The Trunk: Collections-ul.349.mcz

Levente Uzonyi uploaded a new version of Collections to project The Trunk:
http://source.squeak.org/trunk/Collections-ul.349.mcz

==================== Summary ====================

Name: Collections-ul.349
Author: ul
Time: 26 March 2010, 11:31:00.282 pm
UUID: 58393b9b-da51-9249-a692-36539b18270b
Ancestors: Collections-nice.348

- use the TextConverter's to encode/decode strings

=============== Diff against Collections-nice.348 ===============

Item was changed:
  ----- Method: ByteString>>squeakToUtf8 (in category 'converting') -----
  squeakToUtf8
+ "Convert the given string (self) to UTF-8 from Squeak's internal representation."
+
+ ^UTF8TextConverter encodeByteString: self!
- "Convert the given string from UTF-8 using the fast path if converting to Latin-1"
- | outStream lastIndex nextIndex |
- Latin1ToUtf8Map ifNil:[^super squeakToUtf8]. "installation guard"
- lastIndex := 1.
- nextIndex := ByteString findFirstInString: self inSet: Latin1ToUtf8Map startingAt: lastIndex.
- nextIndex = 0 ifTrue:[^self].
- outStream := (String new: self size * 2) writeStream.
- [outStream next: nextIndex-lastIndex putAll: self startingAt: lastIndex.
- outStream nextPutAll: (Latin1ToUtf8Encodings at: (self byteAt: nextIndex)+1).
- lastIndex := nextIndex + 1.
- nextIndex := ByteString findFirstInString: self inSet: Latin1ToUtf8Map startingAt: lastIndex.
- nextIndex = 0] whileFalse.
- outStream next: self size-lastIndex+1 putAll: self startingAt: lastIndex.
- ^outStream contents
- !

Item was changed:
  ----- Method: ByteString>>utf8ToSqueak (in category 'converting') -----
  utf8ToSqueak
+ "Convert the given string (self) from UTF-8 to Squeak's internal representation."
+
+ ^UTF8TextConverter decodeByteString: self
- "Convert the given string from UTF-8 using the fast path if converting to Latin-1"
- | outStream lastIndex nextIndex byte1 byte2 byte3 byte4 unicode |
- Latin1ToUtf8Map ifNil:[^super utf8ToSqueak]. "installation guard"
- lastIndex := 1.
- nextIndex := ByteString findFirstInString: self inSet: Latin1ToUtf8Map startingAt: lastIndex.
- nextIndex = 0 ifTrue:[^self].
- outStream := (String new: self size) writeStream.
- [outStream next: nextIndex-lastIndex putAll: self startingAt: lastIndex.
- byte1 := self byteAt: nextIndex.
- (byte1 bitAnd: 16rE0) = 192 ifTrue: [ "two bytes"
- byte2 := self byteAt: (nextIndex := nextIndex+1).
- (byte2 bitAnd: 16rC0) = 16r80 ifFalse:[^self]. "invalid UTF-8; presume Latin-1"
- unicode := ((byte1 bitAnd: 31) bitShift: 6) + (byte2 bitAnd: 63)].
- (byte1 bitAnd: 16rF0) = 224 ifTrue: [ "three bytes"
- byte2 := self byteAt: (nextIndex := nextIndex+1).
- (byte2 bitAnd: 16rC0) = 16r80 ifFalse:[^self]. "invalid UTF-8; presume Latin-1"
- byte3 := self byteAt: (nextIndex := nextIndex+1).
- (byte3 bitAnd: 16rC0) = 16r80 ifFalse:[^self]. "invalid UTF-8; presume Latin-1"
- unicode := ((byte1 bitAnd: 15) bitShift: 12) + ((byte2 bitAnd: 63) bitShift: 6)
- + (byte3 bitAnd: 63)].
- (byte1 bitAnd: 16rF8) = 240 ifTrue: [ "four bytes"
- byte2 := self byteAt: (nextIndex := nextIndex+1).
- (byte2 bitAnd: 16rC0) = 16r80 ifFalse:[^self]. "invalid UTF-8; presume Latin-1"
- byte3 := self byteAt: (nextIndex := nextIndex+1).
- (byte3 bitAnd: 16rC0) = 16r80 ifFalse:[^self]. "invalid UTF-8; presume Latin-1"
- byte4 := self byteAt: (nextIndex := nextIndex+1).
- (byte4 bitAnd: 16rC0) = 16r80 ifFalse:[^self]. "invalid UTF-8; presume Latin-1"
- unicode := ((byte1 bitAnd: 16r7) bitShift: 18) +
- ((byte2 bitAnd: 63) bitShift: 12) +
- ((byte3 bitAnd: 63) bitShift: 6) +
- (byte4 bitAnd: 63)].
- unicode ifNil:[^self]. "invalid UTF-8; presume Latin-1"
- unicode = 16rFEFF ifFalse: [ "Skip byte order mark"
- outStream nextPut: (Unicode value: unicode) ].
- lastIndex := nextIndex + 1.
- nextIndex := ByteString findFirstInString: self inSet: Latin1ToUtf8Map startingAt: lastIndex.
- nextIndex = 0] whileFalse.
- outStream next: self size-lastIndex+1 putAll: self startingAt: lastIndex.
- ^outStream contents
  !

Item was changed:
  ----- Method: String>>convertToWithConverter: (in category 'converting') -----
  convertToWithConverter: converter
+
+ ^(converter ifNil: [ ^self]) encodeString: self!
- converter
- ifNil: [^ self].
- ^ String
- new: self size
- streamContents: [:writeStream |
- converter
- nextPutAll: self toStream: writeStream;
- emitSequenceToResetStateIfNeededOn: writeStream]!

Item was changed:
  ----- Method: String>>convertFromWithConverter: (in category 'converting') -----
  convertFromWithConverter: converter
 
+ ^(converter ifNil: [ ^self]) decodeString: self!
- converter ifNil: [^ self].
- ^String new: self size streamContents: [ :stream |
- | readStream character |
- readStream := self readStream.
- [ (character := converter nextFromStream: readStream) == nil ]
- whileFalse: [ stream nextPut: character ] ]
-
- !