omnizip 0.3.13 → 0.3.15

data/lib/omnizip/formats/lzma_alone.rb

@@ -138,13 +138,7 @@ module Omnizip
             uncompressed_size: decoder.uncompressed_size,
           }
         end
-
-        # Register LZMA_Alone format when loaded
-        def register!
-          FormatRegistry.register(".lzma", Omnizip::Formats::LzmaAlone)
-        end
       end
     end
   end
 end
-

data/lib/omnizip/formats/msi/entry.rb

@@ -8,8 +8,14 @@ module Omnizip
       # Represents a single file within an MSI package.
       # File information is assembled from File, Component, and Directory tables.
       class Entry
+        include Omnizip::Entry
         include Omnizip::Formats::Msi::Constants
 
+        def entry_name = path
+        def entry_directory? = false
+        def entry_size = size
+        def entry_mtime = nil
+
         # @return [String] Full installation path
         attr_accessor :path
 

data/lib/omnizip/formats/msi.rb

@@ -67,16 +67,7 @@ module Omnizip
         def info(path)
           open(path, &:info)
         end
-
-        # Register MSI format with format registry
-        #
-        # This overrides OLE's .msi registration.
-        def register!
-          FormatRegistry.register(".msi", self)
-          FormatRegistry.register(".msp", self)
-        end
       end
     end
   end
 end
-

data/lib/omnizip/formats/ole/dirent.rb

@@ -8,8 +8,14 @@ module Omnizip
       # Represents a file or directory entry in an OLE compound document.
       # Each dirent is 128 bytes and contains metadata about the entry.
       class Dirent
+        include Omnizip::Entry
         include Omnizip::Formats::Ole::Constants
 
+        def entry_name = name
+        def entry_directory? = dir?
+        def entry_size = nil
+        def entry_mtime = modify_time
+
         # Pack format for dirent structure
         PACK = "a64 v C C V3 a16 V a8 a8 V2 a4"
 

data/lib/omnizip/formats/ole.rb

@@ -117,14 +117,6 @@ module Omnizip
           end
         end
 
-        # Register OLE format in registry
-        def register!
-          FormatRegistry.register(".ole", Storage)
-          FormatRegistry.register(".doc", Storage)
-          FormatRegistry.register(".xls", Storage)
-          FormatRegistry.register(".ppt", Storage)
-        end
-
         private
 
         # Recursively extract dirent to directory
@@ -147,4 +139,3 @@ module Omnizip
     end
   end
 end
-

data/lib/omnizip/formats/rar.rb

@@ -161,13 +161,7 @@ module Omnizip
         def repair(archive_path, output_path)
           ArchiveRepairer.new.repair(archive_path, output_path)
         end
-
-        # Auto-register .rar format
-        def register!
-          FormatRegistry.register(".rar", "Omnizip::Formats::Rar::Reader")
-        end
       end
     end
   end
 end
-

data/lib/omnizip/formats/rar3/reader.rb

@@ -361,10 +361,17 @@ module Omnizip
 
       # RAR archive entry model
       class Entry
+        include Omnizip::Entry
+
         attr_accessor :name, :compressed_size, :uncompressed_size, :crc32,
                       :compression_method, :modified_time, :attributes,
                       :encrypted, :data_offset, :host_os, :salt
 
+        def entry_name = name
+        def entry_directory? = false
+        def entry_size = uncompressed_size
+        def entry_mtime = modified_time
+
         def initialize(name: nil, compressed_size: nil, uncompressed_size: nil,
                        crc32: nil, compression_method: nil, modified_time: nil,
                        attributes: nil, encrypted: nil, data_offset: nil,

data/lib/omnizip/formats/rar5/reader.rb

@@ -316,10 +316,17 @@ module Omnizip
 
       # RAR v5 archive entry model
       class Entry
+        include Omnizip::Entry
+
         attr_accessor :name, :compressed_size, :uncompressed_size, :crc32,
                       :compression_method, :modified_time, :attributes,
                       :encrypted, :data_offset, :host_os, :is_directory
 
+        def entry_name = name
+        def entry_directory? = is_directory
+        def entry_size = uncompressed_size
+        def entry_mtime = modified_time
+
         def initialize(name: nil, compressed_size: nil, uncompressed_size: nil,
                        crc32: nil, compression_method: nil, modified_time: nil,
                        attributes: nil, encrypted: nil, data_offset: nil,

data/lib/omnizip/formats/rpm/entry.rb

@@ -8,8 +8,14 @@ module Omnizip
       # Represents a single file within an RPM package.
       # File information is assembled from multiple header tags.
       class Entry
+        include Omnizip::Entry
         include Omnizip::Formats::Rpm::Constants
 
+        def entry_name = path
+        def entry_directory? = mode && ::File.directory?(mode)
+        def entry_size = size
+        def entry_mtime = mtime
+
         # @return [String] File path
         attr_accessor :path
 

data/lib/omnizip/formats/seven_zip/models/file_entry.rb

@@ -7,11 +7,18 @@ module Omnizip
         # Represents a file entry in .7z archive
         # Contains file metadata, attributes, and extraction information
         class FileEntry
+          include Omnizip::Entry
+
           attr_accessor :name, :size, :compressed_size, :crc, :is_dir,
                         :is_empty, :is_anti, :has_stream, :mtime, :atime,
                         :ctime, :attributes, :folder_index, :file_index,
                         :source_path, :inline_data, :compression_options
 
+          def entry_name = name
+          def entry_directory? = is_dir
+          def entry_size = size
+          def entry_mtime = mtime
+
           # Initialize file entry
           def initialize
             @name = nil

data/lib/omnizip/formats/seven_zip.rb

@@ -174,12 +174,6 @@ module Omnizip
       rescue StandardError
         false
       end
-
-      # Auto-register .7z format when loaded
-      def self.register!
-        FormatRegistry.register(".7z", "Omnizip::Formats::SevenZip::Reader")
-      end
     end
   end
 end
-

data/lib/omnizip/formats/tar/entry.rb

@@ -7,12 +7,18 @@ module Omnizip
       #
       # Represents a single entry (file, directory, link) in a TAR archive
       class Entry
+        include Omnizip::Entry
         include Omnizip::Formats::Tar::Constants
 
         attr_accessor :name, :mode, :uid, :gid, :size, :mtime, :typeflag,
                       :linkname, :uname, :gname, :devmajor, :devminor, :prefix
         attr_reader :data
 
+        def entry_name = name
+        def entry_directory? = directory?
+        def entry_size = size
+        def entry_mtime = mtime
+
         # Initialize a new TAR entry
         #
         # @param name [String] Entry name/path

data/lib/omnizip/formats/tar.rb

@@ -36,22 +36,20 @@ module Omnizip
         # @param file_path [String] Path to TAR archive
         # @yield [Reader] TAR reader instance
         # @return [Reader] TAR reader instance
-        # rubocop:disable Naming/BlockForwarding, Style/ArgumentsForwarding -- Ruby 3.0 compatibility
+        # rubocop:disable-next Naming/BlockForwarding, Style/ArgumentsForwarding -- Ruby 3.0 compatibility
         def open(file_path, &block)
           Reader.open(file_path, &block)
         end
-        # rubocop:enable Naming/BlockForwarding, Style/ArgumentsForwarding
 
         # Create a TAR archive
         #
         # @param file_path [String] Path to output TAR archive
         # @yield [Writer] TAR writer instance
         # @return [Writer] TAR writer instance
-        # rubocop:disable Naming/BlockForwarding, Style/ArgumentsForwarding -- Ruby 3.0 compatibility
+        # rubocop:disable-next Naming/BlockForwarding, Style/ArgumentsForwarding -- Ruby 3.0 compatibility
         def create(file_path, &block)
           Writer.create(file_path, &block)
         end
-        # rubocop:enable Naming/BlockForwarding, Style/ArgumentsForwarding
 
         # Extract a TAR archive
         #
@@ -70,13 +68,7 @@ module Omnizip
           reader = read(file_path)
           reader.list_entries
         end
-
-        # Register TAR format when loaded
-        def register!
-          FormatRegistry.register(".tar", Omnizip::Formats::Tar)
-        end
       end
     end
   end
 end
-

data/lib/omnizip/formats/xar/entry.rb

@@ -8,8 +8,14 @@ module Omnizip
       # Represents a single file/directory/link in a XAR archive.
       # Each entry has metadata and optional data.
       class Entry
+        include Omnizip::Entry
         include Omnizip::Formats::Xar::Constants
 
+        def entry_name = name
+        def entry_directory? = type == :directory
+        def entry_size = size
+        def entry_mtime = mtime
+
         attr_accessor :id, :name, :type, :mode, :uid, :gid, :user, :group,
                       :size, :ctime, :mtime, :atime,
                       :data_offset, :data_length, :data_size, :data_encoding,

data/lib/omnizip/formats/xar.rb

@@ -139,13 +139,7 @@ module Omnizip
         def info(path)
           open(path, &:info) # rubocop:disable Security/Open
         end
-
-        # Auto-register XAR format when loaded
-        def register!
-          FormatRegistry.register(".xar", Reader)
-        end
       end
     end
   end
 end
-

data/lib/omnizip/formats/zip.rb

@@ -44,13 +44,7 @@ module Omnizip
           reader = read(file_path)
           reader.list_entries
         end
-
-        # Auto-register .zip format when loaded
-        def register!
-          FormatRegistry.register(".zip", Omnizip::Formats::Zip)
-        end
       end
     end
   end
 end
-

data/lib/omnizip/implementations/seven_zip/lzma2/encoder.rb

@@ -326,7 +326,9 @@ _input_data)
             prob_is_rep = @models.is_rep[@state.value]
             encoder.queue_bit(prob_is_rep, 0)
 
-            @state.update_match!(distance)
+            # Reps are stored 0-based (distance - 1), matching the decoder's
+            # rep0 convention.
+            @state.update_match!(distance - 1)
 
             encode_match_length(length, pos_state, encoder)
             encode_distance(distance, length, encoder)
@@ -362,39 +364,36 @@ _input_data)
 
                 prob_is_rep2 = @models.is_rep2[@state.value]
                 encoder.queue_bit(prob_is_rep2, rep - 2)
-
-                if rep == 3
-                  @state.reps[3] = @state.reps[2]
-                end
-
-                @state.reps[2] = @state.reps[1]
               end
 
-              @state.reps[1] = @state.reps[0]
-
+              # XZ Utils rep_match: capture the selected distance BEFORE
+              # rotating, then shift reps down.
               distance = @state.reps[rep]
-
-              if distance.nil?
-                raise "Distance is nil for rep #{rep}, reps=#{@state.reps.inspect}"
-              end
-
+              rep.downto(1) { |i| @state.reps[i] = @state.reps[i - 1] }
               @state.reps[0] = distance
             end
 
             if length == 1
               @state.update_short_rep!
             else
-              encode_match_length(length, pos_state, encoder)
+              # Rep matches use their own length models (decoder's
+              # rep_length_coder).
+              encode_match_length(length, pos_state, encoder,
+                                  @models.rep_len_encoder)
               @state.update_long_rep!
             end
 
-            last_byte_pos = match_pos - @state.reps[0] + length - 1
+            # Last byte of the match region in the stream
+            last_byte_pos = match_pos + length - 1
             @prev_byte = @match_finder.buffer.getbyte(last_byte_pos) if last_byte_pos >= 0 && last_byte_pos < @match_finder.buffer.bytesize
           end
 
           def get_literal_state(pos, prev_byte)
             literal_mask = (0x100 << @lp) - (0x100 >> @lc)
-            ((((pos << 8) + prev_byte) & literal_mask) << @lc)
+            # Factor of 3: each literal context owns a 0x300-sized subcoder
+            # (XZ Utils literal_subcoder macro); must match the decoder's
+            # base_offset = 3 * (lit_state << lc).
+            3 * ((((pos << 8) + prev_byte) & literal_mask) << @lc)
           end
 
           def encode_normal_literal(literal_offset, symbol, encoder)
@@ -426,28 +425,31 @@ encoder)
             end
           end
 
-          def encode_match_length(length, pos_state, encoder)
+          # len_encoder selects the length model set: match_len_encoder for
+          # normal matches, rep_len_encoder for repeated matches.
+          def encode_match_length(length, pos_state, encoder,
+                                  len_encoder = @models.match_len_encoder)
             len = length - 2
 
             if len < 8
-              encoder.queue_bit(@models.match_len_encoder.choice, 0)
-              encode_bittree(@models.match_len_encoder.low[pos_state], 3, len,
-                             encoder)
+              encoder.queue_bit(len_encoder.choice, 0)
+              encode_bittree(len_encoder.low[pos_state], 3, len, encoder)
             elsif len < 16
-              encoder.queue_bit(@models.match_len_encoder.choice, 1)
-              encoder.queue_bit(@models.match_len_encoder.choice2, 0)
-              encode_bittree(@models.match_len_encoder.mid[pos_state], 3,
-                             len - 8, encoder)
+              encoder.queue_bit(len_encoder.choice, 1)
+              encoder.queue_bit(len_encoder.choice2, 0)
+              encode_bittree(len_encoder.mid[pos_state], 3, len - 8, encoder)
             else
-              encoder.queue_bit(@models.match_len_encoder.choice, 1)
-              encoder.queue_bit(@models.match_len_encoder.choice2, 1)
-              encode_bittree(@models.match_len_encoder.high, 8, len - 16,
-                             encoder)
+              encoder.queue_bit(len_encoder.choice, 1)
+              encoder.queue_bit(len_encoder.choice2, 1)
+              encode_bittree(len_encoder.high, 8, len - 16, encoder)
             end
           end
 
           def encode_distance(distance, length, encoder)
-            dist_slot = get_dist_slot(distance)
+            # get_dist_slot expects a 0-based distance (xz fastpos.h:
+            # dist 4 -> slot 4, dist 5 -> slot 4).
+            dist0 = distance - 1
+            dist_slot = get_dist_slot(dist0)
             len_state = [length - 2, 3].min
 
             encode_bittree(@models.dist_slot[len_state], 6, dist_slot, encoder)
@@ -455,7 +457,7 @@ encoder)
             if dist_slot >= 4
               footer_bits = (dist_slot >> 1) - 1
               base = (2 | (dist_slot & 1)) << footer_bits
-              dist_reduced = distance - base
+              dist_reduced = dist0 - base
 
               if dist_slot < 14
                 encode_bittree_reverse(@models.dist_special, dist_reduced,