Squeak-dev May 2022

squeak-dev@lists.squeakfoundation.org

37 participants
316 discussions

PSA | Slow .changes access | Real-time protection in Windows Defender
by Marcel Taeumel 25 Sep '22

25 Sep '22

Hi all -- Provided that you have a .changes file that is bigger than a couple of bytes (here: ~200 MiB), you might experience a serious lag (here: ~5 seconds) when: - Starting the .image - Evaluating some do-its (or print-its) - Saving a method All operations that typically read from or write to the .changes file. The Windows Defender's "real-time protection service" seems to monitor file access. Whatever triggers it, the defender might try to get exclusive read access to a file. Then it scans the last bytes and if it finds something strange, it will continue, eventually scanning THE ENTIRE FILE! So, what is "something strange"? Well, try to evaluate this: '<a href="https://www.microsoft.com"></a>' size. Now, depending on the size of your .changes file, the Windows Defender will begin to scan the entire file for more such patterns. Note that the Defender will never complain or blacklist that file. It is just curious for the moment. Why is the Defender able to get exclusive read access on a do-it? Take a look at SmalltalkImage >> #forceChangesToDisk. There, you can see that we close and re-open the .changes file. That's exactly the time when the Windows Defender kicks in. Avoid closing that file, and you will not experience any lag. Try to save a method with such contents. You will be able to observe the same amount of extra lag. Why is regular read access during code browsing not affected? Because we already have the open file handle to the .changes and .sources files. The Defender seems to need exclusive read access. In my experience, this happened, for example, during VM/image startup. (Note that this has nothing to do with the Defender sending stuff to Microsoft via Internet. I did all the tests offline. No Ethernet connection. No WiFi.) *** What can we do about it? Different approaches work: - Save a method (or do-it sth.) to add about 1000 bytes of uninteresting content to the .changes file; then everything is fast again - Shrink your .changes file via #condenseChanges (and don't forget to save your image afterwards!!) to speed up the Defender's scan - Disable the "real-time protection service" in the Windows settings (not recommended) - Avoid strange content in your source code (or do-its) - Do not close the file so often. #primFlush: is not affected. Best, Marcel

4 5

The Trunk: Tools-mt.1087.mcz
by commits＠source.squeak.org 20 Sep '22

20 Sep '22

Marcel Taeumel uploaded a new version of Tools to project The Trunk: http://source.squeak.org/trunk/Tools-mt.1087.mcz ==================== Summary ==================== Name: Tools-mt.1087 Author: mt Time: 15 December 2021, 4:52:00.953892 pm UUID: 627887b0-5064-ab43-b7ac-60b329a8b606 Ancestors: Tools-mt.1086 Tweak BlockClosureInspector to show all the closured stuff including temps from the outer context and inst-vars from the current receiver. Adds some caching to make it faster. Support cache invalidation only for changing the #receiver field. Not sure whether it makes sense to support changing the outerContext as well... =============== Diff against Tools-mt.1086 =============== Item was changed: Inspector subclass: #BlockClosureInspector + instanceVariableNames: 'debuggerMap tempVarNames instVarNames' - instanceVariableNames: '' classVariableNames: '' poolDictionaries: '' category: 'Tools-Inspector'! Item was added: + ----- Method: BlockClosureInspector>>debuggerMap (in category 'private') ----- + debuggerMap + + ^ debuggerMap ifNil: [debuggerMap := self object method debuggerMap]! Item was changed: ----- Method: BlockClosureInspector>>hasBindingOf: (in category 'user interface - styling') ----- hasBindingOf: aString self object method literalsDo: [:literal | (literal isVariableBinding and: [literal key = aString]) ifTrue: [^ true]]. + + (self tempVarNames includes: aString) ifTrue: [^ true]. + (self instVarNames includes: aString) ifTrue: [^ true]. + + ^ false! - ^ self object outerContext tempNames includes: aString! Item was added: + ----- Method: BlockClosureInspector>>instVarNames (in category 'private') ----- + instVarNames + + ^ instVarNames ifNil: [instVarNames := (self object receiver perform: #class "avoid inlining; support proxies") allInstVarNames]! Item was added: + ----- Method: BlockClosureInspector>>replaceSelectionValue: (in category 'selection') ----- + replaceSelectionValue: anObject + + | success | + (success := super replaceSelectionValue: anObject) ifTrue: [ + self selectedField key = #receiver + ifTrue: [instVarNames := nil]]. + ^ success! Item was changed: ----- Method: BlockClosureInspector>>streamBindingsOn: (in category 'fields - streaming') ----- streamBindingsOn: aStream + | compiledCode instNames tempNames | - | compiledCode tempNames | self flag: #decompile. "mt: Use #to: and #do: instead of #to:do: to avoid inlining to preserve bindings in enumeration block for later decompilation. See InspectorField." + + (1 to: (instNames := self instVarNames) size) do: [:index | + aStream nextPut: ((self newFieldForType: #instVar key: index) + name: ('[[[{1}]]]' format: {instNames at: index}); + valueGetter: [:closure | closure receiver instVarAt: index]; + valueSetter: [:closure :value | closure receiver instVarAt: index put: value]; + yourself)]. + + tempNames := self tempVarNames. + "No need to show closured bindings from outerContext. The important temps are already covered in the indexed fields below." + (1 to: tempNames size - self object basicSize) do: [:index | + aStream nextPut: ((self newFieldForType: #tempVar key: index) + name: ('[[{1}]]' format: {tempNames at: index}); + valueGetter: [:closure | closure outerContext tempAt: index]; + valueSetter: [:closure :value | closure outerContext tempAt: index put: value]; + yourself)]. + (1 to: (compiledCode := self object method) numLiterals) do: [:index | | literal | literal := compiledCode literalAt: index. literal isVariableBinding ifTrue: [ aStream nextPut: ((self newFieldForType: #tempVar key: index) name: ('[{1}]' format: {literal key}); valueGetter: [:closure | (closure method literalAt: index) value]; valueSetter: [:closure :value | (closure method literalAt: index) value: value]; yourself)]]. + - - tempNames := self object outerContext tempNames. - "No need to show closured bindings from outerContext. The important temps are already covered in the indexed fields below." - "(1 to: tempNames size - self object basicSize) do: [:index | - aStream nextPut: ((self newFieldForType: #tempVar key: index) - name: ('[{1}]' format: {tempNames at: index}); - valueGetter: [:closure | closure outerContext tempAt: index]; - valueSetter: [:closure :value | closure outerContext tempAt: index put: value]; - yourself)]." - (1 to: self object basicSize) do: [:index | aStream nextPut: ((self newFieldForType: #tempVar key: index) name: ('[{1}]' format: {tempNames at: index + tempNames size - self object basicSize}); valueGetter: [:closure | closure basicAt: index]; valueSetter: [:closure :value | closure basicAt: index put: value]; yourself)]. "See commentary in #doItReceiver. No need to expose the closured 'self' here because there is already the #receiver instVar. fieldReceiver ^ (self newFieldForType: #tempVar key: #receiver) name: '[self]'; valueGetter: [:obj | obj receiver]; valueSetter: [:obj :value | obj receiver: value]; yourself "! Item was added: + ----- Method: BlockClosureInspector>>tempVarNames (in category 'private') ----- + tempVarNames + + ^ tempVarNames ifNil: [tempVarNames := self debuggerMap tempNamesForContext: self object outerContext]!

2 2

Difference between Tools and Apps menu?
by Jaromir Matas 15 Sep '22

15 Sep '22

Hi All, another quick question: What is the difference between Tools and Apps menu? Version 4 introduced Tools menu, version 5 added Apps menu. I can't figure out what the difference is other than Apps are "More Tools". But then why would some entries be in both? So it's probably not it... Or is it about frequency, importance? And then there's "Preferences" in Tools and the same tool named "Preference Browser" in Apps which is also confusing. Thanks for any advice :) -- Jaromír Matas mail(a)jaromir.net

5 5

Installer inbox cleanup
by Tim Johnson 15 Sep '22

15 Sep '22

Hi, Can someone please delete my inbox contribution, Installer-Core-tcj-440.mcz from August 2020? It is superseded by Installer-Core-lrnp.447.mcz which is (I believe) the same intention but better executed. Thanks, Tim

2 1

The Trunk: System-tonyg.1224.mcz
by commits＠source.squeak.org 14 Sep '22

14 Sep '22

Tony Garnock-Jones uploaded a new version of System to project The Trunk: http://source.squeak.org/trunk/System-tonyg.1224.mcz ==================== Summary ==================== Name: System-tonyg.1224 Author: tonyg Time: 29 March 2021, 10:15:02.279821 am UUID: 9d897d72-46fc-4205-a27b-89c7a50ac104 Ancestors: System-mt.1223 Teach SystemOrganizer about multiple environments. =============== Diff against System-mt.1223 =============== Item was changed: Categorizer subclass: #SystemOrganizer + instanceVariableNames: 'environment' - instanceVariableNames: '' classVariableNames: '' poolDictionaries: '' category: 'System-Support'! !SystemOrganizer commentStamp: '<historical>' prior: 0! My instances provide an organization for the classes in the system, just as a ClassOrganizer organizes the messages within a class. The only difference is the methods for fileIn/Out.! Item was changed: ----- Method: SystemOrganizer>>classesIn: (in category 'query') ----- classesIn: categoryName | classes | classes := OrderedCollection new. self categories withIndexCollect: [:cat :idx | (categoryName match: cat) ifTrue: [classes addAll: (self listAtCategoryNumber: idx)] ifFalse: [nil]]. + ^ classes collect: [:clsName | self environment classNamed: clsName]! - ^ classes collect: [:clsName | Smalltalk classNamed: clsName]! Item was changed: ----- Method: SystemOrganizer>>classify:under: (in category 'accessing') ----- classify: element under: newCategory | oldCategory class | self flag: #environments. "do we want notifications for classes in other environments?" oldCategory := self categoryOfElement: element. super classify: element under: newCategory. + class := self environment at: element ifAbsent: [^ self]. - class := Smalltalk at: element ifAbsent: [^ self]. self == SystemOrganization ifTrue: [ SystemChangeNotifier uniqueInstance class: class recategorizedFrom: oldCategory to: newCategory]! Item was changed: ----- Method: SystemOrganizer>>commentInventory: (in category 'query') ----- commentInventory: categoryName "SystemOrganization commentInventory: 'Morphic*'" | classes commentedClasses | classes := OrderedCollection new. self categories withIndexCollect: [:cat :idx | (categoryName match: cat) ifTrue: [classes addAll: (self listAtCategoryNumber: idx)] ifFalse: [nil]]. + commentedClasses := classes select: [:catCls | (self environment at: catCls) hasComment]. - commentedClasses := classes select: [:catCls | (Smalltalk at: catCls) hasComment]. ^ 'There are ' , classes size asString , ' classes in ' , categoryName , ' of which ' , commentedClasses size asString , ' have comments and ', (classes size - commentedClasses size) asString , ' do not yet have comments.' ! Item was added: + ----- Method: SystemOrganizer>>environment (in category 'accessing') ----- + environment + ^ environment ifNil: [Smalltalk globals]! Item was added: + ----- Method: SystemOrganizer>>environment: (in category 'accessing') ----- + environment: anEnvironment + environment := anEnvironment! Item was changed: ----- Method: SystemOrganizer>>fileOutCategory:on:initializing: (in category 'fileIn/Out') ----- fileOutCategory: category on: aFileStream initializing: aBool "Store on the file associated with aFileStream, all the traits and classes associated with the category and any requested shared pools in the right order." | first poolSet tempClass classes traits | traits := self orderedTraitsIn: category. classes := self superclassOrder: category. poolSet := Set new. classes do: [:class | class sharedPools do: [:eachPool | poolSet add: eachPool]]. poolSet size > 0 ifTrue: [ tempClass := Class new. tempClass shouldFileOutPools ifTrue: [ poolSet := poolSet select: [:aPool | + tempClass shouldFileOutPool: (self environment keyAtIdentityValue: aPool)]. - tempClass shouldFileOutPool: (Smalltalk globals keyAtIdentityValue: aPool)]. poolSet do: [:aPool | tempClass fileOutPool: aPool onFileStream: aFileStream]]]. first := true. traits, classes do: [:each | first ifTrue: [first := false] ifFalse: [aFileStream cr; nextPut: Character newPage; cr]. each fileOutOn: aFileStream moveSource: false toFile: 0 initializing: false]. aBool ifTrue: [classes do: [:cls | cls fileOutInitializerOn: aFileStream]].! Item was changed: ----- Method: SystemOrganizer>>orderedTraitsIn: (in category 'fileIn/Out') ----- orderedTraitsIn: category "Answer an OrderedCollection containing references to the traits in the category whose name is the argument, category (a string). The traits are ordered so they can be filed in." | behaviors traits | behaviors := (self listAtCategoryNamed: category asSymbol) + collect: [:title | self environment at: title]. - collect: [:title | Smalltalk at: title]. traits := behaviors reject: [:each | each isBehavior]. ^traits asArray sort: [:t1 :t2 | (t2 traitComposition allTraits includes: t1) or: [(t1 traitComposition allTraits includes: t2) not]]! Item was changed: ----- Method: SystemOrganizer>>removeMissingClasses (in category 'remove') ----- removeMissingClasses "Remove any class names that are no longer in the Smalltalk dictionary. Used for cleaning up after garbage collecting user-generated classes." "SystemOrganization removeMissingClasses" elementArray copy do: [:el | + (self environment includesKey: el) ifFalse: [self removeElement: el]]. - (Smalltalk includesKey: el) ifFalse: [self removeElement: el]]. ! Item was changed: ----- Method: SystemOrganizer>>superclassOrder: (in category 'fileIn/Out') ----- superclassOrder: category "Answer an OrderedCollection containing references to the classes in the category whose name is the argument, category (a string). The classes are ordered with superclasses first so they can be filed in." | behaviors classes | behaviors := (self listAtCategoryNamed: category asSymbol) + collect: [:title | self environment at: title]. - collect: [:title | Smalltalk at: title]. classes := behaviors select: [:each | each isBehavior]. ^ChangeSet superclassOrder: classes!

5 7

ResumableTestFailure
by Thiede, Christoph 13 Sep '22

13 Sep '22

Hi all! :-) I was just working a bit with Squeak by Example (the book) and found out that resumable assertions as described in the book (see chapter SUnit, Advanced, Continuing after a failure) do not any longer work in a trunk image. If you write a test case with some resumable assertions, turn on logging, and run the test, only the first failure will be logged, not the rest, too. (Methods to reproduce: isLogging ^ true "You will need to open a Transcript" testResumable #(1 2 3 4 5) do: [:i | self assert: i even description: ['{1} is not even' format: {i}] resumable: true] ) A second issue I experienced frequently when I felt the desire to work with resumable assertions was that they are basically only available for two selectors: #assert:description:resumable: and #deny:description:identical:. I often would like to insert a resumable into any test without being that verbose, and nearly redoubling the number of assertion selectors by adding a resumable argument to each of them did not feel right, either. Please find the attached changeset, which fixes both issues. * In TestResult >> #runCase:, resumable assertions are identified and - finally - resumed, after marking the test as failed. Unfortunately, other test runners (yes, I'm thinking of smalltalkCI) will need to reimplement this feature unless they reuse the TestResult implementation. * In TestCase, I refactored some methods that raise assertion failures and added the most basic #signalFailure:resumable: as the single source of signalization. Instead of setting the resumable bit manually for each assertion, you now can also make the whole test execution resumable by: * overriding #isResumable, * setting a resumable flag using #resumable:, * setting a resumable flag temporarily using #beResumable:during:, or * adding a <resumable> or <resumable: true> pragma to the test method. * In addition, I tested the added and fixed behavior in ResumableTestFailureTestCase. Looking forward to your review! :-) Best, Christoph

2 2

The Inbox: Kernel-ct.1294.mcz
by commits＠source.squeak.org 09 Sep '22

09 Sep '22

A new version of Kernel was added to project The Inbox: http://source.squeak.org/inbox/Kernel-ct.1294.mcz ==================== Summary ==================== Name: Kernel-ct.1294 Author: ct Time: 9 January 2020, 3:42:36.672555 pm UUID: 7409dcc3-160e-9d44-9712-f266e0f25973 Ancestors: Kernel-nice.1292 Proposal: Add #haltOnceIf: to Object, which combines the strengths of #haltOnce and #haltIf:. Especially useful in performance-critical situations: It HaltOnce is disabled, the condition will not be evaluated at all. =============== Diff against Kernel-nice.1292 =============== Item was changed: ----- Method: Object>>haltIf: (in category 'debugging') ----- + haltIf: aCondition + "This is the typical message to use for inserting breakpoints during debugging. Param can be a block or expression, halt if true. - haltIf: condition - "This is the typical message to use for inserting breakpoints during - debugging. Param can be a block or expression, halt if true. If the Block has one arg, the receiver is bound to that. + If the condition is a selector, we look up in the callchain. Halt if any method's selector equals selector." - If the condition is a selector, we look up in the callchain. Halt if - any method's selector equals selector." - | cntxt | + ^ (self meetsHaltCondition: aCondition) + ifTrue: [self halt]! - condition isSymbol ifTrue:[ - "only halt if a method with selector symbol is in callchain" - cntxt := thisContext. - [cntxt sender isNil] whileFalse: [ - cntxt := cntxt sender. - (cntxt selector = condition) ifTrue: [Halt signal]. - ]. - ^self. - ]. - (condition isBlock - ifTrue: [condition cull: self] - ifFalse: [condition] - ) ifTrue: [ - Halt signal - ].! Item was added: + ----- Method: Object>>haltOnceIf: (in category 'debugging') ----- + haltOnceIf: aCondition + "Check aCondition and halt, unless we have already done it once. See #meetsHaltCondition:." + + self haltOnceEnabled ifFalse: [^ self]. + ^ (self meetsHaltCondition: aCondition) + ifTrue: [ + self clearHaltOnce; halt].! Item was added: + ----- Method: Object>>meetsHaltCondition: (in category 'debugging') ----- + meetsHaltCondition: aCondition + "If the Block has one arg, the receiver is bound to that. + If the condition is a selector, we look up in the callchain. Halt if any method's selector equals selector." + + aCondition isSymbol ifTrue: [ | ctxt | + "only halt if a method with selector symbol is in callchain" + ctxt := thisContext. + [(ctxt := ctxt sender) isNil] whileFalse: [ + (ctxt selector = aCondition) ifTrue: [^ true] ]. + ^ false ]. + ^ aCondition isBlock + ifTrue: [aCondition cull: self] + ifFalse: [aCondition value]!

2 1

[Sounds] How to turn off reverb
by Thiede, Christoph 20 Jul '22

20 Jul '22

Hi all, I never have dealt with the Sound system in Squeak before so here is a newbie question from my side: How can I turn off the reverb which you hear every time you play an FM sound? Take this example: AbstractSound dial: '#'. Either my speakers are broken or there is a reverb that is hidden somewhere in the implementation of FMSound. :-) I can also reproduce this when creating an FMSound manually without defining any envelopes. I also took a look at the ingenious Speech-TTS package from Squeak 3 and could hear a reverb there, too. Is this a bug? Is this a - maybe unfortunate - default setting? How can I turn off this? Thanks in advance! :-) Best, Christoph

8 13

Improving visibility of comments in Shout renders code within SqueakTheme
by tim Rowledge 11 Jul '22

11 Jul '22

I find it very hard to discern the colour used by Shout for comments in the SqueakTheme. I would like to suggest that we change to make it italic as well as coloured. If we edit SqueakTheme class>>#addSyntaxHighlighting: where we set the comment style - set: #comment for: #SHTextStylerST80 to: {Color cyan muchDarker. TextEmphasis italic}; and then run `self create apply` in the browser then things are more visible. tim -- tim Rowledge; tim(a)rowledge.org; http://www.rowledge.org/tim Trancelators interpret messages from the dead

10 18

Association >> #hash performance
by christoph.thiede＠student.hpi.uni-potsdam.de 09 Jul '22

09 Jul '22

Hi all! while playing around with code simulation recently, I found out that the following dictionary lookup consumes extremely many bytecodes: UserInterfaceTheme current get: #return for: SHTextStylerST80. Dictionary >> #scanFor: takes 59 (sic!) attempts to find the right item in the dictionary. In other words, almost 20% of the property table is traversed, which does not conform my idea of a hashtable. The reason for this can be found in Association >> #hash: Since Collections-ar.304, it does not take the value of an association into consideration for the hash any longer. As UserInterfaceTheme uses Associations as keys in the Dictionary, we have a lot of "hash misses". Even though you might not notice this in practice, I noticed this indeed during debugging und simulating, I can tell you. :-) On the other hand, reverting Association >> #hash to honor the value slows down Unicode class compileAll indeed (significantly: 60 ms vs. 2100 ms), as noted in Andreas' original comment: Name: Collections-ar.304 Author: ar Time: 11 February 2010, 11:52:58.959 pm UUID: 9e97b360-adf1-f745-8f8c-562aa08d566e Ancestors: Collections-ar.303 Fix an age-old bug in Association>>hash which causes extreme slowdowns in compiling Unicode methods. Association>>hash does not need to hash the value; it's slow and useless. So my question is: What would be the right fix for this situation? Is it a) Do not use Associations in Dictionarys when its value matters for the lookup; or b) Do only use Associations in Dictionarys when its value maters for the lookup? For a), the dictionary layout of UserInterfaceTheme needs to be changed, for which I am attaching a changeset that replaces Associations by Arrays there.* For b), the layout of litIndSet in Encoder needs to be changed, for instance by using Bindings instead of Associations for classPool entries. However, I'm not familiar enough with Encoder etc. to answer this question. *The changeset will automatically update all UserInterfaceTheme instances. Benchmark: theme := UserInterfaceTheme current. random := Random seed: 20220526. properties := (1 to: 10000) collect: [:i | theme properties keys atRandom: random]. [properties collect: [:ea | theme get: ea]] bench. Old: 45 ms | New: 10 ms Looking forward to your opinion: How should we treat Associations in Dictionarys? Of course, this is not release-critical. :-) Best, Christoph =============== Postscript (accelerateUserInterfaceTheme.2.cs) =============== "Postscript: Update UserInterfaceTheme instances" UserInterfaceTheme allSubInstancesDo: [:theme | | newProperties | newProperties := theme properties copyEmpty. theme properties keysAndValuesDo: [:key :value | newProperties at: ((key isKindOf: Association) ifTrue: [{key key. key value}] ifFalse: [key]) put: value]. theme instVarNamed: 'properties' put: newProperties]. =============== Diff =============== UserInterfaceTheme>>get: {private} · ct 5/26/2022 21:15 (changed) get: keyObject "keyObject is intended to be an Association. We have two lookup strategies: 1) along the superclass chain *of the client*, 2) via a linked theme. Evaluate the result because there can be message sends stored or blocks." | k | properties at: keyObject ifPresent: [:prop | ^ prop value]. - keyObject isVariableBinding "simple key objects" + keyObject isArray "simple key objects" ifFalse: [^ self getViaLink: keyObject]. - k := keyObject key. + k := keyObject at: 1. (self getViaSuperclasses: keyObject) ifNotNil: [:prop | ^ prop]. - keyObject key: k. "restore" + keyObject at: 1 put: k. "restore" ^ self getViaLink: keyObject UserInterfaceTheme>>get:for: {private} · ct 5/26/2022 21:02 (changed) get: propertySymbol for: scope "For convenience. Does support access to non-class keys." | aClass | aClass := (scope isNil or: [scope isBehavior]) ifTrue: [scope] ifFalse: [Smalltalk classNamed: scope]. - aClass ifNotNil: [^ self get: aClass -> propertySymbol]. + aClass ifNotNil: [^ self get: {aClass. propertySymbol}]. properties - at: scope -> propertySymbol + at: {scope. propertySymbol} ifPresent: [:prop | ^ prop value]. - ^ self getViaLink: scope -> propertySymbol + ^ self getViaLink: {scope. propertySymbol} UserInterfaceTheme>>getViaSuperclasses: {private} · ct 5/26/2022 21:15 (changed) getViaSuperclasses: keyObject "keyObject is intended to be an Association. Find the superclass of the key of the keyObject (which will initially be the client's class) and make a new keyObject using that and the original message name, then try searching for that." "We know we're the only referencer of keyObject. Update it rather than create new ones, for performance reasons." - keyObject key: keyObject key superclass. + keyObject at: 1 put: (keyObject at: 1) superclass. - keyObject key ifNil: [^ nil]. + (keyObject at: 1) ifNil: [^ nil]. properties at: keyObject ifPresent: [:prop | ^ prop value]. ^ self getViaSuperclasses: keyObject UserInterfaceTheme>>set:for:to: {building} · ct 5/26/2022 21:03 (changed) set: propertySymbol for: aClassOrSymbol to: valueObject "Where aClass asks its userInterfaceTheme for propertySymbol, provide valueObject." | aClass | aClass := aClassOrSymbol isBehavior ifTrue: [aClassOrSymbol] ifFalse: [Smalltalk classNamed: aClassOrSymbol]. aClass ifNil: [^ self]. ^ self atomicUpdate: [ : props | | key | - key := aClass -> propertySymbol. + key := {aClass. propertySymbol}. valueObject ifNil: [ props removeKey: key ifAbsent: [ "already cleared, don't error" ] ] ifNotNil: [ props at: key put: valueObject ] ] UserInterfaceThemeRequest>>doesNotUnderstand: {lookup} · ct 5/26/2022 21:02 (changed) doesNotUnderstand: aMessage "Look up the visual attribute specified by aMessage's #selector in the current theme for the current target object." aMessage numArgs = 0 ifTrue: [ - ^ (self theme get: self target class -> aMessage selector) + ^ (self theme get: {self target class. aMessage selector}) ifNil: [(self theme respondsTo: aMessage selector) ifTrue: [self theme perform: aMessage selector] ifFalse: [nil "unset property"]]]. aMessage numArgs = 1 ifTrue: [ ^ self theme - set: self target class -> aMessage selector asSimpleGetter + set: {self target class. aMessage selector asSimpleGetter} to: aMessage arguments first]. ^ self theme perform: aMessage selector withArguments: aMessage arguments. --- Sent from Squeak Inbox Talk ["accelerateUserInterfaceTheme.2.cs"]

3 3

← Newer
1
2
3
4
5
6
7
...
32
Older →

Jump to page: