help needed
Locked
 
|
I am trying to look into this issue in Glamour related to morphs: The problem is that, in Glamour, one can choose to display first few items instead of the all and then the user can choose to display more. When the selection is set to the items not shown, there is a DNU in PaginatedMorphTreeMorph. Digging this problem, I have encountered this piece of code for which I would require the help of someone knowing Morph. Is someone knowledgable about this piece of code? Here you can find an image with the script to explore the problem: https://dl.dropbox.com/u/11804892/glamour-treemorphs.zip
MorphTreeMorph >> update: update: aSymbol aSymbol == #selection ifTrue: [^ self listManager updateSelectionFromModel].
((aSymbol isKindOf: Array) and: [aSymbol size > 1 and: [aSymbol first == #selectItems and: [aSymbol second isCollection]]])
ifTrue: [^ self selectedItems: aSymbol second]. aSymbol == self nodeListSelector ifTrue: [^ self updateList].
aSymbol == #selectAll ifTrue: [^ self listManager selectAll]. aSymbol == #deselectAll ifTrue: [^ self listManager deselectAll].
((aSymbol isKindOf: Array) and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [
aSymbol second == #openPath]]]) "allow directed path opening where multiple trees exist" ifTrue:
[^(self allNodeMorphs at: 1 ifAbsent: [^self]) openPath: (aSymbol allButFirst: 2)]. ((aSymbol isKindOf: Array)
and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ aSymbol second == #openItemPath]]]) "allow directed path opening where multiple trees exist"
ifTrue: [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) openItemPath: (aSymbol allButFirst: 2)].
((aSymbol isKindOf: Array) and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [
aSymbol second == #expandAllFromNode:]]]) ifTrue: [^ self expandAllFromNode: aSymbol third].
((aSymbol isKindOf: Array) and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [
aSymbol second == #closeItemPath]]]) "allow directed path closing where multiple trees exist" ifTrue:
[^ (self allNodeMorphs at: 1 ifAbsent: [^self]) closeItemPath: (aSymbol allButFirst: 2)]. ((aSymbol isKindOf: Array)
and: [aSymbol notEmpty and: [aSymbol first == #openPath]]) ifTrue: [^(self allNodeMorphs at: 1 ifAbsent: [^self])
openPath: aSymbol allButFirst]. ((aSymbol isKindOf: Array) and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector and: [
aSymbol second == #expandRoots]]]) ifTrue: [^self expandRoots].
((aSymbol isKindOf: Array) and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector]]) ifTrue: [aSymbol second = #expandAll ifTrue: [^ self expandAll].
aSymbol second = #collapseAll ifTrue: [^ self collapseAll]] |
|
|
Arghhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
What a terrible method……. so ugly. There is something fundamentally wrong. This is why we should rewrite MorphTreeMorph. Stef On Jan 30, 2013, at 4:13 PM, Usman Bhatti wrote: > I am trying to look into this issue in Glamour related to morphs: > > http://code.google.com/p/pharo/issues/detail?id=7345 > > The problem is that, in Glamour, one can choose to display first few items instead of the all and then the user can choose to display more. When the selection is set to the items not shown, there is a DNU in PaginatedMorphTreeMorph. Digging this problem, I have encountered this piece of code for which I would require the help of someone knowing Morph. Is someone knowledgable about this piece of code? Here you can find an image with the script to explore the problem: https://dl.dropbox.com/u/11804892/glamour-treemorphs.zip > > > MorphTreeMorph >> update: > > > update: aSymbol > aSymbol == #selection ifTrue: [^ self listManager updateSelectionFromModel]. > ((aSymbol isKindOf: Array) and: [aSymbol size > 1 and: [aSymbol first == #selectItems and: [aSymbol second isCollection]]]) > ifTrue: [^ self selectedItems: aSymbol second]. > aSymbol == self nodeListSelector ifTrue: [^ self updateList]. > aSymbol == #selectAll ifTrue: [^ self listManager selectAll]. > aSymbol == #deselectAll ifTrue: [^ self listManager deselectAll]. > ((aSymbol isKindOf: Array) > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > aSymbol second == #openPath]]]) "allow directed path opening where multiple trees exist" > ifTrue: > [^(self allNodeMorphs at: 1 ifAbsent: [^self]) > openPath: (aSymbol allButFirst: 2)]. > ((aSymbol isKindOf: Array) > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > aSymbol second == #openItemPath]]]) "allow directed path opening where multiple trees exist" > ifTrue: > [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) > openItemPath: (aSymbol allButFirst: 2)]. > ((aSymbol isKindOf: Array) > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > aSymbol second == #expandAllFromNode:]]]) > ifTrue: > [^ self expandAllFromNode: aSymbol third]. > ((aSymbol isKindOf: Array) > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > aSymbol second == #closeItemPath]]]) "allow directed path closing where multiple trees exist" > ifTrue: > [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) > closeItemPath: (aSymbol allButFirst: 2)]. > ((aSymbol isKindOf: Array) > and: [aSymbol notEmpty and: [aSymbol first == #openPath]]) > ifTrue: > [^(self allNodeMorphs at: 1 ifAbsent: [^self]) > openPath: aSymbol allButFirst]. > ((aSymbol isKindOf: Array) > and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector and: [ > aSymbol second == #expandRoots]]]) > ifTrue: > [^self expandRoots]. > ((aSymbol isKindOf: Array) > and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector]]) > ifTrue: [aSymbol second = #expandAll ifTrue: [^ self expandAll]. > aSymbol second = #collapseAll ifTrue: [^ self collapseAll]] |
|
|
On 31 jan 2013, at 01:04, Stéphane Ducasse <[hidden email]> wrote:
> Arghhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh > What a terrible method……. so ugly. > There is something fundamentally wrong. +1000 -- Cheers, Peter > This is why we should rewrite MorphTreeMorph. > > Stef > > > > On Jan 30, 2013, at 4:13 PM, Usman Bhatti wrote: > >> I am trying to look into this issue in Glamour related to morphs: >> >> http://code.google.com/p/pharo/issues/detail?id=7345 >> >> The problem is that, in Glamour, one can choose to display first few items instead of the all and then the user can choose to display more. When the selection is set to the items not shown, there is a DNU in PaginatedMorphTreeMorph. Digging this problem, I have encountered this piece of code for which I would require the help of someone knowing Morph. Is someone knowledgable about this piece of code? Here you can find an image with the script to explore the problem: https://dl.dropbox.com/u/11804892/glamour-treemorphs.zip >> >> >> MorphTreeMorph >> update: >> >> >> update: aSymbol >> aSymbol == #selection ifTrue: [^ self listManager updateSelectionFromModel]. >> ((aSymbol isKindOf: Array) and: [aSymbol size > 1 and: [aSymbol first == #selectItems and: [aSymbol second isCollection]]]) >> ifTrue: [^ self selectedItems: aSymbol second]. >> aSymbol == self nodeListSelector ifTrue: [^ self updateList]. >> aSymbol == #selectAll ifTrue: [^ self listManager selectAll]. >> aSymbol == #deselectAll ifTrue: [^ self listManager deselectAll]. >> ((aSymbol isKindOf: Array) >> and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ >> aSymbol second == #openPath]]]) "allow directed path opening where multiple trees exist" >> ifTrue: >> [^(self allNodeMorphs at: 1 ifAbsent: [^self]) >> openPath: (aSymbol allButFirst: 2)]. >> ((aSymbol isKindOf: Array) >> and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ >> aSymbol second == #openItemPath]]]) "allow directed path opening where multiple trees exist" >> ifTrue: >> [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) >> openItemPath: (aSymbol allButFirst: 2)]. >> ((aSymbol isKindOf: Array) >> and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ >> aSymbol second == #expandAllFromNode:]]]) >> ifTrue: >> [^ self expandAllFromNode: aSymbol third]. >> ((aSymbol isKindOf: Array) >> and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ >> aSymbol second == #closeItemPath]]]) "allow directed path closing where multiple trees exist" >> ifTrue: >> [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) >> closeItemPath: (aSymbol allButFirst: 2)]. >> ((aSymbol isKindOf: Array) >> and: [aSymbol notEmpty and: [aSymbol first == #openPath]]) >> ifTrue: >> [^(self allNodeMorphs at: 1 ifAbsent: [^self]) >> openPath: aSymbol allButFirst]. >> ((aSymbol isKindOf: Array) >> and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector and: [ >> aSymbol second == #expandRoots]]]) >> ifTrue: >> [^self expandRoots]. >> ((aSymbol isKindOf: Array) >> and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector]]) >> ifTrue: [aSymbol second = #expandAll ifTrue: [^ self expandAll]. >> aSymbol second = #collapseAll ifTrue: [^ self collapseAll]] > > |
|
|
In reply to this post by Usman Bhatti
Morbid fascination
It is late so I may completely off but how many isKindOf: are executed in this method? > ((aSymbol isKindOf: Array) > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > aSymbol second == #openPath]]]) when [aSymbol second == #openPath] is not openPath then we go to the next (not mentioning duplicated code inside the method ---- I will vomit) and we restart to execute isKindOf: ((aSymbol isKindOf: Array) and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ aSymbol second == #expandAllFromNode:]]]) Am'I right? It looks so terrible. Stef > update: aSymbol > aSymbol == #selection ifTrue: [^ self listManager updateSelectionFromModel]. > ((aSymbol isKindOf: Array) and: [aSymbol size > 1 and: [aSymbol first == #selectItems and: [aSymbol second isCollection]]]) > ifTrue: [^ self selectedItems: aSymbol second]. > aSymbol == self nodeListSelector ifTrue: [^ self updateList]. > aSymbol == #selectAll ifTrue: [^ self listManager selectAll]. > aSymbol == #deselectAll ifTrue: [^ self listManager deselectAll]. > ((aSymbol isKindOf: Array) > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > aSymbol second == #openPath]]]) "allow directed path opening where multiple trees exist" > ifTrue: > [^(self allNodeMorphs at: 1 ifAbsent: [^self]) > openPath: (aSymbol allButFirst: 2)]. > ((aSymbol isKindOf: Array) > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > aSymbol second == #openItemPath]]]) "allow directed path opening where multiple trees exist" > ifTrue: > [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) > openItemPath: (aSymbol allButFirst: 2)]. > ((aSymbol isKindOf: Array) > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > aSymbol second == #expandAllFromNode:]]]) > ifTrue: > [^ self expandAllFromNode: aSymbol third]. > ((aSymbol isKindOf: Array) > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > aSymbol second == #closeItemPath]]]) "allow directed path closing where multiple trees exist" > ifTrue: > [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) > closeItemPath: (aSymbol allButFirst: 2)]. > ((aSymbol isKindOf: Array) > and: [aSymbol notEmpty and: [aSymbol first == #openPath]]) > ifTrue: > [^(self allNodeMorphs at: 1 ifAbsent: [^self]) > openPath: aSymbol allButFirst]. > ((aSymbol isKindOf: Array) > and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector and: [ > aSymbol second == #expandRoots]]]) > ifTrue: > [^self expandRoots]. > ((aSymbol isKindOf: Array) > and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector]]) > ifTrue: [aSymbol second = #expandAll ifTrue: [^ self expandAll]. > aSymbol second = #collapseAll ifTrue: [^ self collapseAll]] |
|
|
Le 31/01/2013 02:30, Stéphane Ducasse a écrit :
> Morbid fascination > It is late so I may completely off but > how many isKindOf: are executed in this method? > >> ((aSymbol isKindOf: Array) >> and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ >> aSymbol second == #openPath]]]) > > when [aSymbol second == #openPath] is not openPath > then we go to the next (not mentioning duplicated code inside the method ---- I will vomit) > and we restart to execute isKindOf: > > ((aSymbol isKindOf: Array) > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > aSymbol second == #expandAllFromNode:]]]) > > Am'I right? > It looks so terrible. It is. As a user of MorphTreeMorph, this stuff is hard to use (and doesn't work very well). self history pauseDuring: [ self changed: #(#rootNodes #openItemPath) , nPath reverse ]. :( It's also very well documented, of course. Thierry > Stef > > > >> update: aSymbol >> aSymbol == #selection ifTrue: [^ self listManager updateSelectionFromModel]. >> ((aSymbol isKindOf: Array) and: [aSymbol size > 1 and: [aSymbol first == #selectItems and: [aSymbol second isCollection]]]) >> ifTrue: [^ self selectedItems: aSymbol second]. >> aSymbol == self nodeListSelector ifTrue: [^ self updateList]. >> aSymbol == #selectAll ifTrue: [^ self listManager selectAll]. >> aSymbol == #deselectAll ifTrue: [^ self listManager deselectAll]. >> ((aSymbol isKindOf: Array) >> and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ >> aSymbol second == #openPath]]]) "allow directed path opening where multiple trees exist" >> ifTrue: >> [^(self allNodeMorphs at: 1 ifAbsent: [^self]) >> openPath: (aSymbol allButFirst: 2)]. >> ((aSymbol isKindOf: Array) >> and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ >> aSymbol second == #openItemPath]]]) "allow directed path opening where multiple trees exist" >> ifTrue: >> [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) >> openItemPath: (aSymbol allButFirst: 2)]. >> ((aSymbol isKindOf: Array) >> and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ >> aSymbol second == #expandAllFromNode:]]]) >> ifTrue: >> [^ self expandAllFromNode: aSymbol third]. >> ((aSymbol isKindOf: Array) >> and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ >> aSymbol second == #closeItemPath]]]) "allow directed path closing where multiple trees exist" >> ifTrue: >> [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) >> closeItemPath: (aSymbol allButFirst: 2)]. >> ((aSymbol isKindOf: Array) >> and: [aSymbol notEmpty and: [aSymbol first == #openPath]]) >> ifTrue: >> [^(self allNodeMorphs at: 1 ifAbsent: [^self]) >> openPath: aSymbol allButFirst]. >> ((aSymbol isKindOf: Array) >> and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector and: [ >> aSymbol second == #expandRoots]]]) >> ifTrue: >> [^self expandRoots]. >> ((aSymbol isKindOf: Array) >> and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector]]) >> ifTrue: [aSymbol second = #expandAll ifTrue: [^ self expandAll]. >> aSymbol second = #collapseAll ifTrue: [^ self collapseAll]] > > > -- Thierry Goubier CEA list Laboratoire des Fondations des Systèmes Temps Réel Embarqués 91191 Gif sur Yvette Cedex France Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95 |
|
|
In reply to this post by Stéphane Ducasse
On Thu, Jan 31, 2013 at 2:30 AM, Stéphane Ducasse <[hidden email]> wrote: Morbid fascination Yes. And more so for someone not frequently working with Morphs. So, can someone at least see if the problem can be fixed without a large refactoring that should surely happen, but the glamour bug needs a quick fix. Or at least point to why the missing method on the receiver is abundantly found in the code?
usman
|
|
|
Usman did you check with 1.3 to know if the method was there?
Stef On Jan 31, 2013, at 7:03 AM, Usman Bhatti wrote: > > > > On Thu, Jan 31, 2013 at 2:30 AM, Stéphane Ducasse <[hidden email]> wrote: > Morbid fascination > It is late so I may completely off but > how many isKindOf: are executed in this method? > > > ((aSymbol isKindOf: Array) > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > aSymbol second == #openPath]]]) > > when [aSymbol second == #openPath] is not openPath > then we go to the next (not mentioning duplicated code inside the method ---- I will vomit) > and we restart to execute isKindOf: > > ((aSymbol isKindOf: Array) > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > aSymbol second == #expandAllFromNode:]]]) > > Am'I right? > It looks so terrible. > > Yes. And more so for someone not frequently working with Morphs. > So, can someone at least see if the problem can be fixed without a large refactoring that should surely happen, but the glamour bug needs a quick fix. Or at least point to why the missing method on the receiver is abundantly found in the code? > > usman > > > Stef > > > > > update: aSymbol > > aSymbol == #selection ifTrue: [^ self listManager updateSelectionFromModel]. > > ((aSymbol isKindOf: Array) and: [aSymbol size > 1 and: [aSymbol first == #selectItems and: [aSymbol second isCollection]]]) > > ifTrue: [^ self selectedItems: aSymbol second]. > > aSymbol == self nodeListSelector ifTrue: [^ self updateList]. > > aSymbol == #selectAll ifTrue: [^ self listManager selectAll]. > > aSymbol == #deselectAll ifTrue: [^ self listManager deselectAll]. > > ((aSymbol isKindOf: Array) > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > aSymbol second == #openPath]]]) "allow directed path opening where multiple trees exist" > > ifTrue: > > [^(self allNodeMorphs at: 1 ifAbsent: [^self]) > > openPath: (aSymbol allButFirst: 2)]. > > ((aSymbol isKindOf: Array) > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > aSymbol second == #openItemPath]]]) "allow directed path opening where multiple trees exist" > > ifTrue: > > [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) > > openItemPath: (aSymbol allButFirst: 2)]. > > ((aSymbol isKindOf: Array) > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > aSymbol second == #expandAllFromNode:]]]) > > ifTrue: > > [^ self expandAllFromNode: aSymbol third]. > > ((aSymbol isKindOf: Array) > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > aSymbol second == #closeItemPath]]]) "allow directed path closing where multiple trees exist" > > ifTrue: > > [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) > > closeItemPath: (aSymbol allButFirst: 2)]. > > ((aSymbol isKindOf: Array) > > and: [aSymbol notEmpty and: [aSymbol first == #openPath]]) > > ifTrue: > > [^(self allNodeMorphs at: 1 ifAbsent: [^self]) > > openPath: aSymbol allButFirst]. > > ((aSymbol isKindOf: Array) > > and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector and: [ > > aSymbol second == #expandRoots]]]) > > ifTrue: > > [^self expandRoots]. > > ((aSymbol isKindOf: Array) > > and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector]]) > > ifTrue: [aSymbol second = #expandAll ifTrue: [^ self expandAll]. > > aSymbol second = #collapseAll ifTrue: [^ self collapseAll]] > > > |
|
|
On Thu, Jan 31, 2013 at 1:20 PM, Stéphane Ducasse <[hidden email]> wrote: Usman did you check with 1.3 to know if the method was there? I looked in 1.3. The method is defined on MorphTreeMorph (see sceenshot) but absent from this class in 1.4. Copying and pasting the method from 1.3 to 1.4 in the class does not work, there are some missing instance variables in the code.
The number of implementing classes for setSelectedMorph: does not change between 1.3 and 1.4. To me it seems that the method was moved to MorphTreeListManager in 1.4. It tried to change the receiver to invoke the method on MorphTreeListManager but there are some other problems then.
|
|
|
usman
what is the expected behavior? Because I replace the last line with ^ container emptySelection but I do not know if the result is what is expected. Stef On Jan 31, 2013, at 12:48 PM, Usman Bhatti wrote: > > > > On Thu, Jan 31, 2013 at 1:20 PM, Stéphane Ducasse <[hidden email]> wrote: > Usman did you check with 1.3 to know if the method was there? > > I looked in 1.3. The method is defined on MorphTreeMorph (see sceenshot) but absent from this class in 1.4. Copying and pasting the method from 1.3 to 1.4 in the class does not work, there are some missing instance variables in the code. > The number of implementing classes for setSelectedMorph: does not change between 1.3 and 1.4. To me it seems that the method was moved to MorphTreeListManager in 1.4. It tried to change the receiver to invoke the method on MorphTreeListManager but there are some other problems then. > > <image.png> > > > Stef > > On Jan 31, 2013, at 7:03 AM, Usman Bhatti wrote: > > > > > > > > > On Thu, Jan 31, 2013 at 2:30 AM, Stéphane Ducasse <[hidden email]> wrote: > > Morbid fascination > > It is late so I may completely off but > > how many isKindOf: are executed in this method? > > > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > aSymbol second == #openPath]]]) > > > > when [aSymbol second == #openPath] is not openPath > > then we go to the next (not mentioning duplicated code inside the method ---- I will vomit) > > and we restart to execute isKindOf: > > > > ((aSymbol isKindOf: Array) > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > aSymbol second == #expandAllFromNode:]]]) > > > > Am'I right? > > It looks so terrible. > > > > Yes. And more so for someone not frequently working with Morphs. > > So, can someone at least see if the problem can be fixed without a large refactoring that should surely happen, but the glamour bug needs a quick fix. Or at least point to why the missing method on the receiver is abundantly found in the code? > > > > usman > > > > > > Stef > > > > > > > > > update: aSymbol > > > aSymbol == #selection ifTrue: [^ self listManager updateSelectionFromModel]. > > > ((aSymbol isKindOf: Array) and: [aSymbol size > 1 and: [aSymbol first == #selectItems and: [aSymbol second isCollection]]]) > > > ifTrue: [^ self selectedItems: aSymbol second]. > > > aSymbol == self nodeListSelector ifTrue: [^ self updateList]. > > > aSymbol == #selectAll ifTrue: [^ self listManager selectAll]. > > > aSymbol == #deselectAll ifTrue: [^ self listManager deselectAll]. > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > aSymbol second == #openPath]]]) "allow directed path opening where multiple trees exist" > > > ifTrue: > > > [^(self allNodeMorphs at: 1 ifAbsent: [^self]) > > > openPath: (aSymbol allButFirst: 2)]. > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > aSymbol second == #openItemPath]]]) "allow directed path opening where multiple trees exist" > > > ifTrue: > > > [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) > > > openItemPath: (aSymbol allButFirst: 2)]. > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > aSymbol second == #expandAllFromNode:]]]) > > > ifTrue: > > > [^ self expandAllFromNode: aSymbol third]. > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > aSymbol second == #closeItemPath]]]) "allow directed path closing where multiple trees exist" > > > ifTrue: > > > [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) > > > closeItemPath: (aSymbol allButFirst: 2)]. > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol notEmpty and: [aSymbol first == #openPath]]) > > > ifTrue: > > > [^(self allNodeMorphs at: 1 ifAbsent: [^self]) > > > openPath: aSymbol allButFirst]. > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector and: [ > > > aSymbol second == #expandRoots]]]) > > > ifTrue: > > > [^self expandRoots]. > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector]]) > > > ifTrue: [aSymbol second = #expandAll ifTrue: [^ self expandAll]. > > > aSymbol second = #collapseAll ifTrue: [^ self collapseAll]] > > > > > > > > > |
|
|
In reply to this post by Usman Bhatti
Usman
I also try just returning the container but do not know the expected result. It "works" Stef On Jan 31, 2013, at 12:48 PM, Usman Bhatti wrote: > > > > On Thu, Jan 31, 2013 at 1:20 PM, Stéphane Ducasse <[hidden email]> wrote: > Usman did you check with 1.3 to know if the method was there? > > I looked in 1.3. The method is defined on MorphTreeMorph (see sceenshot) but absent from this class in 1.4. Copying and pasting the method from 1.3 to 1.4 in the class does not work, there are some missing instance variables in the code. > The number of implementing classes for setSelectedMorph: does not change between 1.3 and 1.4. To me it seems that the method was moved to MorphTreeListManager in 1.4. It tried to change the receiver to invoke the method on MorphTreeListManager but there are some other problems then. > > <image.png> > > > Stef > > On Jan 31, 2013, at 7:03 AM, Usman Bhatti wrote: > > > > > > > > > On Thu, Jan 31, 2013 at 2:30 AM, Stéphane Ducasse <[hidden email]> wrote: > > Morbid fascination > > It is late so I may completely off but > > how many isKindOf: are executed in this method? > > > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > aSymbol second == #openPath]]]) > > > > when [aSymbol second == #openPath] is not openPath > > then we go to the next (not mentioning duplicated code inside the method ---- I will vomit) > > and we restart to execute isKindOf: > > > > ((aSymbol isKindOf: Array) > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > aSymbol second == #expandAllFromNode:]]]) > > > > Am'I right? > > It looks so terrible. > > > > Yes. And more so for someone not frequently working with Morphs. > > So, can someone at least see if the problem can be fixed without a large refactoring that should surely happen, but the glamour bug needs a quick fix. Or at least point to why the missing method on the receiver is abundantly found in the code? > > > > usman > > > > > > Stef > > > > > > > > > update: aSymbol > > > aSymbol == #selection ifTrue: [^ self listManager updateSelectionFromModel]. > > > ((aSymbol isKindOf: Array) and: [aSymbol size > 1 and: [aSymbol first == #selectItems and: [aSymbol second isCollection]]]) > > > ifTrue: [^ self selectedItems: aSymbol second]. > > > aSymbol == self nodeListSelector ifTrue: [^ self updateList]. > > > aSymbol == #selectAll ifTrue: [^ self listManager selectAll]. > > > aSymbol == #deselectAll ifTrue: [^ self listManager deselectAll]. > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > aSymbol second == #openPath]]]) "allow directed path opening where multiple trees exist" > > > ifTrue: > > > [^(self allNodeMorphs at: 1 ifAbsent: [^self]) > > > openPath: (aSymbol allButFirst: 2)]. > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > aSymbol second == #openItemPath]]]) "allow directed path opening where multiple trees exist" > > > ifTrue: > > > [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) > > > openItemPath: (aSymbol allButFirst: 2)]. > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > aSymbol second == #expandAllFromNode:]]]) > > > ifTrue: > > > [^ self expandAllFromNode: aSymbol third]. > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > aSymbol second == #closeItemPath]]]) "allow directed path closing where multiple trees exist" > > > ifTrue: > > > [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) > > > closeItemPath: (aSymbol allButFirst: 2)]. > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol notEmpty and: [aSymbol first == #openPath]]) > > > ifTrue: > > > [^(self allNodeMorphs at: 1 ifAbsent: [^self]) > > > openPath: aSymbol allButFirst]. > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector and: [ > > > aSymbol second == #expandRoots]]]) > > > ifTrue: > > > [^self expandRoots]. > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector]]) > > > ifTrue: [aSymbol second = #expandAll ifTrue: [^ self expandAll]. > > > aSymbol second = #collapseAll ifTrue: [^ self collapseAll]] > > > > > > > > > |
|
|
On Thu, Jan 31, 2013 at 10:09 PM, Stéphane Ducasse <[hidden email]> wrote: Usman It kinda "works" because DNU has disappeared. Scenario I have a list, when I select an element from the list, its properties, visu and metrics are shown by glamour.
Ideal Situation: I trigger the selection of a random node (from another visu) and its not part of the shown elements in the list (but is still part of model), it should be selected/gets focus in the list.
what this "Fix" does With the "fix", I can see the properties and visu of the selected node in my browser and DNU is gone. But selection of the node does not happen. For me, sufficient to show a problem-free UI but its not the ideal behavior.
Should this "fix" be a part of pharo? But this code needs a thorough inspection. usman
|
|
|
On Jan 31, 2013, at 7:35 PM, Usman Bhatti wrote: > > > > On Thu, Jan 31, 2013 at 10:09 PM, Stéphane Ducasse <[hidden email]> wrote: > Usman > > I also try just returning the container but do not know the expected result. > It "works" > > It kinda "works" because DNU has disappeared. > > Scenario > I have a list, when I select an element from the list, its properties, visu and metrics are shown by glamour. > > Ideal Situation: > I trigger the selection of a random node (from another visu) and its not part of the shown elements in the list (but is still part of model), it should be selected/gets focus in the list. Ok I will try to see :) > what this "Fix" does > With the "fix", I can see the properties and visu of the selected node in my browser and DNU is gone. But selection of the node does not happen. For me, sufficient to show a problem-free UI but its not the ideal behavior. > > Should this "fix" be a part of pharo? But this code needs a thorough inspection. I'm trying to fix it for real but I'm discovering this class and its superclass. > > usman > > > > Stef > > > > On Jan 31, 2013, at 12:48 PM, Usman Bhatti wrote: > > > > > > > > > On Thu, Jan 31, 2013 at 1:20 PM, Stéphane Ducasse <[hidden email]> wrote: > > Usman did you check with 1.3 to know if the method was there? > > > > I looked in 1.3. The method is defined on MorphTreeMorph (see sceenshot) but absent from this class in 1.4. Copying and pasting the method from 1.3 to 1.4 in the class does not work, there are some missing instance variables in the code. > > The number of implementing classes for setSelectedMorph: does not change between 1.3 and 1.4. To me it seems that the method was moved to MorphTreeListManager in 1.4. It tried to change the receiver to invoke the method on MorphTreeListManager but there are some other problems then. > > > > <image.png> > > > > > > Stef > > > > On Jan 31, 2013, at 7:03 AM, Usman Bhatti wrote: > > > > > > > > > > > > > > On Thu, Jan 31, 2013 at 2:30 AM, Stéphane Ducasse <[hidden email]> wrote: > > > Morbid fascination > > > It is late so I may completely off but > > > how many isKindOf: are executed in this method? > > > > > > > ((aSymbol isKindOf: Array) > > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > > aSymbol second == #openPath]]]) > > > > > > when [aSymbol second == #openPath] is not openPath > > > then we go to the next (not mentioning duplicated code inside the method ---- I will vomit) > > > and we restart to execute isKindOf: > > > > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > aSymbol second == #expandAllFromNode:]]]) > > > > > > Am'I right? > > > It looks so terrible. > > > > > > Yes. And more so for someone not frequently working with Morphs. > > > So, can someone at least see if the problem can be fixed without a large refactoring that should surely happen, but the glamour bug needs a quick fix. Or at least point to why the missing method on the receiver is abundantly found in the code? > > > > > > usman > > > > > > > > > Stef > > > > > > > > > > > > > update: aSymbol > > > > aSymbol == #selection ifTrue: [^ self listManager updateSelectionFromModel]. > > > > ((aSymbol isKindOf: Array) and: [aSymbol size > 1 and: [aSymbol first == #selectItems and: [aSymbol second isCollection]]]) > > > > ifTrue: [^ self selectedItems: aSymbol second]. > > > > aSymbol == self nodeListSelector ifTrue: [^ self updateList]. > > > > aSymbol == #selectAll ifTrue: [^ self listManager selectAll]. > > > > aSymbol == #deselectAll ifTrue: [^ self listManager deselectAll]. > > > > ((aSymbol isKindOf: Array) > > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > > aSymbol second == #openPath]]]) "allow directed path opening where multiple trees exist" > > > > ifTrue: > > > > [^(self allNodeMorphs at: 1 ifAbsent: [^self]) > > > > openPath: (aSymbol allButFirst: 2)]. > > > > ((aSymbol isKindOf: Array) > > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > > aSymbol second == #openItemPath]]]) "allow directed path opening where multiple trees exist" > > > > ifTrue: > > > > [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) > > > > openItemPath: (aSymbol allButFirst: 2)]. > > > > ((aSymbol isKindOf: Array) > > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > > aSymbol second == #expandAllFromNode:]]]) > > > > ifTrue: > > > > [^ self expandAllFromNode: aSymbol third]. > > > > ((aSymbol isKindOf: Array) > > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > > aSymbol second == #closeItemPath]]]) "allow directed path closing where multiple trees exist" > > > > ifTrue: > > > > [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) > > > > closeItemPath: (aSymbol allButFirst: 2)]. > > > > ((aSymbol isKindOf: Array) > > > > and: [aSymbol notEmpty and: [aSymbol first == #openPath]]) > > > > ifTrue: > > > > [^(self allNodeMorphs at: 1 ifAbsent: [^self]) > > > > openPath: aSymbol allButFirst]. > > > > ((aSymbol isKindOf: Array) > > > > and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector and: [ > > > > aSymbol second == #expandRoots]]]) > > > > ifTrue: > > > > [^self expandRoots]. > > > > ((aSymbol isKindOf: Array) > > > > and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector]]) > > > > ifTrue: [aSymbol second = #expandAll ifTrue: [^ self expandAll]. > > > > aSymbol second = #collapseAll ifTrue: [^ self collapseAll]] > > > > > > > > > > > > > > > > > > |
|
|
In reply to this post by Usman Bhatti
On Jan 31, 2013, at 7:35 PM, Usman Bhatti wrote: > > > > On Thu, Jan 31, 2013 at 10:09 PM, Stéphane Ducasse <[hidden email]> wrote: > Usman > > I also try just returning the container but do not know the expected result. > It "works" > > It kinda "works" because DNU has disappeared. > > Scenario > I have a list, when I select an element from the list, its properties, visu and metrics are shown by glamour. > > Ideal Situation: > I trigger the selection of a random node (from another visu) and its not part of the shown elements in the list (but is still part of model), it should be selected/gets focus in the list. I do not understand how a container setSelectedMorph: nil can select a node :) > > what this "Fix" does > With the "fix", I can see the properties and visu of the selected node in my browser and DNU is gone. But selection of the node does not happen. For me, sufficient to show a problem-free UI but its not the ideal behavior. > > Should this "fix" be a part of pharo? But this code needs a thorough inspection. > > usman > > > > Stef > > > > On Jan 31, 2013, at 12:48 PM, Usman Bhatti wrote: > > > > > > > > > On Thu, Jan 31, 2013 at 1:20 PM, Stéphane Ducasse <[hidden email]> wrote: > > Usman did you check with 1.3 to know if the method was there? > > > > I looked in 1.3. The method is defined on MorphTreeMorph (see sceenshot) but absent from this class in 1.4. Copying and pasting the method from 1.3 to 1.4 in the class does not work, there are some missing instance variables in the code. > > The number of implementing classes for setSelectedMorph: does not change between 1.3 and 1.4. To me it seems that the method was moved to MorphTreeListManager in 1.4. It tried to change the receiver to invoke the method on MorphTreeListManager but there are some other problems then. > > > > <image.png> > > > > > > Stef > > > > On Jan 31, 2013, at 7:03 AM, Usman Bhatti wrote: > > > > > > > > > > > > > > On Thu, Jan 31, 2013 at 2:30 AM, Stéphane Ducasse <[hidden email]> wrote: > > > Morbid fascination > > > It is late so I may completely off but > > > how many isKindOf: are executed in this method? > > > > > > > ((aSymbol isKindOf: Array) > > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > > aSymbol second == #openPath]]]) > > > > > > when [aSymbol second == #openPath] is not openPath > > > then we go to the next (not mentioning duplicated code inside the method ---- I will vomit) > > > and we restart to execute isKindOf: > > > > > > ((aSymbol isKindOf: Array) > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > aSymbol second == #expandAllFromNode:]]]) > > > > > > Am'I right? > > > It looks so terrible. > > > > > > Yes. And more so for someone not frequently working with Morphs. > > > So, can someone at least see if the problem can be fixed without a large refactoring that should surely happen, but the glamour bug needs a quick fix. Or at least point to why the missing method on the receiver is abundantly found in the code? > > > > > > usman > > > > > > > > > Stef > > > > > > > > > > > > > update: aSymbol > > > > aSymbol == #selection ifTrue: [^ self listManager updateSelectionFromModel]. > > > > ((aSymbol isKindOf: Array) and: [aSymbol size > 1 and: [aSymbol first == #selectItems and: [aSymbol second isCollection]]]) > > > > ifTrue: [^ self selectedItems: aSymbol second]. > > > > aSymbol == self nodeListSelector ifTrue: [^ self updateList]. > > > > aSymbol == #selectAll ifTrue: [^ self listManager selectAll]. > > > > aSymbol == #deselectAll ifTrue: [^ self listManager deselectAll]. > > > > ((aSymbol isKindOf: Array) > > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > > aSymbol second == #openPath]]]) "allow directed path opening where multiple trees exist" > > > > ifTrue: > > > > [^(self allNodeMorphs at: 1 ifAbsent: [^self]) > > > > openPath: (aSymbol allButFirst: 2)]. > > > > ((aSymbol isKindOf: Array) > > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > > aSymbol second == #openItemPath]]]) "allow directed path opening where multiple trees exist" > > > > ifTrue: > > > > [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) > > > > openItemPath: (aSymbol allButFirst: 2)]. > > > > ((aSymbol isKindOf: Array) > > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > > aSymbol second == #expandAllFromNode:]]]) > > > > ifTrue: > > > > [^ self expandAllFromNode: aSymbol third]. > > > > ((aSymbol isKindOf: Array) > > > > and: [aSymbol size > 1 and: [aSymbol first == self nodeListSelector and: [ > > > > aSymbol second == #closeItemPath]]]) "allow directed path closing where multiple trees exist" > > > > ifTrue: > > > > [^ (self allNodeMorphs at: 1 ifAbsent: [^self]) > > > > closeItemPath: (aSymbol allButFirst: 2)]. > > > > ((aSymbol isKindOf: Array) > > > > and: [aSymbol notEmpty and: [aSymbol first == #openPath]]) > > > > ifTrue: > > > > [^(self allNodeMorphs at: 1 ifAbsent: [^self]) > > > > openPath: aSymbol allButFirst]. > > > > ((aSymbol isKindOf: Array) > > > > and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector and: [ > > > > aSymbol second == #expandRoots]]]) > > > > ifTrue: > > > > [^self expandRoots]. > > > > ((aSymbol isKindOf: Array) > > > > and: [aSymbol size = 2 and: [aSymbol first = self nodeListSelector]]) > > > > ifTrue: [aSymbol second = #expandAll ifTrue: [^ self expandAll]. > > > > aSymbol second = #collapseAll ifTrue: [^ self collapseAll]] > > > > > > > > > > > > > > > > > > |
|
|
On Fri, Feb 1, 2013 at 12:52 AM, Stéphane Ducasse <[hidden email]> wrote:
Agreed :-). I was looking for the behavior which says, the node that you are trying to select is not shown, so create a morph of the node in the tree, select it and show it and I couldn't find this behavior, I doubt it exists.
|
«
Return to Pharo Smalltalk Developers
|
1 view|%1 views
| Free forum by Nabble | Edit this page |