ruby-ole 1.2.8.2 → 1.2.9

data/ChangeLog

@@ -1,3 +1,7 @@
+== 1.2.9 / 2009-07-14
+
+- Lots of performance enhancements for RangesIO.
+
 == 1.2.8.2 / 2009-01-01
 
 - Update code to support ruby 1.9.1

data/lib/ole/ranges_io.rb

@@ -57,23 +57,14 @@ class RangesIO
 		@params = {:close_parent => false}.merge params
 		@mode = IO::Mode.new mode
 		@io = io
-		# convert ranges to arrays. check for negative ranges?
-		ranges ||= [0, io.size]
-		@ranges = ranges.map { |r| Range === r ? [r.begin, r.end - r.begin] : r }
-		# calculate size
-		@size = @ranges.inject(0) { |total, (pos, len)| total + len }
 		# initial position in the file
 		@pos = 0
-
+		self.ranges = ranges || [[0, io.size]]
 		# handle some mode flags
 		truncate 0 if @mode.truncate?
 		seek size if @mode.append?
 	end
-
-#IOError: closed stream
-# get this for reading, writing, everything...
-#IOError: not opened for writing
-
+	
 	# add block form. TODO add test for this
 	def self.open(*args, &block)
 		ranges_io = new(*args)
@@ -86,6 +77,36 @@ class RangesIO
 		end
 	end
 
+	def ranges= ranges
+		# convert ranges to arrays. check for negative ranges?
+		ranges = ranges.map { |r| Range === r ? [r.begin, r.end - r.begin] : r }
+		# combine ranges
+		if @params[:combine] == false
+			# might be useful for debugging...
+			@ranges = ranges
+		else
+			@ranges = []
+			next_pos = nil
+			ranges.each do |pos, len|
+				if next_pos == pos
+					@ranges.last[1] += len
+					next_pos += len
+				else
+					@ranges << [pos, len]
+					next_pos = pos + len
+				end
+			end
+		end
+		# calculate cumulative offsets from range sizes
+		@size = 0
+		@offsets = []
+		@ranges.each do |pos, len|
+			@offsets << @size
+			@size += len
+		end
+		self.pos = @pos
+	end
+
 	def pos= pos, whence=IO::SEEK_SET
 		case whence
 		when IO::SEEK_SET
@@ -95,30 +116,36 @@ class RangesIO
 			pos = @size + pos
 		else raise Errno::EINVAL
 		end
-		raise Errno::EINVAL unless (0...@size) === pos
+		raise Errno::EINVAL unless (0..@size) === pos
 		@pos = pos
+
+		# do a binary search throuh @offsets to find the active range.
+		a, c, b = 0, 0, @offsets.length
+		while a < b
+			c = (a + b) / 2
+			pivot = @offsets[c]
+			if pos == pivot
+				@active = c
+				return
+			elsif pos < pivot
+				b = c
+			else
+				a = c + 1
+			end
+		end
+
+		@active = a - 1
 	end
 
 	alias seek :pos=
 	alias tell :pos
 
-	def close
-		@io.close if @params[:close_parent]
+	def rewind
+		seek 0
 	end
 
-	# returns the [+offset+, +size+], pair inorder to read/write at +pos+
-	# (like a partial range), and its index.
-	def offset_and_size pos
-		total = 0
-		ranges.each_with_index do |(offset, size), i|
-			if pos <= total + size
-				diff = pos - total
-				return [offset + diff, size - diff], i
-			end
-			total += size
-		end
-		# should be impossible for any valid pos, (0...size) === pos
-		raise ArgumentError, "no range for pos #{pos.inspect}"
+	def close
+		@io.close if @params[:close_parent]
 	end
 
 	def eof?
@@ -130,24 +157,26 @@ class RangesIO
 		data = ''
 		return data if eof?
 		limit ||= size
-		partial_range, i = offset_and_size @pos
-		# this may be conceptually nice (create sub-range starting where we are), but
-		# for a large range array its pretty wasteful. even the previous way was. but
-		# i'm not trying to optimize this atm. it may even go to c later if necessary.
-		([partial_range] + ranges[i+1..-1]).each do |pos, len|
+		pos, len = @ranges[@active]
+		diff = @pos - @offsets[@active]
+		pos += diff
+		len -= diff
+		loop do
 			@io.seek pos
 			if limit < len
-				# convoluted, to handle read errors. s may be nil
-				s = @io.read limit
-				@pos += s.length if s
-				break data << s
+				s = @io.read(limit).to_s
+				@pos += s.length
+				data << s
+				break
 			end
-			# convoluted, to handle ranges beyond the size of the file
-			s = @io.read len
-			@pos += s.length if s
+			s = @io.read(len).to_s
+			@pos += s.length
 			data << s
 			break if s.length != len
 			limit -= len
+			break if @active == @ranges.length - 1
+			@active += 1
+			pos, len = @ranges[@active]
 		end
 		data
 	end
@@ -164,8 +193,6 @@ class RangesIO
 	end
 
 	def write data
-		# short cut. needed because truncate 0 may return no ranges, instead of empty range,
-		# thus offset_and_size fails.
 		return 0 if data.empty?
 		data_pos = 0
 		# if we don't have room, we can use the truncate hook to make more space.
@@ -176,8 +203,11 @@ class RangesIO
 				raise IOError, "unable to grow #{inspect} to write #{data.length} bytes" 
 			end
 		end
-		partial_range, i = offset_and_size @pos
-		([partial_range] + ranges[i+1..-1]).each do |pos, len|
+		pos, len = @ranges[@active]
+		diff = @pos - @offsets[@active]
+		pos += diff
+		len -= diff
+		loop do
 			@io.seek pos
 			if data_pos + len > data.length
 				chunk = data[data_pos..-1]
@@ -189,6 +219,9 @@ class RangesIO
 			@io.write data[data_pos, len]
 			@pos += len
 			data_pos += len
+			break if @active == @ranges.length - 1
+			@active += 1
+			pos, len = @ranges[@active]
 		end
 		data_pos
 	end
@@ -202,17 +235,13 @@ class RangesIO
 	def gets
 		s = read 1024
 		i = s.index "\n"
-		@pos -= s.length - (i+1)
+		self.pos -= s.length - (i+1)
 		s[0..i]
 	end
 	alias readline :gets
 
 	def inspect
-		# the rescue is for empty files
-		pos, len = (@ranges[offset_and_size(@pos).last] rescue [nil, nil])
-		range_str = pos ? "#{pos}..#{pos+len}" : 'nil'
-		"#<#{self.class} io=#{io.inspect}, size=#@size, pos=#@pos, "\
-			"range=#{range_str}>"
+		"#<#{self.class} io=#{io.inspect}, size=#{@size}, pos=#{@pos}>"
 	end
 end
 

data/lib/ole/storage/base.rb

@@ -21,7 +21,7 @@ module Ole # :nodoc:
 		class FormatError < StandardError # :nodoc:
 		end
 
-		VERSION = '1.2.8.2'
+		VERSION = '1.2.9'
 
 		# options used at creation time
 		attr_reader :params
@@ -61,7 +61,10 @@ module Ole # :nodoc:
 				else
 					@io.flush
 					# this is for the benefit of ruby-1.9
-					@io.syswrite('') if @io.respond_to?(:syswrite)
+					# generates warnings on jruby though... :/
+					if RUBY_PLATFORM != 'java' and @io.respond_to?(:syswrite)
+						@io.syswrite('')
+					end
 					true
 				end
 			rescue IOError
@@ -602,8 +605,8 @@ module Ole # :nodoc:
 				# note that old_blocks is != @ranges.length necessarily. i'm planning to write a
 				# merge_ranges function that merges sequential ranges into one as an optimization.
 				@bat.resize_chain @blocks, size
-				@ranges = @bat.ranges @blocks, size
-				@pos = @size if @pos > size
+				@pos = size if @pos > size
+				self.ranges = @bat.ranges(@blocks, size)
 				self.first_block = @blocks.empty? ? AllocationTable::EOC : @blocks.first
 
 				# don't know if this is required, but we explicitly request our @io to grow if necessary
@@ -612,8 +615,6 @@ module Ole # :nodoc:
 				# maybe its ok to just seek out there later??
 				max = @ranges.map { |pos, len| pos + len }.max || 0
 				@io.truncate max if max > @io.size
-
-				@size = size
 			end
 		end
 
@@ -633,8 +634,8 @@ module Ole # :nodoc:
 					# bat migration needed! we need to backup some data. the amount of data
 					# should be <= @ole.header.threshold, so we can just hold it all in one buffer.
 					# backup this
-					pos = @pos
-					@pos = 0
+					pos = [@pos, size].min
+					self.pos = 0
 					keep = read [@size, size].min
 					# this does a normal truncate to 0, removing our presence from the old bat, and
 					# rewrite the dirent's first_block
@@ -645,9 +646,9 @@ module Ole # :nodoc:
 					# important to do this now, before the write. as the below write will always
 					# migrate us back to sbat! this will now allocate us +size+ in the new bat.
 					super
-					@pos = 0
+					self.pos = 0
 					write keep
-					@pos = pos
+					self.pos = pos
 				else
 					super
 				end

data/test/test_ranges_io.rb

@@ -31,20 +31,13 @@ class TestRangesIO < Test::Unit::TestCase
 
 	def test_basics
 		assert_equal 160, @io.size
-		assert_match %r{size=160,.*range=100\.\.200}, @io.inspect
+		assert_match %r{size=160}, @io.inspect
 	end
 
 	def test_truncate
 		assert_raises(NotImplementedError) { @io.size += 10 }
 	end
 
-	def test_offset_and_size
-		assert_equal [[100, 100], 0], @io.offset_and_size(0)
-		assert_equal [[150, 50], 0], @io.offset_and_size(50)
-		assert_equal [[5, 5], 1], @io.offset_and_size(105)
-		assert_raises(ArgumentError) { @io.offset_and_size 1000 }
-	end
-
 	def test_seek
 		@io.pos = 10
 		@io.seek(-100, IO::SEEK_END)

data/test/test_storage.rb

@@ -1,14 +1,13 @@
 #! /usr/bin/ruby
 
 $: << File.dirname(__FILE__) + '/../lib'
+#require 'rubygems'
 
 require 'test/unit'
 require 'ole/storage'
 require 'digest/sha1'
 require 'stringio'
 require 'tempfile'
-require 'zlib'
-require 'base64'
 
 #
 # = TODO
@@ -108,49 +107,18 @@ class TestStorageRead < Test::Unit::TestCase
 	def test_read
 		# the regular String#hash was different on the mac, so asserting
 		# against full strings. switch this to sha1 instead of this fugly blob
-		data = <<-end
-			eJxjZGBgYkADAABKAAQ=
-
-			eJxjZPj3n5mLgeE/EDBwMjGAwAEwyeAmAyR8M5OL8ovz00oUwvOLUhTM9Ax0
-			XfKzS3NT80oYuEDywSBxl/xkBgEgD8TWA3LA8npmYGMAKWQZqA==
-
-			eJztVs9rE0EU/mZ32m5smyZtUlKQunhIezAV6knwYGoRxRLQBG89zCabdiE7
-			K9kJpt7Es1DwLxCsP+jJqycv/S/8I9SrmPgmuy0pK1ilNZd8MPt233zz7dud
-			H+99OXBR+pziILgYgglkqaUvIYFs0pWEBSyzIfsHLKeA3Fl0Y6wTX4d2K7bH
-			uEvP6i90zhuf6P21fxpp0C9nOMOvGuMcUQ1811ZuqOjSVWuzszm8eb52cFR+
-			K/k7yd9L/kHyV6OO8gJBezwT7/UVaj/xY9QRjXHhMDoM5rapZ39o/ntRZ2+0
-			sY3xv0C5xkhT1rXo7gHmBr5VWg7f+gbZqU23KTotSqaT5L+Ba2waDmjR1AsQ
-			qZnf6BVRvv29/5rsUtkJhXpWqiohG6LdCOu7ba+pRFud5veuMBisiKl7rmh4
-			ckfn8jnkv2KxC4tNYpujfhk2NjKaZyNVo0PadoLGni4sMshDM3XlcD3D2DzL
-			gW95oaJcmj3Rv6r174gnyguk1p9HvqtHpdmE/pTHDIUBb50VMHFfNlwS5Fig
-			/ihKrcQH+wPoExCflMZJjWRSxZFf3Cd+zfPd0K64T+1HgS8kZshroLrnO0EL
-			00VNKbc90cKqpi9UPN/phBHXrgQ37a2EwpIelI6JVSFD4kQSGxQVs14WgCMK
-			ZXcQ7WHtYxPWw5JupyuJqLLgeLHPEt5jz4r5CxXWcbY=
-
-			eJzt0z1oU2EcRvE3rR+tWtGtOBQ3hS4tODgWKnQRlM5dHAoZCuIHaLcsBaGT
-			hIBksQmlQ1wCTgF3Q+Z0cHN1lZBQQhL/N60fUEEKh5bK+cHleXlJw+2hHY5S
-			yqXjsruvW++/HzzK3/jwdirN3/n4ZSHu2vspXY+9Gc/ro88txN107P3YK7EP
-			Yq/GPo6diV2LvRabj52KfRl7IbYQeyn2Xezl2N3Yi7H12MnYT7ETsc3Yley+
-			ndIoZN9xN55n8eyM32F2/M6F9vHfYyTa3lm/wH/Ipjyb8mzKsynPpjyb8mzK
-			synPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzK
-			synPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzK
-			synPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzK
-			synPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzK
-			synPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzK
-			synPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzK
-			synPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzKsynPpjyb8mzKsynPprx/NO33
-			+7VaLQ6VSqVUKpXL5cFg0Gq1Go1GtVptNpvFYrFer8dNp9M5lRc+B078dzoc
-			Dnu93uGh2+3G+deNDvm/z7Mpz6Y8m/JsyrMpz6Y8m/JsyrMpz6a8vYmU0rdB
-			SnOx2XkxnqX02/j8ZnXzb+fCrcr09r3Puexnc0efz84z8SznnzzfWH9x++HT
-			V+s/7//8zEnOPwDhN6kw
-		end
-		expect = data.split(/\n\s*\n/).map { |chunk| Zlib::Inflate.inflate Base64.decode64(chunk) }
+		sha1sums = %w[
+			d3d1cde9eb43ed4b77d197af879f5ca8b8837577
+			65b75cbdd1f94ade632baeeb0848dec2a342c844
+			cfc230ec7515892cfdb85e4a173e0ce364094970
+			ffd859d94647a11b693f06f092d1a2bccc59d50d
+		]
 
 		# test the ole storage type
 		type = 'Microsoft Word 6.0-Dokument'
 		assert_equal type, (@ole.root/"\001CompObj").read[32..-1][/([^\x00]+)/m, 1]
 		# i was actually not loading data correctly before, so carefully check everything here
-		assert_equal expect, @ole.root.children.map { |child| child.read }
+		assert_equal sha1sums, @ole.root.children.map { |child| Digest::SHA1.hexdigest child.read }
 	end
 
 	def test_dirent
@@ -169,9 +137,7 @@ class TestStorageRead < Test::Unit::TestCase
 
 		dirent.open('r') { |f| assert_equal 2, f.first_block }
 		dirent.open('w') { |f| }
-		assert_raises Errno::EINVAL do
-			dirent.open('a') { |f| }
-		end
+		dirent.open('a') { |f| }
 	end
 
 	def test_delete
@@ -206,8 +172,8 @@ class TestStorageWrite < Test::Unit::TestCase
 		assert_equal '9974e354def8471225f548f82b8d81c701221af7', sha1(io.string)
 		Ole::Storage.open(io, :update_timestamps => false) { }
 		# hash changed. used to be efa8cfaf833b30b1d1d9381771ddaafdfc95305c
-		# thats because i know truncate the io, and am probably removing some trailing allocated
-		# available blocks.
+		# thats because i now truncate the io, and am probably removing some trailing
+		# allocated available blocks.
 		assert_equal 'a39e3c4041b8a893c753d50793af8d21ca8f0a86', sha1(io.string)
 		# add a repack test here
 		Ole::Storage.open io, :update_timestamps => false, &:repack

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: ruby-ole
 version: !ruby/object:Gem::Version 
-  version: 1.2.8.2
+  version: 1.2.9
 platform: ruby
 authors: 
 - Charles Lowe
@@ -9,7 +9,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2009-01-01 00:00:00 +11:00
+date: 2009-07-14 00:00:00 +10:00
 default_executable: 
 dependencies: []
 
@@ -28,29 +28,29 @@ files:
 - ChangeLog
 - data/propids.yaml
 - bin/oletool
+- lib/ole/types.rb
+- lib/ole/storage.rb
 - lib/ole/support.rb
 - lib/ole/base.rb
-- lib/ole/storage.rb
-- lib/ole/file_system.rb
 - lib/ole/ranges_io.rb
+- lib/ole/file_system.rb
+- lib/ole/types/property_set.rb
+- lib/ole/types/base.rb
+- lib/ole/storage/meta_data.rb
 - lib/ole/storage/base.rb
 - lib/ole/storage/file_system.rb
-- lib/ole/storage/meta_data.rb
-- lib/ole/types.rb
-- lib/ole/types/base.rb
-- lib/ole/types/property_set.rb
-- test/test_types.rb
+- test/test_ranges_io.rb
+- test/test_storage.rb
 - test/test_filesystem.rb
 - test/test_support.rb
-- test/test_storage.rb
+- test/test_mbat.rb
+- test/test_types.rb
 - test/test_meta_data.rb
-- test/test_ranges_io.rb
 - test/test_property_set.rb
-- test/test_mbat.rb
-- test/test.doc
 - test/test_word_6.doc
 - test/test_word_97.doc
 - test/test_word_95.doc
+- test/test.doc
 - test/oleWithDirs.ole
 - test/test_SummaryInformation
 has_rdoc: true
@@ -80,16 +80,16 @@ required_rubygems_version: !ruby/object:Gem::Requirement
 requirements: []
 
 rubyforge_project: ruby-ole
-rubygems_version: 1.2.0
+rubygems_version: 1.3.1
 signing_key: 
 specification_version: 2
 summary: Ruby OLE library.
 test_files: 
-- test/test_types.rb
+- test/test_ranges_io.rb
+- test/test_storage.rb
 - test/test_filesystem.rb
 - test/test_support.rb
-- test/test_storage.rb
+- test/test_mbat.rb
+- test/test_types.rb
 - test/test_meta_data.rb
-- test/test_ranges_io.rb
 - test/test_property_set.rb
-- test/test_mbat.rb