- Pharo-dev - lists.pharo.org

[Pharo-project] Issue 3241 in pharo: Eliot's stdio changes
by pharo＠googlecode.com Nov. 11, 2010

Nov. 11, 2010

Status: FixedWaitingToBePharoed Owner: stephane.ducasse Labels: Type-Squeak New issue 3241 by stephane.ducasse: Eliot's stdio changes http://code.google.com/p/pharo/issues/detail?id=3241 Levente Uzonyi uploaded a new version of Files to project The Trunk: http://source.squeak.org/trunk/Files-ul.90.mcz ==================== Summary ==================== Name: Files-ul.90 Author: ul Time: 27 September 2010, 3:56:36.551 am UUID: aff6ed6c-5823-6342-b175-65f0f4e8f403 Ancestors: Files-ul.89 - Integrated Eliot's stdio changes with a minor fix/optimization =============== Diff against Files-ul.89 =============== Item was added: + ----- Method: CrLfFileStream class>>newForStdio (in category 'instance creation') ----- + newForStdio + "Circumvent CrLfFileStream new's retuning an instance of MultiBteFileStream" + ^super new! Item was added: + ----- Method: CrLfFileStream>>openOnHandle:name:forWrite: (in category 'open/close') ----- + openOnHandle: aFileID name: streamName forWrite: writeMode + "Initialize the file with the given handle. If writeMode is true then + allow writing, otherwise put the file in read-only mode." + super openOnHandle: aFileID name: streamName forWrite: writeMode. + lineEndConvention := LineEndDefault! Item was changed: ----- Method: DirectoryEntry>>fullName (in category 'access') ----- fullName + "The fully-qualified name. + Since this method falls into the equality test, make it safe when directory is nil." + ^ directory + ifNotNil: [ directory fullNameFor: self name ] + ifNil: [ self name ]! - "The fully-qualified name." - ^ directory fullNameFor: self name! Item was added: + ----- Method: FileDirectory>>directoryEntryForName: (in category 'private') ----- + directoryEntryForName: aFileName + + "Return a single DirectoryEntry for the given (non-path) entry name, + or nil if the entry could not be found. + Raises InvalidDirectoryError if this directory's path does not identify a directory." + + | entryArray sysPath sysName | + + sysPath := pathName asVmPathName. + sysName := aFileName asVmPathName. + + "New linear-time primitive." + entryArray := self primLookupEntryIn: sysPath name: sysName. + entryArray == #primFailed ifFalse:[ + ^ entryArray ifNotNil: [(DirectoryEntry fromArray: entryArray directory: self) convertFromSystemName] + ]. + + "If the new primitive fails, use the old slow method. + (This fallback can be changed to signal InvalidDirectoryError once + VM's with FilePlugin #primitiveDirectoryEntry have been distributed everywhere; + the new primitive was introduced 6/13/2007." + + ^self isCaseSensitive ifTrue: [ + self entries detect: [:entry | entry name = aFileName ] ifNone: [ nil ] + ] ifFalse: [ + self entries detect: [:entry | entry name sameAs: aFileName ] ifNone: [ nil ] + ] + ! Item was changed: FileStream subclass: #StandardFileStream instanceVariableNames: 'name fileID buffer1' + classVariableNames: 'Registry StdioHandles' - classVariableNames: 'Registry' poolDictionaries: '' category: 'Files-Kernel'! + StandardFileStream class + instanceVariableNames: 'stdioFiles'! !StandardFileStream commentStamp: 'ul 12/6/2009 05:13' prior: 0! Provides a simple, platform-independent, interface to a file system. The instance variable rwmode, inherited from class PositionableStream, here is used to hold a Boolean -- true means opened for read-write, false means opened for read-only. 2/12/96 sw I implement a simple read buffering scheme with the variables defined in PositionableStream (which are unused in me otherwise) in the following way: collection <ByteString> or <ByteArray> This is the buffer. position <Integer> The relative position in the buffer. Greater or equal to zero. readLimit <Integer> The number of bytes buffered. Greater or equal to zero. Read buffering is enabled with #enableReadBuffering, disabled with #disableReadBuffering and it is enabled by default. The buffer is filled when a read attempt of an unbuffered absolute position is requested, or when a negative repositioning is made (with #position: with an argument < than the current absolute position) to an absolute position which is not buffered. In the first case, the buffer is positioned to the given absolute position. In the latter case the repositioning is made to the requested absolute position minus fourth of the buffer size. This means that further small negative repositionings won't result in buffer flushing. This is really useful when filing in code. The read buffer is flushed (#flushReadBuffer) whenever a write attempt is made. The buffer state is valid if and only if collection is not nil and position < readLimit.! + StandardFileStream class + instanceVariableNames: 'stdioFiles'! Item was changed: ----- Method: StandardFileStream class>>fileExistsUserHandling: (in category 'error handling') ----- fileExistsUserHandling: fullFileName | dir localName choice newName newFullFileName | dir := FileDirectory forFileName: fullFileName. localName := FileDirectory localNameFor: fullFileName. + choice := (UIManager default + chooseFrom: #('overwrite that file' 'append (risky!!!!)' 'choose another name' 'cancel') + title: localName, ' already exists.'). - choice := UIManager default chooseFrom:{ - 'overwrite that file'. - 'choose another name'. - 'cancel' - } title: localName, ' - already exists.'. choice = 1 ifTrue: [ dir deleteFileNamed: localName ifAbsent: [self error: 'Could not delete the old version of that file']. ^ self new open: fullFileName forWrite: true]. choice = 2 ifTrue: [ + ^ (self new open: fullFileName forWrite: true) setToEnd]. + + choice = 3 ifTrue: [ newName := UIManager default request: 'Enter a new file name' initialAnswer: fullFileName. newFullFileName := self fullName: newName. ^ self newFileNamed: newFullFileName]. self error: 'Please close this to abort file opening'! Item was added: + ----- Method: StandardFileStream class>>flushAndVoidStdioFiles (in category 'class initialization') ----- + flushAndVoidStdioFiles + stdioFiles ifNotNil: + [stdioFiles do: + [:f| f isReadOnly ifFalse: [[f flush] on: Error do: [:ex|"care less"]]]. + stdioFiles := nil]! Item was added: + ----- Method: StandardFileStream class>>initialize (in category 'class initialization') ----- + initialize + "StandardFileStream initialize" + StdioHandles := Array new: 3. + Smalltalk + addToStartUpList: self after: SecurityManager; "the intent being before: AutoStart" + addToShutDownList: self after: SecurityManager! Item was added: + ----- Method: StandardFileStream class>>newForStdio (in category 'private-file creation') ----- + newForStdio + "Circumvent CrLfFileStream new's returning an instance of MultiBteFileStream" + ^self new! Item was added: + ----- Method: StandardFileStream class>>shutDown: (in category 'class initialization') ----- + shutDown: quitting + quitting ifTrue: + [self withAllSubclassesDo: + [:fileStreamClass| fileStreamClass flushAndVoidStdioFiles]]! Item was added: + ----- Method: StandardFileStream class>>standardIOStreamNamed:forWrite: (in category 'private-file creation') ----- + standardIOStreamNamed: moniker forWrite: forWrite + + | index | + index := #(stdin stdout stderr) identityIndexOf: moniker. + ^((stdioFiles ifNil: [ stdioFiles := Array new: 3 ]) at: index) + ifNil: [ + stdioFiles + at: index + put: ( + (StdioHandles at: index) + ifNil: [ ^self error: moniker, ' is unavailable' ] + ifNotNil: [ :handle| + self newForStdio + openOnHandle: handle + name: moniker + forWrite: forWrite ]) ]! Item was added: + ----- Method: StandardFileStream class>>startUp: (in category 'class initialization') ----- + startUp: resuming + (resuming and: [self == StandardFileStream]) ifTrue: + [self withAllSubclassesDo: + [:fileStreamClass| fileStreamClass voidStdioFiles]. + [StdioHandles := self stdioHandles] + on: Error + do: [:ex| + StdioHandles isArray ifFalse: + [StdioHandles := Array new: 3]]]! Item was added: + ----- Method: StandardFileStream class>>stderr (in category 'file creation') ----- + stderr + ^self standardIOStreamNamed: #stderr forWrite: true! Item was added: + ----- Method: StandardFileStream class>>stdin (in category 'file creation') ----- + stdin + ^self standardIOStreamNamed: #stdin forWrite: false! Item was added: + ----- Method: StandardFileStream class>>stdioHandles (in category 'private-file creation') ----- + stdioHandles + <primitive: 'primitiveFileStdioHandles' module: 'FilePlugin' error: ec> + self primitiveFailed! Item was added: + ----- Method: StandardFileStream class>>stdout (in category 'file creation') ----- + stdout + ^self standardIOStreamNamed: #stdout forWrite: true! Item was added: + ----- Method: StandardFileStream class>>voidStdioFiles (in category 'class initialization') ----- + voidStdioFiles + stdioFiles := nil! Item was added: + ----- Method: StandardFileStream>>openOnHandle:name:forWrite: (in category 'open/close') ----- + openOnHandle: aFileID name: streamName forWrite: writeMode + "Initialize the file with the given handle. If writeMode is true then + allow writing, otherwise put the file in read-only mode. + N.B. Do _not_ register the stream. We do not want it to be + closed implicitly (e.g. on GC). There may be multiple instances + accessing the same stream. The stream is not a file." + fileID := aFileID. + name := streamName. + rwmode := writeMode. + buffer1 := String new: 1. + self enableReadBuffering!

1 0

[Pharo-project] Issue 3238 in pharo: Use Class instead of TextDiffBuilder in changelist
by pharo＠googlecode.com Nov. 11, 2010

Nov. 11, 2010

Status: Accepted Owner: stephane.ducasse Labels: Type-Squeak New issue 3238 by stephane.ducasse: Use Class instead of TextDiffBuilder in changelist http://code.google.com/p/pharo/issues/detail?id=3238 Should check how to invoke this method - use ClassDiffBuilder for diffing class definitions and comments in ChangeList =============== Diff against Tools-ul.275 =============== Item was changed: ----- Method: ChangeList>>contentsDiffedFromCurrent (in category 'viewing access') ----- contentsDiffedFromCurrent "Answer the contents diffed forward from current (in-memory) method version" | aChange aClass | listIndex = 0 ifTrue: [^ '']. aChange := changeList at: listIndex. (aChange type == #method and: [(aClass := aChange methodClass) notNil and: [aClass includesSelector: aChange methodSelector]]) ifTrue: [^Utilities methodDiffFor: aChange text class: aClass selector: aChange methodSelector prettyDiffs: self showingPrettyDiffs]. aChange type == #doIt ifTrue: [| tokens | tokens := Scanner new scanTokens: aChange string. ((tokens select: [:substr| #(subclass: variableByteSubclass: variableWordSubclass: instanceVariableNames: classVariableNames: ) includes: substr]) asSet size >= 3 and: [(aClass := Smalltalk at: tokens third ifAbsent: []) notNil and: [aClass isBehavior]]) ifTrue: + [^ClassDiffBuilder buildDisplayPatchFrom: aClass definition to: aChange string]. - [^TextDiffBuilder buildDisplayPatchFrom: aClass definition to: aChange string]. (tokens size = 4 and: [tokens second == #class and: [tokens third == #instanceVariableNames: and: [(aClass := Smalltalk at: tokens first ifAbsent: []) notNil and: [aClass isBehavior]]]]) ifTrue: + [^ClassDiffBuilder buildDisplayPatchFrom: aClass class definition to: aChange string]]. - [^TextDiffBuilder buildDisplayPatchFrom: aClass class definition to: aChange string]]. (aChange type == #classComment and: [(aClass := aChange commentClass) notNil]) ifTrue: + [^ClassDiffBuilder buildDisplayPatchFrom: aClass comment asString to: aChange string]. - [^TextDiffBuilder buildDisplayPatchFrom: aClass comment asString to: aChange string]. ^(changeList at: listIndex) text!

1 0

[Pharo-project] Issue 3237 in pharo: If MessageTally is not in memory
by pharo＠googlecode.com Nov. 11, 2010

Nov. 11, 2010

Status: FixedWaitingToBePharoed Owner: stephane.ducasse Labels: Type-Squeak New issue 3237 by stephane.ducasse: If MessageTally is not in memory http://code.google.com/p/pharo/issues/detail?id=3237 we have to compare because the implementation in pharo is different. in pharo we do Smalltalk globals at: #MessageTally ifPresent: [ :c | c new close ]. ----- Method: Debugger>>process:controller:context: (in category 'private') ----- process: aProcess controller: aController context: aContext super initialize. + Smalltalk at: #MessageTally ifPresentAndInMemory: [ :tally | + tally terminateTimerProcess]. - Smalltalk at: #MessageTally ifPresent:[:tally| tally terminateTimerProcess]. contents := nil. interruptedProcess := aProcess. interruptedController := aController. contextStackTop := aContext. self newStack: (contextStackTop stackOfSize: 1). contextStackIndex := 1. externalInterrupt := false. selectingPC := true. Smalltalk isMorphic ifTrue: [errorWasInUIProcess := false]!

1 0

[Pharo-project] Issue 3236 in pharo: Cuis DateAndTime Enhancement
by pharo＠googlecode.com Nov. 11, 2010

Nov. 11, 2010

Status: Accepted Owner: stephane.ducasse Labels: Type-Squeak New issue 3236 by stephane.ducasse: Cuis DateAndTime Enhancement http://code.google.com/p/pharo/issues/detail?id=3236 Levente Uzonyi uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-ul.512.mcz ==================== Summary ==================== Name: Kernel-ul.512 Author: ul Time: 7 November 2010, 12:05:13.149 am UUID: f35cd221-68a5-f04b-b6bc-d07df326b3d3 Ancestors: Kernel-nice.511 - DateAndTime startup enhancement from Cuis part 1 =============== Diff against Kernel-nice.511 =============== Item was changed: Magnitude subclass: #DateAndTime instanceVariableNames: 'seconds offset jdn nanos' + classVariableNames: 'ClockProvider DaysSinceEpoch LastMilliSeconds LastTick LastTickSemaphore LocalTimeZone MilliSecondOffset OffsetsAreValid' - classVariableNames: 'ClockProvider DaysSinceEpoch LastMilliSeconds LastTick LastTickSemaphore LocalTimeZone MilliSecondOffset' poolDictionaries: 'ChronologyConstants' category: 'Kernel-Chronology'! !DateAndTime commentStamp: 'brp 5/13/2003 08:07' prior: 0! I represent a point in UTC time as defined by ISO 8601. I have zero duration. My implementation uses three SmallIntegers and a Duration: jdn - julian day number. seconds - number of seconds since midnight. nanos - the number of nanoseconds since the second. offset - duration from UTC. The nanosecond attribute is almost always zero but it defined for full ISO compliance and is suitable for timestamping. ! Item was added: + (PackageInfo named: 'Kernel') postscript: 'DateAndTime classPool at: #OffsetsAreValid put: true'!

1 1

[Pharo-project] Issue 3234 in pharo: fold: reduce: Merge
by pharo＠googlecode.com Nov. 11, 2010

Nov. 11, 2010

Status: FixedWaitingToBePharoed Owner: stephane.ducasse Labels: Type-Squeak New issue 3234 by stephane.ducasse: fold: reduce: Merge http://code.google.com/p/pharo/issues/detail?id=3234 I imagine that reduce: and fold: are not the same in pharo so we should pay attention. Name: Collections-ul.403 Author: ul Time: 7 November 2010, 1:55:55.898 am UUID: 7f81b384-8004-cd44-81e0-9b294d999c8a Ancestors: Collections-ul.402 - merged Collection's #fold: and #reduce: (it's a bit faster than the original versions :)) Item was changed: ----- Method: Collection>>fold: (in category 'enumerating') ----- fold: binaryBlock "Evaluate the block with the first two elements of the receiver, then with the result of the first evaluation and the next element, and so on. Answer the result of the final evaluation. If the receiver is empty, raise an error. If the receiver has a single element, answer that element." "#('if' 'it' 'is' 'to' 'be' 'it' 'is' 'up' 'to' 'me') fold: [:a :b | a, ' ', b]" + ^self reduce: binaryBlock! - | firstValue nextValue | - firstValue := nextValue := Object new. "something that can't be in the receiver" - self do: - [:each | - nextValue := firstValue == nextValue - ifTrue: [each] - ifFalse: [binaryBlock value: nextValue value: each]]. - ^nextValue == firstValue - ifTrue: [self errorEmptyCollection] - ifFalse: [nextValue]! Item was changed: ----- Method: Collection>>reduce: (in category 'enumerating') ----- reduce: binaryBlock "Apply the argument, binaryBlock cumulatively to the elements of the receiver. For sequenceable collections the elements will be used in order, for unordered collections the order is unspecified." | first nextValue | - self emptyCheck. first := true. + self do: [ :each | + first + ifTrue: [ nextValue := each. first := false ] + ifFalse: [ nextValue := binaryBlock value: nextValue value: each ] ]. + first ifTrue: [ self errorEmptyCollection ]. - self do:[:each| - first ifTrue:[nextValue := each. first := false] - ifFalse:[nextValue := binaryBlock value: nextValue value: each]]. ^nextValue!

1 0

[Pharo-project] Issue 3233 in pharo: Tests for WeakSet
by pharo＠googlecode.com Nov. 11, 2010

Nov. 11, 2010

Status: FixedWaitingToBePharoed Owner: stephane.ducasse Labels: Type-Squeak New issue 3233 by stephane.ducasse: Tests for WeakSet http://code.google.com/p/pharo/issues/detail?id=3233 Levente Uzonyi uploaded a new version of CollectionsTests to project The Trunk: http://source.squeak.org/trunk/CollectionsTests-ul.175.mcz ==================== Summary ==================== Name: CollectionsTests-ul.175 Author: ul Time: 7 November 2010, 3:17:19.921 am UUID: 4b82952d-1507-5249-ba8a-9f93a4aaa9f9 Ancestors: CollectionsTests-ul.174 - integrated tests for WeakSet from http://bugs.squeak.org/view.php?id=7350 =============== Diff against CollectionsTests-ul.174 =============== Item was added: + ----- Method: WeakSetTest>>testAddIncludesSizeReclaim (in category 'testing') ----- + testAddIncludesSizeReclaim + | ws o2 o3 | + o2 := 1/2. + o3 := '123' copy. + ws := WeakSet new. + ws add: o2. + ws add: o3. + self assert: ws size = 2. + self assert: (ws includes: o2). + self assert: (ws includes: o3). + + "inclusion test does use equality, not identity" + self assert: (ws includes: o3 copy) description: 'WeakSet are not WeakIdentitySet'. + + "only one copy is added" + ws add: o3 copy. + self assert: ws size = 2. + + "reclame objects so that slots of ws are nilled out" + o2 := o3 := nil. + Smalltalk garbageCollect. + self deny: (ws includes: 1/2). + self deny: (ws includes: '123' copy). + + "fast #size is not updated automatically by dead object reclamation + But there is a slowSize trying to tell the truth" + self assert: ws slowSize = 0.! Item was added: + ----- Method: WeakSetTest>>testCollisions (in category 'testing') ----- + testCollisions + + | ws o1 o2 o5 on remember forget | + + "create a weak set" + ws := WeakSet new: 15. + + "select some fractions wanting same place in ws array" + o1 := (2 to: 100) select: [:i | (ws scanFor: 1 / i) == 1]. + o2 := (2 to: 100) select: [:i | (ws scanFor: 1 / i) == 2]. + o5 := (2 to: 100) select: [:i | (ws scanFor: 1 / i) == 5]. + on := (2 to: 100) select: [:i | (ws scanFor: 1 / i) == (ws array size - 1)]. + + "Add some fractions to the weak set, and remember a pointer for a few of them" + remember := OrderedCollection new. + forget := OrderedCollection new. + ws add: (remember add: 1 / o1 first). + ws add: (forget add: 1 / on second). + ws add: (forget add: 1 / o1 second). + ws add: (forget add: 1 / o5 second). + ws add: (forget add: 1 / o2 second). + ws add: (forget add: 1 / o1 third). + ws add: (remember add: 1 / o2 first). + ws add: (forget add: 1 / o5 third). + ws add: (forget add: 1 / on third). + ws add: (remember add: 1 / o2 fourth). + ws add: (remember add: 1 / on first). + ws add: (remember add: 1 / o5 first). + + "forget and reclaim all entries but those in remember" + forget := nil. + Smalltalk garbageCollect. + + remember do: [:m | self assert: (ws includes: m)]. + ws add: 1/on second. + remember do: [:m | self assert: (ws includes: m)]. + + ws add: (remember add: 1 / o1 fourth). + ws add: (remember add: 1 / on fourth). + remember remove: (ws remove: (1 / o5 first)). + remember remove: (ws remove: (1 / on first)). + remember remove: (ws remove: (1 / o2 first)). + remember remove: (ws remove: (1 / o1 first)). + remember do: [:m | self assert: (ws includes: m)]. + ws add: 1/on second. + ws add: 1/o5 second. + remember do: [:m | self assert: (ws includes: m)]. + ! Item was added: + ----- Method: WeakSetTest>>testDoDontIterateOnReclaimedObjects (in category 'testing') ----- + testDoDontIterateOnReclaimedObjects + + | ws o1 o2 size | + ws := WeakSet new. + 2 to: 20 do: [:i | ws add: 1 / i]. + + o1 := 3.0. + o2 := 4.0. + ws add: o1; add: o2. + + "Reclaim memory" + Smalltalk garbageCollect. + + size := 0. + ws do: [:each | + size := size + 1. + self deny: each isNil]. + + self assert: size = 2! Item was added: + ----- Method: WeakSetTest>>testGrowWhenNecessary (in category 'testing') ----- + testGrowWhenNecessary + | ws objs initialSize | + objs := (2 to: 20) collect: [:i | 1 / i]. + ws := WeakSet new: 5. + initialSize := ws array size. + 1 to: objs size do: [:k | ws add: (objs at: k)]. + self assert: ws array size > initialSize + description: 'The WeakSet grown because not enough room were preallocated'.! Item was added: + ----- Method: WeakSetTest>>testIncludesNil (in category 'testing') ----- + testIncludesNil + + | ws | + ws := WeakSet new. + self deny: (ws includes: nil). + + "After reclamation, should not includes nil: nil counts for nothing" + ws add: 1/2. + ws add: 1/3. + Smalltalk garbageCollect. + self deny: (ws includes: nil).!

1 0

[Pharo-project] Issue 3232 in pharo: SharedQueue code simplication
by pharo＠googlecode.com Nov. 11, 2010

Nov. 11, 2010

Status: FixedWaitingToBePharoed Owner: stephane.ducasse Labels: Type-Squeak New issue 3232 by stephane.ducasse: SharedQueue code simplication http://code.google.com/p/pharo/issues/detail?id=3232 We should really review carefully this changes A new version of Collections was added to project The Inbox: http://source.squeak.org/inbox/Collections-ul.404.mcz ==================== Summary ==================== Name: Collections-ul.404 Author: ul Time: 7 November 2010, 4:52:11.401 am UUID: 2e0defd1-6e46-7243-be6f-b571c6805ae4 Ancestors: Collections-ul.403 - simplified SharedQueue >> #makeRoomAtEnd - enhanced SharedQueue >> #peek =============== Diff against Collections-ul.403 =============== Item was changed: ----- Method: SharedQueue>>makeRoomAtEnd (in category 'private') ----- makeRoomAtEnd + | contentsSize newContentsArray | - | contentsSize | contentsSize := writePosition - readPosition. + newContentsArray := contentsSize * 2 > contentsArray size + ifTrue: [ contentsArray class new: contentsArray size * 2 ] - contentsSize * 2 > contentsArray size - ifTrue: [ - "grow" - contentsArray := (contentsArray class new: contentsArray size * 2) - replaceFrom: 1 - to: contentsSize - with: contentsArray - startingAt: readPosition; - yourself ] ifFalse: [ (contentsArray size > 10 and: [ contentsSize * 4 <= contentsArray size ]) + ifTrue: [ contentsArray class new: (contentsSize * 2 max: 10) ] + ifFalse: [ contentsArray ] ]. + newContentsArray + replaceFrom: 1 + to: contentsSize + with: contentsArray + startingAt: readPosition. + contentsArray == newContentsArray + ifFalse: [ contentsArray := newContentsArray ] + ifTrue: [ contentsArray from: contentsSize + 1 to: contentsArray size put: nil ]. - ifTrue: [ - "shrink" - contentsArray := (contentsArray class new: (contentsSize * 2 max: 10)) - replaceFrom: 1 - to: contentsSize - with: contentsArray - startingAt: readPosition; - yourself ] - ifFalse: [ - "just move the elements to the front" - contentsArray - replaceFrom: 1 - to: contentsSize - with: contentsArray - startingAt: readPosition. - contentsArray - from: contentsSize + 1 - to: contentsArray size - put: nil ] ]. readPosition := 1. writePosition := contentsSize + 1! Item was changed: ----- Method: SharedQueue>>peek (in category 'accessing') ----- peek "Answer the object that was sent through the receiver first and has not yet been received by anyone but do not remove it from the receiver. If no object has been sent, return nil" + readSynch isSignaled ifFalse: [ ^nil ]. + ^accessProtect critical: [ + readPosition >= writePosition ifFalse: [ + contentsArray at: readPosition ] ]! - ^accessProtect - critical: [ - | value | - readPosition >= writePosition - ifTrue: [readPosition := 1. - writePosition := 1. - value := nil] - ifFalse: [value := contentsArray at: readPosition]. - value].!

1 0

[Pharo-project] Issue 3230 in pharo: Fixes for Latin1Environment from EToys
by pharo＠googlecode.com Nov. 11, 2010

Nov. 11, 2010

Status: FixedWaitingToBePharoed Owner: stephane.ducasse Labels: Type-Squeak New issue 3230 by stephane.ducasse: Fixes for Latin1Environment from EToys http://code.google.com/p/pharo/issues/detail?id=3230 ***Need to be really verified if this applies to pharo*** Name: Multilingual-sn.131 Author: sn Time: 8 November 2010, 12:54:12.262 am UUID: 64aa8ddf-9938-4373-a45d-7bde1c0ccced Ancestors: Multilingual-ul.130 Fixes for Latin1Environment from EToys =============== Diff against Multilingual-ul.130 =============== Item was changed: ----- Method: Latin1Environment class>>fileNameConverterClass (in category 'subclass responsibilities') ----- fileNameConverterClass + ^ self systemConverterClass - ^ Latin1TextConverter ! Item was changed: ----- Method: Latin1Environment class>>inputInterpreterClass (in category 'subclass responsibilities') ----- inputInterpreterClass | platformName osVersion | platformName := SmalltalkImage current platformName. osVersion := SmalltalkImage current osVersion. (platformName = 'Win32' and: [osVersion ~= 'CE']) ifTrue: [^ (self win32VMUsesUnicode) ifTrue: [UTF32InputInterpreter] ifFalse: [MacRomanInputInterpreter]]. platformName = 'Mac OS' + ifTrue: [^ MacUnicodeInputInterpreter]. - ifTrue: [^ MacRomanInputInterpreter]. platformName = 'unix' ifTrue: [^ UTF32InputInterpreter]. + ^ MacUnicodeInputInterpreter! - ^ MacRomanInputInterpreter! Item was changed: ----- Method: Latin1Environment class>>systemConverterClass (in category 'subclass responsibilities') ----- systemConverterClass + | platformName osVersion | + platformName := SmalltalkImage current platformName. + osVersion := SmalltalkImage current getSystemAttribute: 1002. + (platformName = 'Win32' + and: [osVersion = 'CE']) + ifTrue: [^ MacRomanTextConverter]. + platformName = 'Win32' + ifTrue: [^ (self win32VMUsesUnicode) ifTrue: [UTF8TextConverter] ifFalse: [Latin1TextConverter]]. + platformName = 'Mac OS' + ifTrue: [^ ('10*' match: SmalltalkImage current osVersion) + ifTrue: [UTF8TextConverter] + ifFalse: [MacRomanTextConverter]]. + platformName = 'unix' + ifTrue: [^ UTF8TextConverter]. + ^ MacRomanTextConverter! - ^ Latin1TextConverter. - ! Item was changed: ----- Method: MacUnicodeInputInterpreter>>initialize (in category 'initialization') ----- initialize | satisfiesVersion | satisfiesVersion := self majorMinorBuildFrom: SmalltalkImage current vmVersion + satisfies: [:major :minor :build | + major >= 4 or: [ + major >= 3 and: [minor >= 8 and: [build >= 7]]]]. - satisfies: [:major :minor :build | (major >= 3 - and: [minor >= 8 - and: [build >= 7]]) or: [major >= 4]]. - satisfiesVersion ifTrue: [keyValueIndex := 6] ifFalse: [keyValueIndex := 3]!

1 0

[Pharo-project] Issue 3228 in pharo: fixed font rendering to support translucent colors
by pharo＠googlecode.com Nov. 11, 2010

Nov. 11, 2010

Status: FixedWaitingToBePharoed Owner: stephane.ducasse Labels: Type-Squeak New issue 3228 by stephane.ducasse: fixed font rendering to support translucent colors http://code.google.com/p/pharo/issues/detail?id=3228 To be verified Name: Graphics-mt.152 Author: mt Time: 10 November 2010, 4:14:35.129 pm UUID: 04582fa8-3ca1-e440-987a-74483e96950d Ancestors: Graphics-dtl.151 - fixed font rendering to support translucent colors =============== Diff against Graphics-dtl.151 =============== Item was changed: ----- Method: BitBlt>>installStrikeFont:foregroundColor:backgroundColor: (in category 'private') ----- installStrikeFont: aStrikeFont foregroundColor: foregroundColor backgroundColor: backgroundColor | lastSourceDepth targetColor | sourceForm ifNotNil:[lastSourceDepth := sourceForm depth]. sourceForm := aStrikeFont glyphs. "Ignore any halftone pattern since we use a color map approach here" halftoneForm := nil. sourceY := 0. height := aStrikeFont height. sourceForm depth = 1 ifTrue: [ self combinationRule: Form paint. (colorMap notNil and:[lastSourceDepth = sourceForm depth]) ifFalse: [ "Set up color map for a different source depth (color font)" "Uses caching for reasonable efficiency" colorMap := self cachedFontColormapFrom: sourceForm depth to: destForm depth. colorMap at: 1 put: (destForm pixelValueFor: backgroundColor)]. colorMap at: 2 put: (destForm pixelValueFor: foregroundColor). ] ifFalse: [ (Preferences subPixelRenderFonts and: [ foregroundColor = Color black or: [ Preferences subPixelRenderColorFonts ]]) ifTrue: [ destForm depth > 8 ifTrue: [ + "Only force #rgbMul if color is not translucent for better result quality. Alpha values are ignored with #rgbMul." + foregroundColor isTranslucent + ifFalse: [self combinationRule: Form rgbMul]. - "rgbMul is equivalent to component alpha blend if text is black (only faster, hehe)" - self combinationRule: 37. "RGBMul" colorMap := (destForm depth = 32 or: [ (foregroundColor = Color black) not ]) ifTrue: [ "rgbMul / rgbAdd IS component alpha blend for any color of text (neat trick, eh!!)" "This colorMap is to be used on the second pass with rule 20 (rgbAdd) See #displayString:from:to:at:strikeFont:kern:" "Note: In 32bpp we always need the second pass, as the source could have transparent pixels, and we need to add to the alpha channel" self colorConvertingMap: foregroundColor from: sourceForm depth to: destForm depth keepSubPixelAA: true]] ifFalse: [ self combinationRule: 25. "Paint" targetColor := foregroundColor = Color black ifFalse: [ foregroundColor ]. colorMap := self colorConvertingMap: targetColor from: sourceForm depth to: destForm depth keepSubPixelAA: true] ] ifFalse: [ "Do not use rule 34 for 16bpp display. TTCFont uses it, but it builds a glyphs cache for each color used!!" self combinationRule: (destForm depth = 32 ifTrue: [34 "alphaBlendScaled"] ifFalse: [25 "Paint"]). colorMap := self colorConvertingMap: foregroundColor from: sourceForm depth to: destForm depth keepSubPixelAA: false ] ].!

1 0

Nov. 11, 2010

On Nov 11, 2010, at 5:14 PM, Tony Fleig wrote: > Hi, > > I find myself wanting a Character>>isConsonant method in addition to the existing Character>>isVowel method. It might be defined as > > Character>>isConsonant > ^ self isLetter & self isVowel not > > My question is: what is the best way to do this? > > Should I simply add the method to Character? If I do, I guess I must remember to merge my Monticello changes into each new version of Character? Is there some way to bundle this change in my package so I don't forget? but if you add this method in the character class as a class extension of your package then the addition will be done automatically each time you load it. > > I have created a method in my class MyClass>>characterIsConsonant: aCharacter, but it doesn't read well in the code. > > I suppose I could add the method dynamically in my class's initialize method... Not sure how to do that exactly. > > Is there another way to do this that I haven't thought of? > > Thanks, > TF

2 1