Squeak-dev September 2020

squeak-dev@lists.squeakfoundation.org

37 participants
269 discussions

The Inbox: SUnit-ct.125.mcz
by commits＠source.squeak.org 29 Dec '21

29 Dec '21

Christoph Thiede uploaded a new version of SUnit to project The Inbox: http://source.squeak.org/inbox/SUnit-ct.125.mcz ==================== Summary ==================== Name: SUnit-ct.125 Author: ct Time: 29 March 2020, 1:02:50.60056 pm UUID: 814e778e-04dc-954c-93e3-fb19e307d4c8 Ancestors: SUnit-mt.121 Proposal: Update debugging logic of TestCase. Instead of abusing #halt, open a debugger directly on the test selector. Deprecates #openDebuggerOnFailingTestMethod. Also slightly refactors internal behavior to avoid some duplication, and to get rid of unnecessary semaphores which should not add any value compared to #ensure: unless some really low-level things are broken. Please correct me if I am wrong here! =============== Diff against SUnit-mt.121 =============== Item was added: + ----- Method: TestCase>>assureResourcesDuring: (in category 'private') ----- + assureResourcesDuring: aBlock + + | resources | + resources := self resources. + resources do: [:resource | + resource isAvailable ifFalse: [ + ^ resource signalInitializationError]]. + ^ aBlock ensure: [ + resources do: [:resource | + resource reset]].! Item was changed: ----- Method: TestCase>>debug (in category 'running') ----- debug + "Run the receiver and open a debugger on the first failure or error." + + ^ self assureResourcesDuring: [self runCaseWithoutTimeout]! - self resources do: - [ : res | res isAvailable ifFalse: [ ^ res signalInitializationError ] ]. - [ self runCase ] ensure: - [ self resources do: - [ : each | each reset ] ]! Item was changed: ----- Method: TestCase>>debugAsFailure (in category 'running') ----- debugAsFailure + "Spawn a debugger that is ready to debug the receiver." + + (Process + forBlock: [self debug] + runUntil: [:context | context selector = testSelector]) + debug.! - | semaphore | - semaphore := Semaphore new. - self resources do: [:res | - res isAvailable ifFalse: [^res signalInitializationError]]. - [semaphore wait. self resources do: [:each | each reset]] fork. - (self class selector: testSelector) runCaseAsFailure: semaphore.! Item was changed: ----- Method: TestCase>>openDebuggerOnFailingTestMethod (in category 'running') ----- openDebuggerOnFailingTestMethod + + self deprecated: 'ct: Use #debugAsFailure'. + "SUnit has halted one step in front of the failing test method. Step over the 'self halt' and send into 'self perform: testSelector' to see the failure from the beginning" - self halt; performTest! Item was changed: ----- Method: TestCase>>runCaseAsFailure: (in category 'running') ----- runCaseAsFailure: aSemaphore + + self deprecated: 'ct: Use #runCaseWithoutTimeout and #ensure:'. + ^ [self runCaseWithoutTimeout] + ensure: [aSemaphore signal]! - [self setUp. - self openDebuggerOnFailingTestMethod] ensure: [ - self tearDown. - aSemaphore signal]! Item was added: + ----- Method: TestCase>>runCaseWithoutTimeout (in category 'running') ----- + runCaseWithoutTimeout + + [self setUp. + self performTest] + ensure: [self tearDown].!

4 4

Changeset: Deprecate SyntaxErrorNotification >> #errorMessage
by Thiede, Christoph 08 Dec '21

08 Dec '21

Hi all, please have a look at the attached changeset. It deprecates the IMO redundant instance variable 'errorMessage' in SyntaxErrorNotification and uses super's 'messageText' instead. The two single senders in the Trunk image are patched as well. I wonder at all how the usage in DecompilerTests could have worked before, giving the prior implementation of #messageText in SyntaxErrorNotification ... Please review and merge if you like it. Best, Christoph

2 1

Changeset: forceString.1.cs
by Thiede, Christoph 08 Dec '21

08 Dec '21

This changeset proposes an additional property for PluggableTextMorph/PluggableTextSpec, named #forceString. If it is activated, any formattings will be erased from the text morph instantly. This can be useful for widgets that are not actually designed to accept texts and thus display a heap of unwanted formattings. An example is the SearchBar, which at least in my image too often looks totally misstyled after several copy'n'paste operations: [cid:4dc50742-4fb0-49c7-8e91-e28c30faaff0] (Or should maybe use a text styler there, instead?) Another, probably better example is FillInTheBlankMorph. As it only returns a string, there is no need to display styled texts. How would you think about it? Please let me know. :-) Best, Christoph

3 5

The Inbox: System-ct.1136.mcz
by commits＠source.squeak.org 07 Dec '21

07 Dec '21

Christoph Thiede uploaded a new version of System to project The Inbox: http://source.squeak.org/inbox/System-ct.1136.mcz ==================== Summary ==================== Name: System-ct.1136 Author: ct Time: 6 February 2020, 7:50:02.954648 pm UUID: 0a637557-d32d-9347-bc4b-4f2bb95a707d Ancestors: System-cmm.1131 Make Smalltalk more robust against interruptions/timeouts during sources compilation A popular example is a test case that compiles a lot of methods and then is terminated via timeout. See [1] for an example. This commit ensures that the changes file does not remain closed when the method is curtailed. Before probably destroyed your image, now works: [MCSnapshotResource mockPackage unload] valueWithin: 1 seconds "you may need to adapt the limit to reproduce" onTimeout: []. [1] http://forum.world.st/BUG-MultiByteFileStream-Object-gt-gt-error-primGetPos… =============== Diff against System-cmm.1131 =============== Item was changed: ----- Method: SmalltalkImage>>forceChangesToDisk (in category 'sources, changes log') ----- forceChangesToDisk "Ensure that the changes file has been fully written to disk by closing and re-opening it. This makes the system more robust in the face of a power failure or hard-reboot." | changesFile | changesFile := SourceFiles at: 2. (changesFile isKindOf: FileStream) ifTrue: [ changesFile flush. + SecurityManager default hasFileAccess ifTrue: [ + [changesFile close] ensure: [ + changesFile open: changesFile name forWrite: true]]. + changesFile setToEnd].! - SecurityManager default hasFileAccess ifTrue:[ - changesFile close. - changesFile open: changesFile name forWrite: true]. - changesFile setToEnd. - ]. - !

6 7

The Inbox: Tools-ct.962.mcz
by commits＠source.squeak.org 07 Dec '21

07 Dec '21

Christoph Thiede uploaded a new version of Tools to project The Inbox: http://source.squeak.org/inbox/Tools-ct.962.mcz ==================== Summary ==================== Name: Tools-ct.962 Author: ct Time: 20 March 2020, 6:58:18.856237 pm UUID: c4cbd7b0-ddf5-3b4b-93ed-55e5b2c80465 Ancestors: Tools-mt.955 Makes variable bindings in Workspace more comfortable: - Ask for a useful variable name when dropping something into a workspace - Validate variable name before creating inaccessible bindings - Do not propose invalid names at all (for example, dropped class were given an invalid name in the past) - Ask before removing variable bindings. This also increases responsiveness of the UI. - Improves multilingual support =============== Diff against Tools-mt.955 =============== Item was changed: ----- Method: Workspace>>acceptDroppingMorph:event:inMorph: (in category 'drag and drop') ----- acceptDroppingMorph: dropee event: evt inMorph: targetMorph + "Return the dropee to its old position, and add a reference to it at the cursor point." + - "Return the dropee to its old position, and add a reference to it at the - cursor point." | bindingName externalName reference | (dropee isKindOf: TransferMorph) ifTrue: [reference := dropee passenger. externalName := dropee passenger className] ifFalse: [reference := dropee. externalName := dropee externalName]. externalName := externalName isOctetString + ifTrue: [externalName] + ifFalse: ['a' , externalName]. + + bindingName := Project uiManager + request: 'Please enter a name for the dropped object:' translated + initialAnswer: (externalName withFirstCharacterDownshifted , reference identityHash printString) asLegalSelector. + bindingName isEmptyOrNil ifTrue: [^ false]. + bindingName := Parser new parseSelector: bindingName. + bindingName ifNil: [ + self inform: 'Invalid selector.' translated. + ^ false]. + - ifTrue: [externalName] - ifFalse: ['a' , externalName]. - bindingName := externalName withFirstCharacterDownshifted , reference identityHash printString. - targetMorph correctSelectionWithString: bindingName , ' '. (self bindingOf: bindingName) value: reference. + targetMorph correctSelectionWithString: bindingName , ' '. (dropee isKindOf: TransferMorph) ifFalse: [dropee rejectDropMorphEvent: evt]. + ^ true "success"! - ^ true"success"! Item was changed: ----- Method: Workspace>>addModelItemsToWindowMenu: (in category 'menu commands') ----- addModelItemsToWindowMenu: aMenu aMenu addLine. aMenu + add: 'save contents to file...' translated - add: 'save contents to file...' target: self action: #saveContentsInFile. aMenu + add: 'inspect variables' translated - add: 'inspect variables' target: self action: #inspectBindings. aMenu + add: 'reset variables' translated - add: 'reset variables' target: self + action: #resetBindings. - action: #initializeBindings. aMenu addUpdating: #mustDeclareVariableWording target: self action: #toggleVariableDeclarationMode. aMenu addUpdating: #acceptDroppedMorphsWording target: self action: #toggleDroppingMorphForReference. + self addToggleStylingMenuItemTo: aMenu.! - self addToggleStylingMenuItemTo: aMenu. - ! Item was added: + ----- Method: Workspace>>resetBindings (in category 'menu commands') ----- + resetBindings + + (self bindings notEmpty ==> [self confirm: ('Are you sure you would like to remove {1} variables?' translated format: {self bindings size})]) + ifFalse: [^ self]. + self initializeBindings.!

5 7

The Inbox: System-ct.1156.mcz
by commits＠source.squeak.org 07 Dec '21

07 Dec '21

Christoph Thiede uploaded a new version of System to project The Inbox: http://source.squeak.org/inbox/System-ct.1156.mcz ==================== Summary ==================== Name: System-ct.1156 Author: ct Time: 5 May 2020, 5:41:10.96399 pm UUID: 2ce2496e-92e9-f744-9a58-d9e9320adfab Ancestors: System-mt.1155 Fixes double prompt for handling a dropped file when the first dialog was cancelled. Cleans up duplication between ExternalDropHandler >> #lookupServiceBasedHandler: and DefaultExternalDropHandler by deprecating the first one. Maybe we can also remove this method? =============== Diff against System-mt.1155 =============== Item was changed: ----- Method: DefaultExternalDropHandler>>chooseServiceFrom: (in category 'private') ----- chooseServiceFrom: aCollection + "private - choose a service from aCollection asking the user if needed" + ^ aCollection size = 1 - "private - choose a service from aCollection asking the user if needed" - ^aCollection size = 1 ifTrue: [aCollection anyOne] + ifFalse: [UIManager default + chooseFrom: (aCollection collect: [:each | each label]) - ifFalse:[UIManager default - chooseFrom: (aCollection collect:[:each| each label]) values: aCollection title: 'You dropped a file. Please choose a service:' translated withCRs].! Item was changed: ----- Method: DefaultExternalDropHandler>>handle:in:dropEvent: (in category 'event handling') ----- handle: dropStream in: pasteUp dropEvent: anEvent "the file was just droped, let's do our job" | fileName services chosen | fileName := dropStream name. services := self servicesForFileNamed: fileName. + - "no service, default behavior" + services ifEmpty: [ + ^ UIManager default edit: dropStream contentsOfEntireFile]. + - services isEmpty - ifTrue: [^UIManager default edit: dropStream contentsOfEntireFile]. - chosen := self chooseServiceFrom: services. + chosen ifNotNil: [chosen performServiceFor: dropStream].! - chosen ifNotNil:[chosen performServiceFor: dropStream]! Item was changed: ----- Method: DefaultExternalDropHandler>>servicesForFileNamed: (in category 'private') ----- servicesForFileNamed: aString + "private - answer a collection of file-services for the file named aString" + + ^ (FileServices itemsForFile: aString) + copyWithoutAll: self unwantedSelectors! - "private - answer a collection of file-services for the file named - aString" - | allServices | - allServices := FileServices itemsForFile: aString. - ^ allServices - reject: [:svc | self unwantedSelectors includes: svc selector]! Item was changed: ----- Method: ExternalDropHandler class>>defaultHandler (in category 'accessing') ----- defaultHandler + + ^ DefaultHandler ifNil: [ + DefaultHandler := ExternalDropHandler + type: nil + extension: nil + action: #edit]! - DefaultHandler ifNil: [DefaultHandler := ExternalDropHandler type: nil extension: nil action: [:dropStream | dropStream edit]]. - ^DefaultHandler! Item was changed: ----- Method: ExternalDropHandler class>>lookupExternalDropHandler: (in category 'accessing') ----- lookupExternalDropHandler: stream | types extension serviceHandler | types := stream mimeTypes. types ifNotNil: [ + self registeredHandlers + detect: [:handler | handler matchesTypes: types] + ifFound: [:handler | ^ handler]]. - self registeredHandlers do: [:handler | - (handler matchesTypes: types) - ifTrue: [^handler]]]. extension := FileDirectory extensionFor: stream name. + self registeredHandlers + detect: [:handler | handler matchesExtension: extension] + ifFound: [:handler | ^ handler]. + + ^ serviceHandler ifNil: [self defaultHandler]! - self registeredHandlers do: [:handler | - (handler matchesExtension: extension) - ifTrue: [^handler]]. - serviceHandler := self lookupServiceBasedHandler: stream. - ^serviceHandler - ifNil: [self defaultHandler]! Item was changed: ----- Method: ExternalDropHandler class>>lookupServiceBasedHandler: (in category 'accessing') ----- lookupServiceBasedHandler: dropStream + - "The file was just dropped; let's do our job" | fileName services theOne | + self deprecated: 'ct: Use DefaultExternalDropHandler'. fileName := dropStream name. - services := (FileServices itemsForFile: fileName) reject: [:svc | self unwantedSelectors includes: svc selector]. + - "no service, default behavior" + services ifEmpty: [^ nil]. + - services isEmpty - ifTrue: [^nil]. - theOne := self chooseServiceFrom: services. + ^ theOne ifNotNil: [ + ExternalDropHandler type: nil extension: nil action: [:stream | theOne performServiceFor: stream]]! - ^theOne - ifNotNil: [ExternalDropHandler type: nil extension: nil action: [:stream | theOne performServiceFor: stream]]!

3 3

The Inbox: ShoutCore-ct.69.mcz
by commits＠source.squeak.org 06 Dec '21

06 Dec '21

A new version of ShoutCore was added to project The Inbox: http://source.squeak.org/inbox/ShoutCore-ct.69.mcz ==================== Summary ==================== Name: ShoutCore-ct.69 Author: ct Time: 14 August 2019, 10:43:04.7164 pm UUID: 7d165f4e-09bf-a445-a664-5e8edb9867b8 Ancestors: ShoutCore-ul.68 Fix bug in SHParserST80: Wrong styling after blocks with arguments without statements Thanks Jakob for the report! =============== Diff against ShoutCore-ul.68 =============== Item was changed: ----- Method: SHParserST80>>parseBlockArguments (in category 'parse') ----- parseBlockArguments currentTokenFirst == $: ifFalse: [ ^self ]. [ currentTokenFirst == $: ] whileTrue: [ self scanPast: #blockArgColon; parseArgument: #blockPatternArg ]. + ((self parseVerticalBarForTemporaries: #blockArgsBar) and: [ + currentTokenFirst ~= $] ]) + ifFalse: [ + self fail ": Missing block args bar" ]! - (self parseVerticalBarForTemporaries: #blockArgsBar) ifFalse: [ - self fail ": Missing block args bar" ]!

12 25

The Inbox: Tools-ct.937.mcz
by commits＠source.squeak.org 06 Dec '21

06 Dec '21

Christoph Thiede uploaded a new version of Tools to project The Inbox: http://source.squeak.org/inbox/Tools-ct.937.mcz ==================== Summary ==================== Name: Tools-ct.937 Author: ct Time: 29 January 2020, 6:26:57.711894 pm UUID: 34af35bc-83e2-9b40-9aa3-8f6af607ac6e Ancestors: Tools-mt.929 Fixes a styling issue in the Workspace that missstyled class variables from Object as invalid (that is DependentsFields). This is a bug because you can access DependentsFields indeed. There is a subtle but important difference between the classOrMetaclass nil and the classOrMetaclass Object: The former does not respect any state for styling, but the latter respects class variables from Object (as well as, hypothetically, ProtoObject). Please note that it would be even more accurate to specify nil class instead of Object here, but I don't know whether we really want to emphasize the nuances of UndefinedObject. It feels rather like an implementation detail for me, at least in this context. (Just imagine UndefinedObject had a UniqueInstance class variable - would we really to access it from any Workspace?) Open for opinions. =============== Diff against Tools-mt.929 =============== Item was changed: ----- Method: Workspace>>aboutToStyle: (in category 'code pane') ----- aboutToStyle: aStyler self shouldStyle ifFalse: [ ^false ]. + aStyler + parseAMethod: false; + classOrMetaClass: Object; - aStyler - classOrMetaClass: nil; workspace: self. ^true!

4 3

Merge Request: #caseOf:otherwise: with arguments
by Thiede, Christoph 28 Nov '21

28 Nov '21

This changeset makes it possible to accept an argument in the otherwise block of a #caseOf:otherwise: call. Diff of the #caseOf:otherwise: implementation: caseOf: aBlockAssociationCollection otherwise: aBlock "The elements of aBlockAssociationCollection are associations between blocks. Answer the evaluated value of the first association in aBlockAssociationCollection whose evaluated key equals the receiver. If no match is found, answer the result of evaluating aBlock." aBlockAssociationCollection associationsDo: [:assoc | (assoc key value = self) ifTrue: [^assoc value value]]. - ^ aBlock value + ^ aBlock cull: self "| z | z := {[#a]->[1+1]. ['b' asSymbol]->[2+2]. [#c]->[3+3]}. #b caseOf: z otherwise: [0]" "| z | z := {[#a]->[1+1]. ['d' asSymbol]->[2+2]. [#c]->[3+3]}. #b caseOf: z otherwise: [0]" "The following are compiled in-line:" "#b caseOf: {[#a]->[1+1]. ['b' asSymbol]->[2+2]. [#c]->[3+3]} otherwise: [0]" "#b caseOf: {[#a]->[1+1]. ['d' asSymbol]->[2+2]. [#c]->[3+3]} otherwise: [0]" + "#b caseOf: {[#a]->[1+1]. ['d' asSymbol]->[2+2]. [#c]->[3+3]} otherwise: [:x | x halt]" Furthermore, the changeset includes a necessary modification of MessageNode >> #transformCase: so that the otherwise argument can be compiled in-line. Last but not least, I wrote some tests for #caseOf:[otherwise:]. Please review! (In a later change, it would be possible to allow arguments for the association key blocks as well. But I love short feedback loops, so let's assess this one first :-)) Best, Christoph

4 9

BUG/REGRESSION while debugging Generator >> #nextPut:
by Thiede, Christoph 15 Nov '21

15 Nov '21

Hi all, I discovered a bug when stepping over a call to Generator >> #nextPut:. !! Save your image before trying the following !! Generator on: [:stream | stream nextPut: #foo] [cid:95952b5e-98f1-4447-bd09-29d396961281] 1. Due to multiprocessing implementation of the Generator, pressing over will hang up the current process forever, because the generator is never queried. 2. If I press cmd-dot, I get an infinite number of debugger windows that show Context >> #cannotReturn:. This will probably damage your image. The first is annoying for newcomers, but I see this behavior is reasonable and without the effect of the second, it wouldn't be a big problem. The second, however, looks like a bigger issue to me. Actually, I think the most serious aspect is that in Squeak 5.1, the same procedure does not crash your image, but you are forwarded to the emergency debugger and can terminate the process: [cid:c8eee8c3-d2de-49c0-b5d4-33bd340d6c98] [cid:6044ecf2-91ab-4f6e-b863-b980586868e9] Also, this is not the first scenario in the last time where I got an infinite debugger chain. See [BUG(s)] in Context control (#jump, #runUntilErrorOrReturnFrom:)<http://forum.world.st/BUG-s-in-Context-control-jump-runUntilErrorOrReturnFr…> for a similar issue. And there were even more situations which I could not yet reproduce exactly. I'm afraid that the recent changes to the debuggers might have weakened its ability to detect recursive errors. Can someone else tell about these problems? I suppose we ignore a large number of recursive errors as in MorphicDebugger >> #openOn:context:label:contents:fullView:, the uiBlock is often triggered as a separate UI message which will be executed after the recursion flag has been cleared by the base class. Would be great if someone could have a look at it or share more information :) Best, Christoph

8 25

← Newer
1
2
3
4
5
6
7
...
27
Older →

Jump to page: