[vwnc] Did RBParser changed in 7.6

> Hi
>
> I encountered a strange bug that you can reproduce with the two  
> methods below.
> Apparently RBParser changed in 7.6.
> Is it true?
>
> Then the bug is fishy because we get a NonBoolean receiver but when I  
> inspect and executed each of the statement
>
> of
>
> patchLiteralArrayToken
> (self patchLiteralArrayEmbeddedAssignmentToken
> or:
> [self patchLiteralArraySpecialToken
> or:
> [self patchLiteralArrayDottedNameToken
> ifFalse: [self     patchLiteralArraySymbolToken     ]]])
> ifFalse:
> [^self parserError: #InvalidToken << #browser >> 'Invalid token']
>
> I get booleans.
> and patchLiteralArraySymbolToken is returning boolean.
> So I do not understand why I get an exception telling to me that  
> receiver is not a boolean.
> In addition we looked at the byte and it looks correct.
> May be we missed something obvious.

> Then the bug is fishy because we get a NonBoolean receiver but when I  
> inspect and executed each of the statement
>
> of
>
> patchLiteralArrayToken
> (self patchLiteralArrayEmbeddedAssignmentToken
> or:
> [self patchLiteralArraySpecialToken
> or:
> [self patchLiteralArrayDottedNameToken
> ifFalse: [self     patchLiteralArraySymbolToken     ]]])
> ifFalse:
> [^self parserError: #InvalidToken << #browser >> 'Invalid token']
>
> I get booleans.
> and patchLiteralArraySymbolToken is returning boolean.

> Hi
>
> I encountered a strange bug that you can reproduce with the two
> methods below.
> Apparently RBParser changed in 7.6.
> Is it true?
>
> Then the bug is fishy because we get a NonBoolean receiver but when I
> inspect and executed each of the statement
>
> of
>
> patchLiteralArrayToken
> (self patchLiteralArrayEmbeddedAssignmentToken
> or:
> [self patchLiteralArraySpecialToken
> or:
> [self patchLiteralArrayDottedNameToken
> ifFalse: [self     patchLiteralArraySymbolToken     ]]])
> ifFalse:
> [^self parserError: #InvalidToken << #browser >> 'Invalid token']
>
> I get booleans.
> and patchLiteralArraySymbolToken is returning boolean.
> So I do not understand why I get an exception telling to me that
> receiver is not a boolean.
> In addition we looked at the byte and it looks correct.
> May be we missed something obvious.
>
> Stef
>
>
>
> test
> "self new test"
> ^Refactory.Browser.RBParser parseMethod: (ZorkTest sourceCodeAt:
> #modelVersion3)
> onError: [:aString :pos | ^nil]
>
> modelVersion3
> ^
> #(Moose.Model (entity
> (FAMIX.Namespace (id: 1) (name aNamespace))
>
> (FAMIX.Class (id: 2) (name ClassA) (belongsTo (idref: 1)) (packagedIn
> (idref: 201)))
> (FAMIX.Method (name methodA1) (signature 'methodA1()') (belongsTo
> (idref: 2)) (LOC 4))
> (FAMIX.Method (name methodA2) (signature 'methodA2()') (belongsTo
> (idref: 2)) (LOC 5))
> (FAMIX.Attribute (name attributeA1) (belongsTo (idref: 2)))
> (FAMIX.Attribute (name attributeA2) (belongsTo (idref: 2)))
>
> (FAMIX.Class (id: 3) (name ClassB) (belongsTo (idref: 1)) (packagedIn
> (idref: 201)))
> (FAMIX.Attribute (name attributeB1) (belongsTo (idref: 3)))
>
> (FAMIX.InheritanceDefinition (subclass (idref: 3)) (superclass
> (idref: 2)))
>
> (FAMIX.Class (id: 4) (name ClassC) (belongsTo (idref: 1)) (packagedIn
> (idref: 202)))
> (FAMIX.Method (name methodC1) (signature 'methodC1()') (belongsTo
> (idref: 4)) (LOC 4))
> (FAMIX.Method (name methodC2) (signature 'methodC2()') (belongsTo
> (idref: 4)) (LOC 10))
> (FAMIX.Method (name methodC3) (signature 'methodC3()') (belongsTo
> (idref: 4)) (LOC 7))
> (FAMIX.Attribute (name attributeC1) (belongsTo (idref: 4)))
> (FAMIX.Attribute (name attributeC2) (belongsTo (idref: 4)))
>
> (FAMIX.InheritanceDefinition (subclass (idref: 4)) (superclass
> (idref: 2)))
>
> (FAMIX.Class (id: 5) (name ClassD) (belongsTo (idref: 1)) (packagedIn
> (idref: 202)))
> (FAMIX.Method (name methodD1) (signature 'methodD1()') (belongsTo
> (idref: 5)) (LOC 3))
>
> (FAMIX.InheritanceDefinition (subclass (idref: 5)) (superclass
> (idref: 4)))
>
> (FAMIX.Package (id: 201) (name aPackage))
> (FAMIX.Package (id: 202) (name anotherPackage))
> ))

>
> On Apr 21, 2008, at 1:36 PM, stéphane ducasse wrote:
>
>> Hi
>>
>> I encountered a strange bug that you can reproduce with the two
>> methods below.
>> Apparently RBParser changed in 7.6.
>> Is it true?
>>
>> Then the bug is fishy because we get a NonBoolean receiver but when I
>> inspect and executed each of the statement
>>
>> of
>>
>> patchLiteralArrayToken
>> (self patchLiteralArrayEmbeddedAssignmentToken
>> or:
>> [self patchLiteralArraySpecialToken
>> or:
>> [self patchLiteralArrayDottedNameToken
>> ifFalse: [self     patchLiteralArraySymbolToken     ]]])
>> ifFalse:
>> [^self parserError: #InvalidToken << #browser >> 'Invalid token']
>>
>> I get booleans.
>> and patchLiteralArraySymbolToken is returning boolean.
>> So I do not understand why I get an exception telling to me that
>> receiver is not a boolean.
>> In addition we looked at the byte and it looks correct.
>> May be we missed something obvious.
>>
>> Stef
>>
>>
>>
>> test
>> "self new test"
>> ^Refactory.Browser.RBParser parseMethod: (ZorkTest sourceCodeAt:
>> #modelVersion3)
>> onError: [:aString :pos | ^nil]
>>
>> modelVersion3
>> ^
>> #(Moose.Model (entity
>> (FAMIX.Namespace (id: 1) (name aNamespace))
>>
>> (FAMIX.Class (id: 2) (name ClassA) (belongsTo (idref: 1))  
>> (packagedIn
>> (idref: 201)))
>> (FAMIX.Method (name methodA1) (signature 'methodA1()') (belongsTo
>> (idref: 2)) (LOC 4))
>> (FAMIX.Method (name methodA2) (signature 'methodA2()') (belongsTo
>> (idref: 2)) (LOC 5))
>> (FAMIX.Attribute (name attributeA1) (belongsTo (idref: 2)))
>> (FAMIX.Attribute (name attributeA2) (belongsTo (idref: 2)))
>>
>> (FAMIX.Class (id: 3) (name ClassB) (belongsTo (idref: 1))  
>> (packagedIn
>> (idref: 201)))
>> (FAMIX.Attribute (name attributeB1) (belongsTo (idref: 3)))
>>
>> (FAMIX.InheritanceDefinition (subclass (idref: 3)) (superclass
>> (idref: 2)))
>>
>> (FAMIX.Class (id: 4) (name ClassC) (belongsTo (idref: 1))  
>> (packagedIn
>> (idref: 202)))
>> (FAMIX.Method (name methodC1) (signature 'methodC1()') (belongsTo
>> (idref: 4)) (LOC 4))
>> (FAMIX.Method (name methodC2) (signature 'methodC2()') (belongsTo
>> (idref: 4)) (LOC 10))
>> (FAMIX.Method (name methodC3) (signature 'methodC3()') (belongsTo
>> (idref: 4)) (LOC 7))
>> (FAMIX.Attribute (name attributeC1) (belongsTo (idref: 4)))
>> (FAMIX.Attribute (name attributeC2) (belongsTo (idref: 4)))
>>
>> (FAMIX.InheritanceDefinition (subclass (idref: 4)) (superclass
>> (idref: 2)))
>>
>> (FAMIX.Class (id: 5) (name ClassD) (belongsTo (idref: 1))  
>> (packagedIn
>> (idref: 202)))
>> (FAMIX.Method (name methodD1) (signature 'methodD1()') (belongsTo
>> (idref: 5)) (LOC 3))
>>
>> (FAMIX.InheritanceDefinition (subclass (idref: 5)) (superclass
>> (idref: 4)))
>>
>> (FAMIX.Package (id: 201) (name aPackage))
>> (FAMIX.Package (id: 202) (name anotherPackage))
>> ))
>
> Nope, just a 7.6 bug. I introduced it. I'll add another test now. And
> whip self. :)
>
> Stefan Schmiedl noticed this the other day too.
>
> Thanks for reporting it.
>
> --
> Travis Griggs
> Objologist
> "An idea, like a ghost, must be spoken to a little before it will
> explain itself." - Charles Dickens
>
>
> _______________________________________________
> vwnc mailing list
> [hidden email]
> http://lists.cs.uiuc.edu/mailman/listinfo/vwnc
>

>
> On Apr 21, 2008, at 1:36 PM, stéphane ducasse wrote:
>
>> Hi
>>
>> I encountered a strange bug that you can reproduce with the two
>> methods below.
>> Apparently RBParser changed in 7.6.
>> Is it true?
>>
>> Then the bug is fishy because we get a NonBoolean receiver but when I
>> inspect and executed each of the statement
>>
>> of
>>
>> patchLiteralArrayToken
>> (self patchLiteralArrayEmbeddedAssignmentToken
>> or:
>> [self patchLiteralArraySpecialToken
>> or:
>> [self patchLiteralArrayDottedNameToken
>> ifFalse: [self     patchLiteralArraySymbolToken     ]]])
>> ifFalse:
>> [^self parserError: #InvalidToken << #browser >> 'Invalid token']
>>
>> I get booleans.
>> and patchLiteralArraySymbolToken is returning boolean.
>> So I do not understand why I get an exception telling to me that
>> receiver is not a boolean.
>> In addition we looked at the byte and it looks correct.
>> May be we missed something obvious.
>>
>> Stef
>>
>>
>>
>> test
>> "self new test"
>> ^Refactory.Browser.RBParser parseMethod: (ZorkTest sourceCodeAt:
>> #modelVersion3)
>> onError: [:aString :pos | ^nil]
>>
>> modelVersion3
>> ^
>> #(Moose.Model (entity
>> (FAMIX.Namespace (id: 1) (name aNamespace))
>>
>> (FAMIX.Class (id: 2) (name ClassA) (belongsTo (idref: 1))  
>> (packagedIn
>> (idref: 201)))
>> (FAMIX.Method (name methodA1) (signature 'methodA1()') (belongsTo
>> (idref: 2)) (LOC 4))
>> (FAMIX.Method (name methodA2) (signature 'methodA2()') (belongsTo
>> (idref: 2)) (LOC 5))
>> (FAMIX.Attribute (name attributeA1) (belongsTo (idref: 2)))
>> (FAMIX.Attribute (name attributeA2) (belongsTo (idref: 2)))
>>
>> (FAMIX.Class (id: 3) (name ClassB) (belongsTo (idref: 1))  
>> (packagedIn
>> (idref: 201)))
>> (FAMIX.Attribute (name attributeB1) (belongsTo (idref: 3)))
>>
>> (FAMIX.InheritanceDefinition (subclass (idref: 3)) (superclass
>> (idref: 2)))
>>
>> (FAMIX.Class (id: 4) (name ClassC) (belongsTo (idref: 1))  
>> (packagedIn
>> (idref: 202)))
>> (FAMIX.Method (name methodC1) (signature 'methodC1()') (belongsTo
>> (idref: 4)) (LOC 4))
>> (FAMIX.Method (name methodC2) (signature 'methodC2()') (belongsTo
>> (idref: 4)) (LOC 10))
>> (FAMIX.Method (name methodC3) (signature 'methodC3()') (belongsTo
>> (idref: 4)) (LOC 7))
>> (FAMIX.Attribute (name attributeC1) (belongsTo (idref: 4)))
>> (FAMIX.Attribute (name attributeC2) (belongsTo (idref: 4)))
>>
>> (FAMIX.InheritanceDefinition (subclass (idref: 4)) (superclass
>> (idref: 2)))
>>
>> (FAMIX.Class (id: 5) (name ClassD) (belongsTo (idref: 1))  
>> (packagedIn
>> (idref: 202)))
>> (FAMIX.Method (name methodD1) (signature 'methodD1()') (belongsTo
>> (idref: 5)) (LOC 3))
>>
>> (FAMIX.InheritanceDefinition (subclass (idref: 5)) (superclass
>> (idref: 4)))
>>
>> (FAMIX.Package (id: 201) (name aPackage))
>> (FAMIX.Package (id: 202) (name anotherPackage))
>> ))
>
> Nope, just a 7.6 bug. I introduced it. I'll add another test now. And
> whip self. :)
>
> Stefan Schmiedl noticed this the other day too.
>
> Thanks for reporting it.
>
> --
> Travis Griggs
> Objologist
> "An idea, like a ghost, must be spoken to a little before it will
> explain itself." - Charles Dickens
>
>
> _______________________________________________
> vwnc mailing list
> [hidden email]
> http://lists.cs.uiuc.edu/mailman/listinfo/vwnc
>