diff-lcs 1.5.1 → 1.6.1

data/lib/diff/lcs/hunk.rb

@@ -22,11 +22,13 @@ class Diff::LCS::Hunk
     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
+    @old_empty = data_old.empty? || (data_old.size == 1 && data_old[0].empty?)
+    @new_empty = data_new.empty? || (data_new.size == 1 && data_new[0].empty?)
 
     before = after = file_length_difference
     after += @blocks[0].diff_size
@@ -79,14 +81,12 @@ class Diff::LCS::Hunk
     old_size = @data_old.size
 
     add_end =
-      if (@end_old + context) > old_size
-        old_size - @end_old
+      if (@end_old + context) >= old_size
+        old_size - @end_old - 1
       else
         context
       end
 
-    add_end = @max_diff_size if add_end >= old_size
-
     @end_old += add_end
     @end_new += add_end
   end
@@ -131,11 +131,16 @@ class Diff::LCS::Hunk
 
   # 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(_last = false)
+  def old_diff(last = false)
     warn "Expecting only one block in an old diff hunk!" if @blocks.size > 1
 
     block = @blocks[0]
 
+    if last
+      old_missing_newline = !@old_empty && missing_last_newline?(@data_old)
+      new_missing_newline = !@new_empty && missing_last_newline?(@data_new)
+    end
+
     # 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.
@@ -146,19 +151,22 @@ class Diff::LCS::Hunk
       @data_old[@start_old..@end_old].each { |e| s << encode("< ") + e.chomp + encode("\n") }
     end
 
+    s << encode("\\n") if old_missing_newline && !new_missing_newline
     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 << encode("\\n") if new_missing_newline && !old_missing_newline
+
     s
   end
   private :old_diff
 
   def unified_diff(last = false)
     # Calculate item number range.
-    s = encode("@@ -#{unified_range(:old, last)} +#{unified_range(:new, last)} @@\n")
+    s = encode("@@ -#{unified_range(:old)} +#{unified_range(:new)} @@\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
@@ -178,8 +186,8 @@ class Diff::LCS::Hunk
     last_block = blocks[-1]
 
     if last
-      old_missing_newline = missing_last_newline?(@data_old)
-      new_missing_newline = missing_last_newline?(@data_new)
+      old_missing_newline = !@old_empty && missing_last_newline?(@data_old)
+      new_missing_newline = !@new_empty && missing_last_newline?(@data_new)
     end
 
     @blocks.each do |block|
@@ -213,8 +221,8 @@ class Diff::LCS::Hunk
 
   def context_diff(last = false)
     s = encode("***************\n")
-    s << encode("*** #{context_range(:old, ",", last)} ****\n")
-    r = context_range(:new, ",", last)
+    s << encode("*** #{context_range(:old, ",")} ****\n")
+    r = context_range(:new, ",")
 
     if last
       old_missing_newline = missing_last_newline?(@data_old)
@@ -273,14 +281,24 @@ class Diff::LCS::Hunk
   end
   private :context_diff
 
-  def ed_diff(format, _last = false)
+  def ed_diff(format, last)
     warn "Expecting only one block in an old diff hunk!" if @blocks.size > 1
+    if last
+      # ed script doesn't support well incomplete lines
+      warn "<old_file>: No newline at end of file\n" if !@old_empty && missing_last_newline?(@data_old)
+      warn "<new_file>: No newline at end of file\n" if !@new_empty && missing_last_newline?(@data_new)
+
+      if @blocks[0].op == "!"
+        return +"" if @blocks[0].changes[0].element == @blocks[0].changes[1].element + "\n"
+        return +"" if @blocks[0].changes[0].element + "\n" == @blocks[0].changes[1].element
+      end
+    end
 
     s =
       if format == :reverse_ed
-        encode("#{ED_DIFF_OP_ACTION[@blocks[0].op]}#{context_range(:old, ",")}\n")
+        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")
+        encode("#{context_range(:old, ",")}#{ED_DIFF_OP_ACTION[@blocks[0].op]}\n")
       end
 
     unless @blocks[0].insert.empty?
@@ -295,7 +313,7 @@ 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, last = false)
+  def context_range(mode, op)
     case mode
     when :old
       s, e = (@start_old + 1), (@end_old + 1)
@@ -303,9 +321,6 @@ class Diff::LCS::Hunk
       s, e = (@start_new + 1), (@end_new + 1)
     end
 
-    e -= 1 if last
-    e = 1 if e.zero?
-
     (s < e) ? "#{s}#{op}#{e}" : e.to_s
   end
   private :context_range
@@ -313,18 +328,19 @@ class Diff::LCS::Hunk
   # 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, last)
+  def unified_range(mode)
     case mode
     when :old
+      return "0,0" if @old_empty
       s, e = (@start_old + 1), (@end_old + 1)
     when :new
+      return "0,0" if @new_empty
       s, e = (@start_new + 1), (@end_new + 1)
     end
 
-    length = e - s + (last ? 0 : 1)
+    length = e - s + 1
 
-    first = (length < 2) ? e : s # "strange, but correct"
-    (length <= 1) ? first.to_s : "#{first},#{length}"
+    (length <= 1) ? e.to_s : "#{s},#{length}"
   end
   private :unified_range
 

data/lib/diff/lcs/ldiff.rb

@@ -1,14 +1,13 @@
 # frozen_string_literal: true
 
 require "optparse"
-require "ostruct"
 require "diff/lcs/hunk"
 
-module Diff::LCS::Ldiff # :nodoc:
+class Diff::LCS::Ldiff # :nodoc:
   # standard:disable Layout/HeredocIndentation
   BANNER = <<-COPYRIGHT
 ldiff #{Diff::LCS::VERSION}
-  Copyright 2004-2019 Austin Ziegler
+  Copyright 2004-2025 Austin Ziegler
 
   Part of Diff::LCS.
   https://github.com/halostatue/diff-lcs
@@ -18,16 +17,28 @@ ldiff #{Diff::LCS::VERSION}
   MIT licence.
   COPYRIGHT
   # standard:enable Layout/HeredocIndentation
-end
 
-class << Diff::LCS::Ldiff
+  InputInfo = Struct.new(:filename, :data, :stat) do
+    def initialize(filename)
+      super(filename, ::File.read(filename), ::File.stat(filename))
+    end
+  end
+
   attr_reader :format, :lines # :nodoc:
   attr_reader :file_old, :file_new # :nodoc:
   attr_reader :data_old, :data_new # :nodoc:
 
-  def run(args, _input = $stdin, output = $stdout, error = $stderr) # :nodoc:
+  def self.run(args, input = $stdin, output = $stdout, error = $stderr) # :nodoc:
+    new.run(args, input, output, error)
+  end
+
+  def initialize
     @binary = nil
+    @format = :old
+    @lines = 0
+  end
 
+  def run(args, _input = $stdin, output = $stdout, error = $stderr) # :nodoc:
     args.options do |o|
       o.banner = "Usage: #{File.basename($0)} [options] oldfile newfile"
       o.separator ""
@@ -86,8 +97,18 @@ class << Diff::LCS::Ldiff
     @lines ||= 0
 
     file_old, file_new = *ARGV
+    diff?(
+      InputInfo.new(file_old),
+      InputInfo.new(file_new),
+      @format,
+      output,
+      binary: @binary,
+      lines: @lines
+    ) ? 1 : 0
+  end
 
-    case @format
+  def diff?(info_old, info_new, format, output, binary: nil, lines: 0)
+    case format
     when :context
       char_old = "*" * 3
       char_new = "-" * 3
@@ -100,74 +121,69 @@ class << Diff::LCS::Ldiff
     # items we've read from each file will differ by FLD (could be 0).
     file_length_difference = 0
 
-    data_old = File.read(file_old)
-    data_new = File.read(file_new)
-
     # Test binary status
-    if @binary.nil?
-      old_txt = data_old[0, 4096].scan(/\0/).empty?
-      new_txt = data_new[0, 4096].scan(/\0/).empty?
-      @binary = !old_txt || !new_txt
-    end
-
-    unless @binary
-      data_old = data_old.lines.to_a
-      data_new = data_new.lines.to_a
+    if binary.nil?
+      old_bin = info_old.data[0, 4096].include?("\0")
+      new_bin = info_new.data[0, 4096].include?("\0")
+      binary = old_bin || new_bin
     end
 
     # diff yields lots of pieces, each of which is basically a Block object
-    if @binary
-      diffs = (data_old == data_new)
+    if binary
+      has_diffs = (info_old.data != info_new.data)
+      if format != :report
+        if has_diffs
+          output << "Binary files #{info_old.filename} and #{info_new.filename} differ\n"
+          return true
+        end
+        return false
+      end
     else
+      data_old = info_old.data.lines.to_a
+      data_new = info_new.data.lines.to_a
       diffs = Diff::LCS.diff(data_old, data_new)
-      diffs = nil if diffs.empty?
-    end
-
-    return 0 unless diffs
-
-    if @format == :report
-      output << "Files #{file_old} and #{file_new} differ\n"
-      return 1
+      return false if diffs.empty?
     end
 
-    if (@format == :unified) || (@format == :context)
-      ft = File.stat(file_old).mtime.localtime.strftime("%Y-%m-%d %H:%M:%S.000000000 %z")
-      output << "#{char_old} #{file_old}\t#{ft}\n"
-      ft = File.stat(file_new).mtime.localtime.strftime("%Y-%m-%d %H:%M:%S.000000000 %z")
-      output << "#{char_new} #{file_new}\t#{ft}\n"
+    case format
+    when :report
+      output << "Files #{info_old.filename} and #{info_new.filename} differ\n"
+      return true
+    when :unified, :context
+      ft = info_old.stat.mtime.localtime.strftime("%Y-%m-%d %H:%M:%S.000000000 %z")
+      output << "#{char_old} #{info_old.filename}\t#{ft}\n"
+      ft = info_new.stat.mtime.localtime.strftime("%Y-%m-%d %H:%M:%S.000000000 %z")
+      output << "#{char_new} #{info_new.filename}\t#{ft}\n"
+    when :ed
+      real_output = output
+      output = []
     end
 
     # Loop over hunks. If a hunk overlaps with the last hunk, join them.
     # Otherwise, print out the old one.
     oldhunk = hunk = nil
-
-    if @format == :ed
-      real_output = output
-      output = []
-    end
-
     diffs.each do |piece|
-      begin # rubocop:disable Style/RedundantBegin
-        hunk = Diff::LCS::Hunk.new(data_old, data_new, piece, @lines, file_length_difference)
+      begin
+        hunk = Diff::LCS::Hunk.new(data_old, data_new, piece, lines, file_length_difference)
         file_length_difference = hunk.file_length_difference
 
         next unless oldhunk
-        next if @lines.positive? && hunk.merge(oldhunk)
+        next if lines.positive? && hunk.merge(oldhunk)
 
-        output << oldhunk.diff(@format)
-        output << "\n" if @format == :unified
+        output << oldhunk.diff(format)
+        output << "\n" if format == :unified
       ensure
         oldhunk = hunk
       end
     end
 
-    last = oldhunk.diff(@format, true)
-    last << "\n" if last.respond_to?(:end_with?) && !last.end_with?("\n")
+    last = oldhunk.diff(format, true)
+    last << "\n" unless last.is_a?(Diff::LCS::Hunk) || last.empty? || last.end_with?("\n")
 
     output << last
 
-    output.reverse_each { |e| real_output << e.diff(:ed_finish) } if @format == :ed
+    output.reverse_each { |e| real_output << e.diff(:ed_finish, e == output[0]) } if format == :ed
 
-    1
+    true
   end
 end

data/lib/diff/lcs/version.rb

@@ -0,0 +1,7 @@
+# frozen_string_literal: true
+
+module Diff
+  module LCS
+    VERSION = "1.6.1"
+  end
+end

data/lib/diff/lcs.rb

@@ -49,9 +49,9 @@ module Diff; end unless defined? Diff
 #          a x b y c z p d q
 #    a b c a x b y c z
 module Diff::LCS
-  VERSION = "1.5.1"
 end
 
+require "diff/lcs/version"
 require "diff/lcs/callbacks"
 require "diff/lcs/internals"
 
@@ -62,11 +62,12 @@ module Diff::LCS
   #   lcs = seq1.lcs(seq2)
   #
   # A note when using objects: Diff::LCS only works properly when each object
-  # can be used as a key in a Hash, which typically means that the objects must
-  # implement Object#eql? in a way that two identical values compare
-  # identically for key purposes. That is:
+  # can be used as a key in a Hash. This means that those objects must implement
+  # the methods +#hash+ and +#eql?+ such that two objects containing identical values
+  # compare identically for key purposes. That is:
   #
-  #   O.new('a').eql?(O.new('a')) == true
+  #   O.new('a').eql?(O.new('a')) == true &&
+  #   O.new('a').hash == O.new('a').hash
   def lcs(other, &block) # :yields: self[i] if there are matched subsequences
     Diff::LCS.lcs(self, other, &block)
   end
@@ -256,7 +257,7 @@ class << Diff::LCS
   #
   # The methods for <tt>callbacks#match</tt>, <tt>callbacks#discard_a</tt>, and
   # <tt>callbacks#discard_b</tt> are invoked with an event comprising the
-  # action ("=", "+", or "-", respectively), the indicies +i+ and +j+, and the
+  # action ("=", "+", or "-", respectively), the indexes +i+ and +j+, and the
   # elements <tt>A[i]</tt> and <tt>B[j]</tt>. Return values are discarded by
   # #traverse_sequences.
   #
@@ -421,7 +422,7 @@ class << Diff::LCS
   #                                 occurred.
   #
   # #traverse_balanced might be a bit slower than #traverse_sequences,
-  # noticable only while processing huge amounts of data.
+  # noticeable only while processing huge amounts of data.
   #
   # == Algorithm
   #
@@ -465,7 +466,7 @@ class << Diff::LCS
   # The methods for <tt>callbacks#match</tt>, <tt>callbacks#discard_a</tt>,
   # <tt>callbacks#discard_b</tt>, and <tt>callbacks#change</tt> are invoked
   # with an event comprising the action ("=", "+", "-", or "!", respectively),
-  # the indicies +i+ and +j+, and the elements <tt>A[i]</tt> and <tt>B[j]</tt>.
+  # the indexes +i+ and +j+, and the elements <tt>A[i]</tt> and <tt>B[j]</tt>.
   # Return values are discarded by #traverse_balanced.
   #
   # === Context
@@ -482,7 +483,7 @@ class << Diff::LCS
 
     # Process all the lines in the match vector.
     loop do
-      # Find next match indices +ma+ and +mb+
+      # Find next match indexes +ma+ and +mb+
       loop do
         ma += 1
         break unless ma < matches.size && matches[ma].nil?

data/mise.toml

@@ -0,0 +1,5 @@
+[tools]
+ruby = "3.4"
+
+[env]
+MAINTENANCE = "true"

data/spec/fixtures/123_x

@@ -0,0 +1,2 @@
+123
+x

data/spec/fixtures/456_x

@@ -0,0 +1,2 @@
+456
+x

data/spec/fixtures/four_lines

@@ -0,0 +1,4 @@
+one
+two
+three
+four

data/spec/fixtures/four_lines_with_missing_new_line

@@ -0,0 +1,4 @@
+one
+two
+three
+four

data/spec/fixtures/ldiff/diff.missing_new_line1-e

@@ -0,0 +1 @@
+No newline at end of file

data/spec/fixtures/ldiff/diff.missing_new_line1-f

@@ -0,0 +1 @@
+No newline at end of file

data/spec/fixtures/ldiff/diff.missing_new_line2-e

@@ -0,0 +1 @@
+No newline at end of file

data/spec/fixtures/ldiff/diff.missing_new_line2-f

@@ -0,0 +1 @@
+No newline at end of file

data/spec/fixtures/ldiff/error.diff.chef-e

@@ -0,0 +1,2 @@
+<old_file>: No newline at end of file
+<new_file>: No newline at end of file

data/spec/fixtures/ldiff/error.diff.chef-f

@@ -0,0 +1,2 @@
+<old_file>: No newline at end of file
+<new_file>: No newline at end of file

data/spec/fixtures/ldiff/error.diff.missing_new_line1-e

@@ -0,0 +1 @@
+<new_file>: No newline at end of file

data/spec/fixtures/ldiff/error.diff.missing_new_line1-f

@@ -0,0 +1 @@
+<new_file>: No newline at end of file

data/spec/fixtures/ldiff/error.diff.missing_new_line2-e

@@ -0,0 +1 @@
+<old_file>: No newline at end of file

data/spec/fixtures/ldiff/error.diff.missing_new_line2-f

@@ -0,0 +1 @@
+<old_file>: No newline at end of file

data/spec/fixtures/ldiff/output.diff.bin2

@@ -0,0 +1 @@
+Binary files spec/fixtures/file1.bin and spec/fixtures/file2.bin differ

data/spec/fixtures/ldiff/output.diff.bin2-c

@@ -0,0 +1 @@
+Binary files spec/fixtures/file1.bin and spec/fixtures/file2.bin differ

data/spec/fixtures/ldiff/output.diff.bin2-e

@@ -0,0 +1 @@
+Binary files spec/fixtures/file1.bin and spec/fixtures/file2.bin differ

data/spec/fixtures/ldiff/output.diff.bin2-f

@@ -0,0 +1 @@
+Binary files spec/fixtures/file1.bin and spec/fixtures/file2.bin differ

data/spec/fixtures/ldiff/output.diff.bin2-u

@@ -0,0 +1 @@
+Binary files spec/fixtures/file1.bin and spec/fixtures/file2.bin differ

data/spec/fixtures/ldiff/output.diff.empty.vs.four_lines

@@ -0,0 +1,5 @@
+0a1,4
+> one
+> two
+> three
+> four

data/spec/fixtures/ldiff/output.diff.empty.vs.four_lines-c

@@ -0,0 +1,9 @@
+*** spec/fixtures/empty	2025-01-31 12:14:52.856031635 +0100
+--- spec/fixtures/four_lines	2025-01-31 12:13:45.476036544 +0100
+***************
+*** 0 ****
+--- 1,4 ----
++ one
++ two
++ three
++ four

data/spec/fixtures/ldiff/output.diff.empty.vs.four_lines-e

@@ -0,0 +1,6 @@
+0a
+one
+two
+three
+four
+.

data/spec/fixtures/ldiff/output.diff.empty.vs.four_lines-f

@@ -0,0 +1,6 @@
+a0
+one
+two
+three
+four
+.

data/spec/fixtures/ldiff/output.diff.empty.vs.four_lines-u

@@ -0,0 +1,7 @@
+--- spec/fixtures/empty	2025-01-31 12:14:52.856031635 +0100
++++ spec/fixtures/four_lines	2025-01-31 12:13:45.476036544 +0100
+@@ -0,0 +1,4 @@
++one
++two
++three
++four

data/spec/fixtures/ldiff/output.diff.four_lines.vs.empty

@@ -0,0 +1,5 @@
+1,4d0
+< one
+< two
+< three
+< four

data/spec/fixtures/ldiff/output.diff.four_lines.vs.empty-c

@@ -0,0 +1,9 @@
+*** spec/fixtures/four_lines	2025-01-31 12:13:45.476036544 +0100
+--- spec/fixtures/empty	2025-01-31 12:14:52.856031635 +0100
+***************
+*** 1,4 ****
+- one
+- two
+- three
+- four
+--- 0 ----

data/spec/fixtures/ldiff/output.diff.four_lines.vs.empty-e

@@ -0,0 +1 @@
+1,4d

data/spec/fixtures/ldiff/output.diff.four_lines.vs.empty-f

@@ -0,0 +1 @@
+d1 4

data/spec/fixtures/ldiff/output.diff.four_lines.vs.empty-u

@@ -0,0 +1,7 @@
+--- spec/fixtures/four_lines	2025-01-31 12:13:45.476036544 +0100
++++ spec/fixtures/empty	2025-01-31 12:14:52.856031635 +0100
+@@ -1,4 +0,0 @@
+-one
+-two
+-three
+-four

data/spec/fixtures/ldiff/output.diff.issue95_trailing_context

@@ -0,0 +1,4 @@
+1c1
+< 123
+---
+> 456

data/spec/fixtures/ldiff/output.diff.issue95_trailing_context-c

@@ -0,0 +1,9 @@
+*** spec/fixtures/123_x	2025-01-31 17:00:17.070615716 +0100
+--- spec/fixtures/456_x	2025-01-31 16:58:26.380624827 +0100
+***************
+*** 1,2 ****
+! 123
+  x
+--- 1,2 ----
+! 456
+  x

data/spec/fixtures/ldiff/output.diff.issue95_trailing_context-e

@@ -0,0 +1,3 @@
+1c
+456
+.

data/spec/fixtures/ldiff/output.diff.issue95_trailing_context-f

@@ -0,0 +1,3 @@
+c1
+456
+.

data/spec/fixtures/ldiff/output.diff.issue95_trailing_context-u

@@ -0,0 +1,6 @@
+--- spec/fixtures/123_x	2025-01-31 17:00:17.070615716 +0100
++++ spec/fixtures/456_x	2025-01-31 16:58:26.380624827 +0100
+@@ -1,2 +1,2 @@
+-123
++456
+ x

data/spec/fixtures/ldiff/output.diff.missing_new_line1

@@ -0,0 +1,5 @@
+4c4
+< four
+---
+> four
+

data/spec/fixtures/ldiff/output.diff.missing_new_line1-c

@@ -0,0 +1,14 @@
+*** spec/fixtures/four_lines	2025-01-31 12:17:43.926013315 +0100
+--- spec/fixtures/four_lines_with_missing_new_line	2025-01-31 12:17:43.926013315 +0100
+***************
+*** 1,4 ****
+  one
+  two
+  three
+! four
+--- 1,4 ----
+  one
+  two
+  three
+! four
+