = @blocks.select { |e| not e.remove.empty? }
-    if removes
-      outlist = @data_old[lo .. hi].map { |e| e.insert(0, encode(' ')) }
+    removes = @blocks.reject { |e| e.remove.empty? }
+
+    unless removes.empty?
+      outlist = @data_old[lo..hi].map { |e| String.new("#{encode('  ')}#{e.chomp}") }
+
+      last_block = removes[-1]
 
       removes.each do |block|
         block.remove.each do |item|
           outlist[item.position - lo][0, 1] = encode(block.op) # - or !
         end
+
+        if last && block == last_block && old_missing_newline
+          outlist << encode('\')
+        end
       end
-      s << outlist.join("\n")
+
+      s << outlist.join(encode("\n")) << encode("\n")
     end
 
-    s << encode("\n--- #{r} ----\n")
+    s << encode("--- #{r} ----\n")
     lo, hi = @start_new, @end_new
-    inserts = @blocks.select { |e| not e.insert.empty? }
-    if inserts
-      outlist = @data_new[lo .. hi].collect { |e| e.insert(0, encode(' ')) }
+    inserts = @blocks.reject { |e| e.insert.empty? }
+
+    unless inserts.empty?
+      outlist = @data_new[lo..hi].map { |e| String.new("#{encode('  ')}#{e.chomp}") }
+
+      last_block = inserts[-1]
+
       inserts.each do |block|
         block.insert.each do |item|
           outlist[item.position - lo][0, 1] = encode(block.op) # + or !
         end
+
+        if last && block == last_block && new_missing_newline
+          outlist << encode('\')
+        end
       end
-      s << outlist.join("\n")
+      s << outlist.join(encode("\n"))
     end
+
     s
   end
   private :context_diff
 
-  def ed_diff(format)
-    op_act = { "+" => 'a', "-" => 'd', "!" => "c" }
-    warn "Expecting only one block in an old diff hunk!" if @blocks.size > 1
+  def ed_diff(format, _last = false)
+    warn 'Expecting only one block in an old diff hunk!' if @blocks.size > 1
 
-    if format == :reverse_ed
-      s = encode("#{op_act[@blocks[0].op]}#{context_range(:old)}\n")
-    else
-      s = encode("#{context_range(:old, ' ')}#{op_act[@blocks[0].op]}\n")
-    end
+    s =
+      if format == :reverse_ed
+        encode("#{ED_DIFF_OP_ACTION[@blocks[0].op]}#{context_range(:old, ',')}\n")
+      else
+        encode("#{context_range(:old, ' ')}#{ED_DIFF_OP_ACTION[@blocks[0].op]}\n")
+      end
 
     unless @blocks[0].insert.empty?
-      @data_new[@start_new .. @end_new].each { |e| s << e + encode("\n") }
+      @data_new[@start_new..@end_new].each do |e|
+        s << e.chomp + encode("\n")
+      end
       s << encode(".\n")
     end
     s
@@ -225,7 +290,7 @@ class Diff::LCS::Hunk
 
   # Generate a range of item numbers to print. Only print 1 number if the
   # range has only one item in it. Otherwise, it's 'start,end'
-  def context_range(mode, op = ',')
+  def context_range(mode, op, last = false)
     case mode
     when :old
       s, e = (@start_old + 1), (@end_old + 1)
@@ -233,14 +298,17 @@ class Diff::LCS::Hunk
       s, e = (@start_new + 1), (@end_new + 1)
     end
 
-    (s < e) ? "#{s}#{op}#{e}" : "#{e}"
+    e -= 1 if last
+    e = 1 if e.zero?
+
+    s < e ? "#{s}#{op}#{e}" : e.to_s
   end
   private :context_range
 
   # Generate a range of item numbers to print for unified diff. Print number
   # where block starts, followed by number of lines in the block
   # (don't print number of lines if it's 1)
-  def unified_range(mode)
+  def unified_range(mode, last)
     case mode
     when :old
       s, e = (@start_old + 1), (@end_old + 1)
@@ -248,12 +316,25 @@ class Diff::LCS::Hunk
       s, e = (@start_new + 1), (@end_new + 1)
     end
 
-    length = e - s + 1
-    first = (length < 2) ? e : s # "strange, but correct"
-    (length == 1) ? "#{first}" : "#{first},#{length}"
+    length = e - s + (last ? 0 : 1)
+
+    first = length < 2 ? e : s # "strange, but correct"
+    length <= 1 ? first.to_s : "#{first},#{length}"
   end
   private :unified_range
 
+  def missing_last_newline?(data)
+    newline = encode("\n")
+
+    if data[-2]
+      data[-2].end_with?(newline) && !data[-1].end_with?(newline)
+    elsif data[-1]
+      !data[-1].end_with?(newline)
+    else
+      true
+    end
+  end
+
   if String.method_defined?(:encoding)
     def encode(literal, target_encoding = @preferred_data_encoding)
       literal.encode target_encoding
@@ -263,10 +344,11 @@ class Diff::LCS::Hunk
       args.map { |arg| arg.encode(string.encoding) }
     end
   else
-    def encode(literal, target_encoding = nil)
+    def encode(literal, _target_encoding = nil)
       literal
     end
-    def encode_as(string, *args)
+
+    def encode_as(_string, *args)
       args
     end
   end

data/lib/diff/lcs/internals.rb

@@ -1,4 +1,4 @@
-# -*- ruby encoding: utf-8 -*-
+# frozen_string_literal: true
 
 class << Diff::LCS
   def diff_traversal(method, seq1, seq2, callbacks, &block)
@@ -44,47 +44,49 @@ class << Diff::LCS::Internals
     b_finish = b.size - 1
     vector = []
 
-    # Prune off any common elements at the beginning...
-    while ((a_start <= a_finish) and (b_start <= b_finish) and
-           (a[a_start] == b[b_start]))
+    # Collect any common elements at the beginning...
+    while (a_start <= a_finish) and (b_start <= b_finish) and (a[a_start] == b[b_start])
       vector[a_start] = b_start
       a_start += 1
       b_start += 1
     end
-    b_start = a_start
 
     # Now the end...
-    while ((a_start <= a_finish) and (b_start <= b_finish) and
-           (a[a_finish] == b[b_finish]))
+    while (a_start <= a_finish) and (b_start <= b_finish) and (a[a_finish] == b[b_finish])
       vector[a_finish] = b_finish
       a_finish -= 1
       b_finish -= 1
     end
 
     # Now, compute the equivalence classes of positions of elements.
+    # An explanation for how this works: https://codeforces.com/topic/92191
     b_matches = position_hash(b, b_start..b_finish)
 
     thresh = []
     links  = []
     string = a.kind_of?(String)
 
-    (a_start .. a_finish).each do |i|
+    (a_start..a_finish).each do |i|
       ai = string ? a[i, 1] : a[i]
       bm = b_matches[ai]
       k = nil
       bm.reverse_each do |j|
+        # Although the threshold check is not mandatory for this to work,
+        # it may have an optimization purpose
+        # An attempt to remove it: https://github.com/halostatue/diff-lcs/pull/72
+        # Why it is reintroduced: https://github.com/halostatue/diff-lcs/issues/78
         if k and (thresh[k] > j) and (thresh[k - 1] < j)
           thresh[k] = j
         else
           k = replace_next_larger(thresh, j, k)
         end
-        links[k] = [ (k > 0) ? links[k - 1] : nil, i, j ] unless k.nil?
+        links[k] = [k.positive? ? links[k - 1] : nil, i, j] unless k.nil?
       end
     end
 
     unless thresh.empty?
       link = links[thresh.size - 1]
-      while not link.nil?
+      until link.nil?
         vector[link[1]] = link[2]
         link = link[0]
       end
@@ -93,14 +95,15 @@ class << Diff::LCS::Internals
     vector
   end
 
-  # This method will analyze the provided patchset to provide a
-  # single-pass normalization (conversion of the array form of
-  # Diff::LCS::Change objects to the object form of same) and detection of
-  # whether the patchset represents changes to be made.
+  # This method will analyze the provided patchset to provide a single-pass
+  # normalization (conversion of the array form of Diff::LCS::Change objects to
+  # the object form of same) and detection of whether the patchset represents
+  # changes to be made.
   def analyze_patchset(patchset, depth = 0)
-    raise "Patchset too complex" if depth > 1
+    fail 'Patchset too complex' if depth > 1
 
     has_changes = false
+    new_patchset = []
 
     # Format:
     # [ # patchset
@@ -110,29 +113,28 @@ class << Diff::LCS::Internals
     #   ]
     # ]
 
-    patchset = patchset.map do |hunk|
+    patchset.each do |hunk|
       case hunk
       when Diff::LCS::Change
         has_changes ||= !hunk.unchanged?
-        hunk
+        new_patchset << hunk
       when Array
-        # Detect if the 'hunk' is actually an array-format
-        # Change object.
+        # Detect if the 'hunk' is actually an array-format change object.
         if Diff::LCS::Change.valid_action? hunk[0]
           hunk = Diff::LCS::Change.from_a(hunk)
           has_changes ||= !hunk.unchanged?
-          hunk
+          new_patchset << hunk
         else
           with_changes, hunk = analyze_patchset(hunk, depth + 1)
           has_changes ||= with_changes
-          hunk.flatten
+          new_patchset.concat(hunk)
         end
       else
-        raise ArgumentError, "Cannot normalise a hunk of class #{hunk.class}."
+        fail ArgumentError, "Cannot normalise a hunk of class #{hunk.class}."
       end
     end
 
-    [ has_changes, patchset.flatten(1) ]
+    [has_changes, new_patchset]
   end
 
   # Examine the patchset and the source to see in which direction the
@@ -173,13 +175,11 @@ class << Diff::LCS::Internals
         when '!'
           if le == change.old_element
             left_match += 1
+          elsif re == change.new_element
+            right_match += 1
           else
-            if re == change.new_element
-              right_match += 1
-            else
-              left_miss += 1
-              right_miss += 1
-            end
+            left_miss += 1
+            right_miss += 1
           end
         end
       when Diff::LCS::Change
@@ -209,16 +209,16 @@ class << Diff::LCS::Internals
         end
       end
 
-      break if (not limit.nil?) && (count > limit)
+      break if !limit.nil? && (count > limit)
     end
 
-    no_left = (left_match == 0) && (left_miss > 0)
-    no_right = (right_match == 0) && (right_miss > 0)
+    no_left = left_match.zero? && left_miss.positive?
+    no_right = right_match.zero? && right_miss.positive?
 
-    case [ no_left, no_right ]
-    when [ false, true ]
+    case [no_left, no_right]
+    when [false, true]
       :patch
-    when [ true, false ]
+    when [true, false]
       :unpatch
     else
       case left_match <=> right_match
@@ -235,7 +235,8 @@ class << Diff::LCS::Internals
           :patch
         end
       else
-        raise "The provided patchset does not appear to apply to the provided enumerable as either source or destination value."
+        fail "The provided patchset does not appear to apply to the provided \
+enumerable as either source or destination value."
       end
     end
   end
@@ -256,16 +257,16 @@ class << Diff::LCS::Internals
     end
 
     # Binary search for the insertion point
-    last_index ||= enum.size
+    last_index ||= enum.size - 1
     first_index = 0
-    while (first_index <= last_index)
+    while first_index <= last_index
       i = (first_index + last_index) >> 1
 
       found = enum[i]
 
-      if value == found
-        return nil
-      elsif value > found
+      return nil if value == found
+
+      if value > found
         first_index = i + 1
       else
         last_index = i - 1
@@ -275,7 +276,7 @@ class << Diff::LCS::Internals
     # The insertion point is in first_index; overwrite the next larger
     # value.
     enum[first_index] = value
-    return first_index
+    first_index
   end
   private :replace_next_larger