archive-zip 0.2.0 → 0.3.0

data/NEWS

@@ -6,6 +6,13 @@ detailed information is available in the rest of the documentation.
 <b>NOTE:</b> Date stamps in the following entries are in YYYY/MM/DD format.
 
 
+== v0.3.0 (2009/01/23)
+
+* Made a significant performance improvement for the extraction of compressed
+  entries for performance on par with InfoZIP's unzip.  Parsing archives with
+  many entries is still a bit subpar however.
+
+
 == v0.2.0 (2008/08/06)
 
 * Traditional (weak) encryption is now supported.

data/lib/archive/support/zlib.rb

@@ -34,8 +34,8 @@ module Zlib # :nodoc:
     # and _strategy_ are all passed directly to Zlib::Deflate.new().  See the
     # documentation of that method for their meanings.
     #
-    # NOTE: Due to limitations in Ruby's finalization capabilities, the #close
-    # method is _not_ automatically called when this object is garbage
+    # <b>NOTE:</b> Due to limitations in Ruby's finalization capabilities, the
+    # #close method is _not_ automatically called when this object is garbage
     # collected.  Make sure to call #close when finished with this object.
     def initialize(io, level = Zlib::DEFAULT_COMPRESSION, window_bits = nil, mem_level = nil, strategy = nil)
       @delegate = io
@@ -46,9 +46,9 @@ module Zlib # :nodoc:
 
     # The CRC32 checksum of the uncompressed data written using this object.
     #
-    # NOTE: Anything still in the internal write buffer has not been processed,
-    # so calling #flush prior to examining this attribute may be necessary for
-    # an accurate computation.
+    # <b>NOTE:</b> Anything still in the internal write buffer has not been
+    # processed, so calling #flush prior to examining this attribute may be
+    # necessary for an accurate computation.
     attr_reader :crc32
 
     protected
@@ -70,18 +70,18 @@ module Zlib # :nodoc:
 
     # Returns the number of bytes of compressed data produced so far.
     #
-    # NOTE: Anything still in the internal write buffer has not been processed,
-    # so calling #flush prior to calling this method may be necessary for an
-    # accurate count.
+    # <b>NOTE:</b> Anything still in the internal write buffer has not been
+    # processed, so calling #flush prior to calling this method may be necessary
+    # for an accurate count.
     def compressed_size
       @deflater.total_out
     end
 
     # Returns the number of bytes sent to be compressed so far.
     #
-    # NOTE: Anything still in the internal write buffer has not been processed,
-    # so calling #flush prior to calling this method may be necessary for an
-    # accurate count.
+    # <b>NOTE:</b> Anything still in the internal write buffer has not been
+    # processed, so calling #flush prior to calling this method may be necessary
+    # for an accurate count.
     def uncompressed_size
       @deflater.total_in
     end
@@ -111,6 +111,10 @@ module Zlib # :nodoc:
   class ZReader
     include IO::Like
 
+    # The number of bytes to read from the delegate object each time the
+    # internal read buffer is filled.
+    DEFAULT_DELEGATE_READ_SIZE = 4096
+
     # Creates a new instance of this class with the given arguments using #new
     # and then passes the instance to the given block.  The #close method is
     # guaranteed to be called after the block completes.
@@ -133,11 +137,12 @@ module Zlib # :nodoc:
     # meaning.  If _io_ also responds to _rewind_, then the _rewind_ method of
     # this class can be used to reset the whole stream back to the beginning.
     #
-    # NOTE: Due to limitations in Ruby's finalization capabilities, the #close
-    # method is _not_ automatically called when this object is garbage
+    # <b>NOTE:</b> Due to limitations in Ruby's finalization capabilities, the
+    # #close method is _not_ automatically called when this object is garbage
     # collected.  Make sure to call #close when finished with this object.
     def initialize(io, window_bits = nil)
       @delegate = io
+      @delegate_read_size = DEFAULT_DELEGATE_READ_SIZE
       @window_bits = window_bits
       @inflater = Zlib::Inflate.new(@window_bits)
       @crc32 = 0
@@ -146,11 +151,15 @@ module Zlib # :nodoc:
 
     # The CRC32 checksum of the uncompressed data read using this object.
     #
-    # NOTE: The contents of the internal read buffer are immediately processed
-    # any time the buffer is filled, so this count is only accurate if all data
-    # has been read out of this object.
+    # <b>NOTE:</b> The contents of the internal read buffer are immediately
+    # processed any time the buffer is filled, so this count is only accurate if
+    # all data has been read out of this object.
     attr_reader :crc32
 
+    # The number of bytes to read from the delegate object each time the
+    # internal read buffer is filled.
+    attr_accessor :delegate_read_size
+
     protected
 
     # The delegate object from which compressed data is read.
@@ -180,11 +189,14 @@ module Zlib # :nodoc:
     private
 
     def unbuffered_read(length)
-      raise EOFError, 'end of file reached' if @inflater.finished?
+      if @decompress_buffer.empty? && @inflater.finished? then
+        raise EOFError, 'end of file reached'
+      end
 
       begin
         while @decompress_buffer.length < length && ! @inflater.finished? do
-          @decompress_buffer << @inflater.inflate(delegate.read(1))
+          @decompress_buffer <<
+            @inflater.inflate(delegate.read(@delegate_read_size))
         end
       rescue Errno::EINTR, Errno::EAGAIN
         raise if @decompress_buffer.empty?

data/lib/archive/zip.rb

@@ -242,9 +242,10 @@ module Archive # :nodoc:
     # <b>:flatten</b>::
     #   When set to +false+ (the default), the directory paths containing
     #   archived files will be included in the zip paths of entries representing
-    #   the files.  When set to +true+ files are archived without any containing
-    #   directory structure in the zip paths.  Setting to +true+ implies that
-    #   <b>:directories</b> is +false+ and <b>:path_prefix</b> is empty.
+    #   the files.  When set to +true+, files are archived without any
+    #   containing directory structure in the zip paths.  Setting to +true+
+    #   implies that <b>:directories</b> is +false+ and <b>:path_prefix</b> is
+    #   empty.
     # <b>:exclude</b>::
     #   Specifies a proc or lambda which takes a single argument containing a
     #   prospective zip entry and returns +true+ if the entry should be excluded
@@ -444,13 +445,13 @@ module Archive # :nodoc:
     #   <tt>:all</tt>.
     # <b>:create</b>::
     #   When set to +true+ (the default), files and directories which do not
-    #   already exist will be extracted.  When set to +false+ only files and
+    #   already exist will be extracted.  When set to +false+, only files and
     #   directories which already exist will be extracted (depending on the
     #   setting of <b>:overwrite</b>).
     # <b>:flatten</b>::
     #   When set to +false+ (the default), the directory paths containing
     #   extracted files will be created within +destination+ in order to contain
-    #   the files.  When set to +true+ files are extracted directly to
+    #   the files.  When set to +true+, files are extracted directly to
     #   +destination+ and directory entries are skipped.
     # <b>:exclude</b>::
     #   Specifies a proc or lambda which takes a single argument containing a
@@ -626,8 +627,8 @@ module Archive # :nodoc:
 
     private
 
-    # NOTE: For now _io_ MUST be seekable and report such by returning +true+
-    # from its seekable? method.  See IO#seekable?.
+    # <b>NOTE:</b> For now _io_ MUST be seekable and report such by returning
+    # +true+ from its seekable? method.  See IO#seekable?.
     #
     # Raises Archive::Zip::IOError if _io_ is not seekable.
     def parse(io)

data/lib/archive/zip/data_descriptor.rb

@@ -25,10 +25,10 @@ module Archive; class Zip
     # like-named attributes of _other_ and raises Archive::Zip::Error for any
     # mismatches.
     #
-    # NOTE: The compressed_size attribute is not checked because encrypted
-    # entries may have misleading compressed sizes.  Checking only the CRC32 and
-    # uncompressed size of the data should be sufficient to ensure that an entry
-    # has been successfully extracted.
+    # <b>NOTE:</b> The compressed_size attribute is not checked because
+    # encrypted entries may have misleading compressed sizes.  Checking only the
+    # CRC32 and uncompressed size of the data should be sufficient to ensure
+    # that an entry has been successfully extracted.
     def verify(other)
       unless crc32 == other.crc32 then
         raise Zip::Error,

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: archive-zip
 version: !ruby/object:Gem::Version 
-  version: 0.2.0
+  version: 0.3.0
 platform: ruby
 authors: 
 - Jeremy Bopp
@@ -9,7 +9,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2008-08-07 00:00:00 -05:00
+date: 2009-01-23 00:00:00 -06:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
@@ -37,24 +37,24 @@ extra_rdoc_files:
 - NEWS
 - README
 files: 
+- lib/archive/support/io-like.rb
+- lib/archive/support/io.rb
+- lib/archive/support/iowindow.rb
+- lib/archive/support/stringio.rb
+- lib/archive/support/time.rb
+- lib/archive/support/zlib.rb
+- lib/archive/zip/codec/deflate.rb
+- lib/archive/zip/codec/null_encryption.rb
+- lib/archive/zip/codec/store.rb
+- lib/archive/zip/codec/traditional_encryption.rb
 - lib/archive/zip/codec.rb
-- lib/archive/zip/entry.rb
 - lib/archive/zip/data_descriptor.rb
+- lib/archive/zip/entry.rb
+- lib/archive/zip/error.rb
 - lib/archive/zip/extra_field/extended_timestamp.rb
-- lib/archive/zip/extra_field/unix.rb
 - lib/archive/zip/extra_field/raw.rb
-- lib/archive/zip/error.rb
+- lib/archive/zip/extra_field/unix.rb
 - lib/archive/zip/extra_field.rb
-- lib/archive/zip/codec/null_encryption.rb
-- lib/archive/zip/codec/store.rb
-- lib/archive/zip/codec/deflate.rb
-- lib/archive/zip/codec/traditional_encryption.rb
-- lib/archive/support/io.rb
-- lib/archive/support/iowindow.rb
-- lib/archive/support/zlib.rb
-- lib/archive/support/io-like.rb
-- lib/archive/support/time.rb
-- lib/archive/support/stringio.rb
 - lib/archive/zip.rb
 - CONTRIBUTORS
 - HACKING