Issue 75: add stringClass attributes to FileSystem

src/FileSystem-Core/AbstractFileReference.class.st

@@ -743,10 +743,13 @@ AbstractFileReference >> readStreamDo: doBlock ifAbsent: absentBlock [
 
 { #category : 'streams' }
 AbstractFileReference >> readStreamEncoded: anEncoding [
-
 	^ ZnCharacterReadStream
 		on: self binaryReadStream
 		encoding: anEncoding
+		stringClass:
+			(String isInUnicodeComparisonMode
+				ifTrue: [ Unicode7 ]
+				ifFalse: [ String ])
 ]
 
 { #category : 'streams' }

src/FileSystem-GemStone-Kernel/CharacterCollection.extension.st

@@ -20,12 +20,15 @@ CharacterCollection >> asResolvedBy: aFileSystem [
 { #category : '*filesystem-gemstone-kernel' }
 CharacterCollection >> asZnCharacterEncoder [
 	"Return a ZnCharacterEncoder instance using the receiver as identifier"
-	
+
 	" 'UTF-8' asZnCharacterEncoder "
-
-	((self select: [ :each | each isAlphaNumeric ]) asLowercase) = 'utf8' ifFalse: [ self error: 'Only utf8 encoding supported'].
-	^ ZnUTF8Encoder new
 
+	(self select: [ :each | each isAlphaNumeric ]) asLowercase = 'utf8'
+		ifTrue: [ ^ ZnUTF8Encoder new ]
+		ifFalse: [ 
+			(self select: [ :each | each isAlphaNumeric ]) asLowercase = '8bit'
+				ifFalse: [ self error: 'only 8bit or  utf8 encoding supported' ] ].
+	^ Zn8BITEncoder new
 ]
 
 { #category : '*filesystem-gemstone-kernel' }

src/FileSystem-Tests-Core/FileReferenceTest.class.st

@@ -1164,7 +1164,7 @@ FileReferenceTest >> testReadStreamIfAbsent [
 { #category : 'tests' }
 FileReferenceTest >> testRelativeTo [
 
-	| alpha beta reference path result |
+	| alpha beta |
 	alpha := sandbox / 'alpha'.
 	beta := alpha / 'beta'.
 	self
@@ -1326,7 +1326,7 @@ FileReferenceTest >> testWithExtension [
 { #category : 'tests' }
 FileReferenceTest >> testWithoutExtension [
 
-	| reference result |
+	| reference |
 	reference := sandbox / 'alpha.beta.gamma'.
 	reference := reference withoutExtension.
 	self

src/Zinc-Character-Encoding-Core/Zn8BITEncoder.class.st

@@ -0,0 +1,123 @@
+"
+Part of FileSystem
+
+=========
+
+I implement the encoding and decoding of Extended ASCII (8 bit character encoding) that produces instances of class String.
+
+The encoding is consistent with topaz 'fileformat 8BIT' (see section 1.3 Handling text outside the ASCII range in the topaz manual[1]  for more details).
+
+[1] https://downloads.gemtalksystems.com/docs/GemStone64/3.6.x/GS64-Topaz-3.6/GS64-Topaz-3.6.htm?https://downloads.gemtalksystems.com/docs/GemStone64/3.6.x/GS64-Topaz-3.6/1-Tutorial.htm#pgfId-1130673
+"
+Class {
+	#name : 'Zn8BITEncoder',
+	#superclass : 'ZnCharacterEncoder',
+	#classVars : [
+		'Default'
+	],
+	#category : 'Zinc-Character-Encoding-Core'
+}
+
+{ #category : 'accessing' }
+Zn8BITEncoder class >> default [
+	"Return a cached instance of the most commonly used encoder,
+	which is faster than going via #newForEncoding: that does a subclass search"
+
+	^ Default ifNil: [ Default := self new ]
+]
+
+{ #category : 'accessing' }
+Zn8BITEncoder class >> handlesEncoding: string [
+	"Return true when my instances handle the encoding described by string"
+
+	^ (self canonicalEncodingIdentifier: string) = '8bit'
+]
+
+{ #category : 'accessing' }
+Zn8BITEncoder class >> knownEncodingIdentifiers [
+	^ #( #'8bit' )
+]
+
+{ #category : 'instance creation' }
+Zn8BITEncoder class >> newForEncoding: string stringClass: stringClass [
+	"Return a new character encoder object for an encoding described by string.
+	Search for a subclass that handles it and delegate (subclassResponsibility)."
+
+	^ self new stringClass: stringClass
+]
+
+{ #category : 'converting' }
+Zn8BITEncoder >> backOnStream: stream [
+	"Move back one character on stream"
+
+	stream position = 0
+		ifTrue: [Error signal: 'Cannot move backward past the start of the stream.'].
+	stream skip: -1
+]
+
+{ #category : 'convenience' }
+Zn8BITEncoder >> decodeAsCodePoints: bytes [
+	"Decode bytes and return the resulting code points"
+
+	^ String withBytes: bytes
+]
+
+{ #category : 'convenience' }
+Zn8BITEncoder >> decodeBytes: bytes [
+	"Decode bytes and return the resulting string"
+
+	^ self stringClass withBytes: bytes
+]
+
+{ #category : 'converting' }
+Zn8BITEncoder >> encodedByteCountFor: character [
+	"Return how many bytes are needed to encode character"
+
+	^ 1
+]
+
+{ #category : 'convenience' }
+Zn8BITEncoder >> encodeString: string [
+	"Encode string and return the resulting Utf8 instance"
+
+	^ string asByteArray
+]
+
+{ #category : 'accessing' }
+Zn8BITEncoder >> identifier [
+	^ #'8bit'
+]
+
+{ #category : 'converting' }
+Zn8BITEncoder >> nextCodePointFromStream: stream [
+	"Read and return the next integer code point from stream"
+
+	^ stream next
+]
+
+{ #category : 'converting' }
+Zn8BITEncoder >> nextFromStream: stream [
+	"Read and return the next character from stream"
+
+	^ Character codePoint: stream next
+]
+
+{ #category : 'converting' }
+Zn8BITEncoder >> nextPutCodePoint: codePoint toStream: stream [
+	"Write the encoding for Integer code point to stream"
+
+	^ stream nextPut: (Character codePoint: codePoint)
+]
+
+{ #category : 'convenience' }
+Zn8BITEncoder >> readInto: string startingAt: offset count: requestedCount fromStream: stream [
+	"Read requestedCount characters into string starting at offset,
+	returning the number read, there could be less available when stream is atEnd."
+
+	| stringBuffer |
+	stringBuffer := string.
+	offset to: offset + requestedCount - 1 do: [ :index | 
+		stream atEnd ifTrue: [ ^ index - offset ].  
+		stringBuffer codePointAt: index put: (self nextCodePointFromStream: stream)].
+	^ requestedCount
+]

src/Zinc-Character-Encoding-Core/ZnBufferedReadStream.class.st

@@ -68,7 +68,13 @@ ZnBufferedReadStream >> closed [
 ZnBufferedReadStream >> collectionSpecies [
 	^ stream isBinary
 		ifTrue: [ ByteArray ]
-		ifFalse: [ String ]
+		ifFalse: [ 
+			(stream respondsTo: #'stringClass')
+				ifTrue: [ stream stringClass ]
+				ifFalse: [ 
+					String isInUnicodeComparisonMode
+						ifTrue: [ Unicode7 ]
+						ifFalse: [ String ] ] ]
 ]
 
 { #category : 'accessing' }
@@ -403,25 +409,31 @@ ZnBufferedReadStream >> uint8 [
 { #category : 'accessing' }
 ZnBufferedReadStream >> upTo: value [ 
 	"Read upto but not including value and return them as a collection.
-	If value is not found, return the entire contents of the stream.
-	This could be further optimzed."
+	If value is not found, return the entire contents of the stream."
 
 	^ self collectionSpecies 
-		streamContents: [ :writeStream | | element |
-			[ self atEnd or: [ (element := self next) = value ] ] whileFalse: [ 
-				writeStream nextPut: element ] ]
+		streamContents: [ :writeStream | | ch |
+			[ self atEnd or: [ (ch := self next) = value ] ] whileFalse: [ 
+				writeStream nextPut: ch ] ]
 ]
 
 { #category : 'accessing' }
 ZnBufferedReadStream >> upToAll: aCollection [ 
 "Answer a subcollection from the current access position to the occurrence (if any, but not inclusive) of aCollection. If aCollection is not in the stream, answer the entire rest of the stream."
 
 | startPos endMatch result x |
+aCollection isEmpty ifTrue: [ ^aCollection ].
 startPos := self position.
+"upTo: will stop before aCollection first"
 x := self upTo: aCollection first.
-self atEnd ifTrue: [ ^ x ].
-2 to: aCollection size do: [:i |
-	self peek = (aCollection at: i)
+self atEnd
+	ifTrue: [ 
+		aCollection size <= 1
+			ifTrue: [ ^ x ].
+		self position: startPos.
+		^ self upToEnd].
+2 to: aCollection size do: [:i | | y |
+	(y := self peek) = (aCollection at: i)
 		ifTrue: [ self next ]
 		ifFalse: [ self position: startPos.
 		^ self upToEnd ] ].

src/Zinc-Character-Encoding-Core/ZnCharacterEncoder.class.st

@@ -41,6 +41,9 @@ Part of Zinc HTTP Components.
 Class {
 	#name : 'ZnCharacterEncoder',
 	#superclass : 'Object',
+	#instVars : [
+		'stringClass'
+	],
 	#category : 'Zinc-Character-Encoding-Core'
 }
 
@@ -70,12 +73,25 @@ ZnCharacterEncoder class >> knownEncodingIdentifiers [
 ZnCharacterEncoder class >> newForEncoding: string [
 	"Return a new character encoder object for an encoding described by string.
 	Search for a subclass that handles it and delegate (subclassResponsibility)."
+
+	^ self
+		newForEncoding: string
+		stringClass:
+			(String isInUnicodeComparisonMode
+				ifTrue: [ Unicode7 ]
+				ifFalse: [ String ])
+]
+
+{ #category : 'instance creation' }
+ZnCharacterEncoder class >> newForEncoding: string stringClass: stringClass [
+	"Return a new character encoder object for an encoding described by string.
+	Search for a subclass that handles it and delegate (subclassResponsibility)."
 
 	| concreteSubclass |
 	concreteSubclass := self allSubclasses 
 		detect: [ :each | each handlesEncoding: string ] 
 		ifNone: [ ^ self error: 'The ', string printString, ' is not currently supported.' ].
-	^ concreteSubclass newForEncoding: string
+	^ concreteSubclass newForEncoding: string stringClass: stringClass
 ]
 
 { #category : 'converting' }
@@ -157,3 +173,17 @@ ZnCharacterEncoder >> nextPut: character toStream: stream [
 
 	self nextPutCodePoint: character asInteger toStream: stream
 ]
+
+{ #category : 'accessing' }
+ZnCharacterEncoder >> stringClass [
+	^ stringClass
+		ifNil: [ 
+			stringClass := String isInUnicodeComparisonMode
+				ifTrue: [ Unicode7 ]
+				ifFalse: [ String ] ]
+]
+
+{ #category : 'accessing' }
+ZnCharacterEncoder >> stringClass: object [
+	stringClass := object
+]

src/Zinc-Character-Encoding-Core/ZnCharacterReadStream.class.st

@@ -13,12 +13,29 @@ Part of Zinc HTTP Components.
 Class {
 	#name : 'ZnCharacterReadStream',
 	#superclass : 'ZnEncodedReadStream',
+	#instVars : [
+		'stringClass'
+	],
 	#category : 'Zinc-Character-Encoding-Core'
 }
 
+{ #category : 'instance creation' }
+ZnCharacterReadStream class >> on: wrappedStream encoding: encoding stringClass: stringClass [
+	^ self new
+		on: wrappedStream;
+		encoding: encoding;
+		stringClass: stringClass;
+		yourself
+]
+
 { #category : 'accessing' }
 ZnCharacterReadStream >> collectionSpecies [
-	^ String
+	^ self stringClass
+]
+
+{ #category : 'accessing' }
+ZnCharacterReadStream >> encoder [
+	^ encoder ifNil: [ encoder := super encoder stringClass: self stringClass ]
 ]
 
 { #category : 'accessing' }
@@ -91,6 +108,16 @@ ZnCharacterReadStream >> readInto: collection startingAt: offset count: requeste
 
 ]
 
+{ #category : 'accessing' }
+ZnCharacterReadStream >> stringClass [
+	^ stringClass ifNil: [ stringClass := String ]
+]
+
+{ #category : 'accessing' }
+ZnCharacterReadStream >> stringClass: object [
+	stringClass := object
+]
+
 { #category : 'accessing' }
 ZnCharacterReadStream >> upToAll: aCollection [
 	"Answer a subcollection from the current access position to the occurrence (if any, but not inclusive) of aCollection. If aCollection is not in the stream, answer the entire rest of the stream."