TestRunner not assigning strings/remembering old values of temp vars.

> I am testing the CryptoDESTest.  I am getting some strange behavior.  When I
> run the following code in a Workspace, repeatably, I get the same results:
>
> code:
>
>        | plain key d |
>        plain := 'squeaker'.
>        key := 'hacking!'.
>        Transcript cr; show: 'plain: ', plain printString.
>        d := DES key: key.
>        d encryptBlock: plain.
>        Transcript cr; show: 'plain: ', plain printString.
>        self assert: plain asByteArray = #[64 222 70 13 59 145 76 186].
>        plain destroy.
>
> output:
>
> plain: 'squeaker'
> plain: '@ÞF
> ;‘Lº'
> plain: 'squeaker'
> plain: '@ÞF
> ;‘Lº'
> plain: 'squeaker'
> plain: '@ÞF
> ;‘Lº'
>
> However, when I run this as a test method:
>
> testDES3
>        | plain key d |
>        plain := 'squeaker'.
>        key := 'hacking!'.
>        Transcript cr; show: 'plain: ', plain printString.
>        d := DES key: key.
>        d encryptBlock: plain.
>        Transcript cr; show: 'plain: ', plain printString.
>        self assert: plain asByteArray = #[64 222 70 13 59 145 76 186].
>               plain destroy.
>
> I get the following output and the test fails:
>
> plain: 'squeaker'
> plain: '@ÞF
> ;‘Lº'
> plain: '        '
> plain: '£ƒ¼ïùmÞ$'
> plain: '£ƒ¼ïùmÞ$'
> plain: '§,}ºé¨öú'
> plain: '§,}ºé¨öú'
> plain: '3qX¦wRË|'
> plain: '3qX¦wRË|'
> plain: 'Ê0ê^ó{ì '
> plain: 'Ê0ê^ó{ì '
> plain: 'çî³›,ίw'
>
> Why is test runner remembering the old value and not setting it with: plain
> := 'squeaker'.?
>
> TIA,
> Rob
>
>

> Of course, 'squeaker' is a literal stored in the CompiledMethod.
> If you write over it, then you break the literal for next method
> execution.
>
> Each time you evaluate in a workspace, you create a new CompiledMethod
> and a new literal.
> When you compile in a class, that's different because you compiled
> only once and execute many.
>
> Nicolas
>
> 2010/7/25 Rob Withers <[hidden email]>:
>> I am testing the CryptoDESTest.  I am getting some strange behavior.
>> When I
>> run the following code in a Workspace, repeatably, I get the same
>> results:
>>
>> code:
>>
>>        | plain key d |
>>        plain := 'squeaker'.
>>        key := 'hacking!'.
>>        Transcript cr; show: 'plain: ', plain printString.
>>        d := DES key: key.
>>        d encryptBlock: plain.
>>        Transcript cr; show: 'plain: ', plain printString.
>>        self assert: plain asByteArray = #[64 222 70 13 59 145 76 186].
>>        plain destroy.
>>
>> output:
>>
>> plain: 'squeaker'
>> plain: '@ÞF
>> ;‘Lº'
>> plain: 'squeaker'
>> plain: '@ÞF
>> ;‘Lº'
>> plain: 'squeaker'
>> plain: '@ÞF
>> ;‘Lº'
>>
>> However, when I run this as a test method:
>>
>> testDES3
>>        | plain key d |
>>        plain := 'squeaker'.
>>        key := 'hacking!'.
>>        Transcript cr; show: 'plain: ', plain printString.
>>        d := DES key: key.
>>        d encryptBlock: plain.
>>        Transcript cr; show: 'plain: ', plain printString.
>>        self assert: plain asByteArray = #[64 222 70 13 59 145 76 186].
>>               plain destroy.
>>
>> I get the following output and the test fails:
>>
>> plain: 'squeaker'
>> plain: '@ÞF
>> ;‘Lº'
>> plain: '        '
>> plain: '£ƒ¼ïùmÞ$'
>> plain: '£ƒ¼ïùmÞ$'
>> plain: '§,}ºé¨öú'
>> plain: '§,}ºé¨öú'
>> plain: '3qX¦wRË|'
>> plain: '3qX¦wRË|'
>> plain: 'Ê0ê^ó{ì '
>> plain: 'Ê0ê^ó{ì '
>> plain: 'çî³›,ίw'
>>
>> Why is test runner remembering the old value and not setting it with:
>> plain
>> := 'squeaker'.?
>>
>> TIA,
>> Rob
>>
>>
>

> Thanks, that makes sense.  Isn't that a bug since Strings are mutable?
>
> --------------------------------------------------
> From: "Nicolas Cellier" <[hidden email]>
> Sent: Sunday, July 25, 2010 3:51 AM
> To: "The general-purpose Squeak developers list"
> <[hidden email]>
> Subject: Re: [squeak-dev] TestRunner not assigning strings/remembering old
> values of temp vars.
>
>> Of course, 'squeaker' is a literal stored in the CompiledMethod.
>> If you write over it, then you break the literal for next method
>> execution.
>>
>> Each time you evaluate in a workspace, you create a new CompiledMethod
>> and a new literal.
>> When you compile in a class, that's different because you compiled
>> only once and execute many.
>>
>> Nicolas
>>
>> 2010/7/25 Rob Withers <[hidden email]>:
>>>
>>> I am testing the CryptoDESTest.  I am getting some strange behavior. When
>>> I
>>> run the following code in a Workspace, repeatably, I get the same
>>> results:
>>>
>>> code:
>>>
>>>       | plain key d |
>>>       plain := 'squeaker'.
>>>       key := 'hacking!'.
>>>       Transcript cr; show: 'plain: ', plain printString.
>>>       d := DES key: key.
>>>       d encryptBlock: plain.
>>>       Transcript cr; show: 'plain: ', plain printString.
>>>       self assert: plain asByteArray = #[64 222 70 13 59 145 76 186].
>>>       plain destroy.
>>>
>>> output:
>>>
>>> plain: 'squeaker'
>>> plain: '@ÞF
>>> ;‘Lº'
>>> plain: 'squeaker'
>>> plain: '@ÞF
>>> ;‘Lº'
>>> plain: 'squeaker'
>>> plain: '@ÞF
>>> ;‘Lº'
>>>
>>> However, when I run this as a test method:
>>>
>>> testDES3
>>>       | plain key d |
>>>       plain := 'squeaker'.
>>>       key := 'hacking!'.
>>>       Transcript cr; show: 'plain: ', plain printString.
>>>       d := DES key: key.
>>>       d encryptBlock: plain.
>>>       Transcript cr; show: 'plain: ', plain printString.
>>>       self assert: plain asByteArray = #[64 222 70 13 59 145 76 186].
>>>              plain destroy.
>>>
>>> I get the following output and the test fails:
>>>
>>> plain: 'squeaker'
>>> plain: '@ÞF
>>> ;‘Lº'
>>> plain: '        '
>>> plain: '£ƒ¼ïùmÞ$'
>>> plain: '£ƒ¼ïùmÞ$'
>>> plain: '§,}ºé¨öú'
>>> plain: '§,}ºé¨öú'
>>> plain: '3qX¦wRË|'
>>> plain: '3qX¦wRË|'
>>> plain: 'Ê0ê^ó{ì '
>>> plain: 'Ê0ê^ó{ì '
>>> plain: 'çî³›,ίw'
>>>
>>> Why is test runner remembering the old value and not setting it with:
>>> plain
>>> := 'squeaker'.?
>>>
>>> TIA,
>>> Rob
>>>
>>>
>>
>
>

> There is no immutable support in Squeak.
> Smalltalk custom is to avoid overwriting literals like String Array...
> ...Same for instances of other classes like Point Rectangle which are
> likely to be shared by several objects.
> Most algorithms will rather modify and answer a copy.
> This default behaviour has a cost (creates lots of objects) but is
> fool proof (no danger to overwrite a shared information).
>
> For DES, I don't know, maybe it's just an optimization you must be aware
> of.
> In that case, you are responsible for creating the copy by yourself
> when required.

>
> Nicolas
>
> 2010/7/25 Rob Withers <[hidden email]>:
>> Thanks, that makes sense.  Isn't that a bug since Strings are mutable?
>>
>> --------------------------------------------------
>> From: "Nicolas Cellier" <[hidden email]>
>> Sent: Sunday, July 25, 2010 3:51 AM
>> To: "The general-purpose Squeak developers list"
>> <[hidden email]>
>> Subject: Re: [squeak-dev] TestRunner not assigning strings/remembering
>> old
>> values of temp vars.
>>
>>> Of course, 'squeaker' is a literal stored in the CompiledMethod.
>>> If you write over it, then you break the literal for next method
>>> execution.
>>>
>>> Each time you evaluate in a workspace, you create a new CompiledMethod
>>> and a new literal.
>>> When you compile in a class, that's different because you compiled
>>> only once and execute many.
>>>
>>> Nicolas
>>>
>>> 2010/7/25 Rob Withers <[hidden email]>:
>>>>
>>>> I am testing the CryptoDESTest.  I am getting some strange behavior.
>>>> When
>>>> I
>>>> run the following code in a Workspace, repeatably, I get the same
>>>> results:
>>>>
>>>> code:
>>>>
>>>>       | plain key d |
>>>>       plain := 'squeaker'.
>>>>       key := 'hacking!'.
>>>>       Transcript cr; show: 'plain: ', plain printString.
>>>>       d := DES key: key.
>>>>       d encryptBlock: plain.
>>>>       Transcript cr; show: 'plain: ', plain printString.
>>>>       self assert: plain asByteArray = #[64 222 70 13 59 145 76 186].
>>>>       plain destroy.
>>>>
>>>> output:
>>>>
>>>> plain: 'squeaker'
>>>> plain: '@ÞF
>>>> ;‘Lº'
>>>> plain: 'squeaker'
>>>> plain: '@ÞF
>>>> ;‘Lº'
>>>> plain: 'squeaker'
>>>> plain: '@ÞF
>>>> ;‘Lº'
>>>>
>>>> However, when I run this as a test method:
>>>>
>>>> testDES3
>>>>       | plain key d |
>>>>       plain := 'squeaker'.
>>>>       key := 'hacking!'.
>>>>       Transcript cr; show: 'plain: ', plain printString.
>>>>       d := DES key: key.
>>>>       d encryptBlock: plain.
>>>>       Transcript cr; show: 'plain: ', plain printString.
>>>>       self assert: plain asByteArray = #[64 222 70 13 59 145 76 186].
>>>>              plain destroy.
>>>>
>>>> I get the following output and the test fails:
>>>>
>>>> plain: 'squeaker'
>>>> plain: '@ÞF
>>>> ;‘Lº'
>>>> plain: '        '
>>>> plain: '£ƒ¼ïùmÞ$'
>>>> plain: '£ƒ¼ïùmÞ$'
>>>> plain: '§,}ºé¨öú'
>>>> plain: '§,}ºé¨öú'
>>>> plain: '3qX¦wRË|'
>>>> plain: '3qX¦wRË|'
>>>> plain: 'Ê0ê^ó{ì '
>>>> plain: 'Ê0ê^ó{ì '
>>>> plain: 'çî³›,ίw'
>>>>
>>>> Why is test runner remembering the old value and not setting it with:
>>>> plain
>>>> := 'squeaker'.?
>>>>
>>>> TIA,
>>>> Rob
>>>>
>>>>
>>>
>>
>>
>

>
>
> --------------------------------------------------
> From: "Nicolas Cellier" <[hidden email]>
> Sent: Sunday, July 25, 2010 4:09 AM
> To: "The general-purpose Squeak developers list"
> <[hidden email]>
> Subject: Re: [squeak-dev] TestRunner not assigning strings/remembering old
> values of temp vars.
>
>> There is no immutable support in Squeak.
>> Smalltalk custom is to avoid overwriting literals like String Array...
>> ...Same for instances of other classes like Point Rectangle which are
>> likely to be shared by several objects.
>> Most algorithms will rather modify and answer a copy.
>> This default behaviour has a cost (creates lots of objects) but is
>> fool proof (no danger to overwrite a shared information).
>>
>> For DES, I don't know, maybe it's just an optimization you must be aware
>> of.
>> In that case, you are responsible for creating the copy by yourself
>> when required.
>
>
> You mean before calling encryptBlock:/decryptBlock: with string data, we
> should copy the string data.  That's unfortunate.  It works fine with
> ByteArrays.
>

> Thanks for the info!
> Rob
>
>>
>> Nicolas
>>
>> 2010/7/25 Rob Withers <[hidden email]>:
>>>
>>> Thanks, that makes sense.  Isn't that a bug since Strings are mutable?
>>>
>>> --------------------------------------------------
>>> From: "Nicolas Cellier" <[hidden email]>
>>> Sent: Sunday, July 25, 2010 3:51 AM
>>> To: "The general-purpose Squeak developers list"
>>> <[hidden email]>
>>> Subject: Re: [squeak-dev] TestRunner not assigning strings/remembering
>>> old
>>> values of temp vars.
>>>
>>>> Of course, 'squeaker' is a literal stored in the CompiledMethod.
>>>> If you write over it, then you break the literal for next method
>>>> execution.
>>>>
>>>> Each time you evaluate in a workspace, you create a new CompiledMethod
>>>> and a new literal.
>>>> When you compile in a class, that's different because you compiled
>>>> only once and execute many.
>>>>
>>>> Nicolas
>>>>
>>>> 2010/7/25 Rob Withers <[hidden email]>:
>>>>>
>>>>> I am testing the CryptoDESTest.  I am getting some strange behavior.
>>>>> When
>>>>> I
>>>>> run the following code in a Workspace, repeatably, I get the same
>>>>> results:
>>>>>
>>>>> code:
>>>>>
>>>>>      | plain key d |
>>>>>      plain := 'squeaker'.
>>>>>      key := 'hacking!'.
>>>>>      Transcript cr; show: 'plain: ', plain printString.
>>>>>      d := DES key: key.
>>>>>      d encryptBlock: plain.
>>>>>      Transcript cr; show: 'plain: ', plain printString.
>>>>>      self assert: plain asByteArray = #[64 222 70 13 59 145 76 186].
>>>>>      plain destroy.
>>>>>
>>>>> output:
>>>>>
>>>>> plain: 'squeaker'
>>>>> plain: '@ÞF
>>>>> ;‘Lº'
>>>>> plain: 'squeaker'
>>>>> plain: '@ÞF
>>>>> ;‘Lº'
>>>>> plain: 'squeaker'
>>>>> plain: '@ÞF
>>>>> ;‘Lº'
>>>>>
>>>>> However, when I run this as a test method:
>>>>>
>>>>> testDES3
>>>>>      | plain key d |
>>>>>      plain := 'squeaker'.
>>>>>      key := 'hacking!'.
>>>>>      Transcript cr; show: 'plain: ', plain printString.
>>>>>      d := DES key: key.
>>>>>      d encryptBlock: plain.
>>>>>      Transcript cr; show: 'plain: ', plain printString.
>>>>>      self assert: plain asByteArray = #[64 222 70 13 59 145 76 186].
>>>>>             plain destroy.
>>>>>
>>>>> I get the following output and the test fails:
>>>>>
>>>>> plain: 'squeaker'
>>>>> plain: '@ÞF
>>>>> ;‘Lº'
>>>>> plain: '        '
>>>>> plain: '£ƒ¼ïùmÞ$'
>>>>> plain: '£ƒ¼ïùmÞ$'
>>>>> plain: '§,}ºé¨öú'
>>>>> plain: '§,}ºé¨öú'
>>>>> plain: '3qX¦wRË|'
>>>>> plain: '3qX¦wRË|'
>>>>> plain: 'Ê0ê^ó{ì '
>>>>> plain: 'Ê0ê^ó{ì '
>>>>> plain: 'çî³›,ίw'
>>>>>
>>>>> Why is test runner remembering the old value and not setting it with:
>>>>> plain
>>>>> := 'squeaker'.?
>>>>>
>>>>> TIA,
>>>>> Rob
>>>>>
>>>>>
>>>>
>>>
>>>
>>
>
>

> 2010/7/25 Rob Withers <[hidden email]>:
>>
>>
>> --------------------------------------------------
>> From: "Nicolas Cellier" <[hidden email]>
>> Sent: Sunday, July 25, 2010 4:09 AM
>> To: "The general-purpose Squeak developers list"
>> <[hidden email]>
>> Subject: Re: [squeak-dev] TestRunner not assigning strings/remembering
>> old
>> values of temp vars.
>>
>>> There is no immutable support in Squeak.
>>> Smalltalk custom is to avoid overwriting literals like String Array...
>>> ...Same for instances of other classes like Point Rectangle which are
>>> likely to be shared by several objects.
>>> Most algorithms will rather modify and answer a copy.
>>> This default behaviour has a cost (creates lots of objects) but is
>>> fool proof (no danger to overwrite a shared information).
>>>
>>> For DES, I don't know, maybe it's just an optimization you must be aware
>>> of.
>>> In that case, you are responsible for creating the copy by yourself
>>> when required.
>>
>>
>> You mean before calling encryptBlock:/decryptBlock: with string data, we
>> should copy the string data.  That's unfortunate.  It works fine with
>> ByteArrays.
>>
>
> No, if you use a literal ByteArray with #[   ] notations, you'll get
> same symptoms as with a String literal.
>

>
> On 25.07.2010, at 04:22, Rob Withers wrote:
>
>> Thanks for pointing this out.  I am not happy about this at all.  It is
>> very non-intuitive, all due to an optimization that limits my
>> expressiveness.
>
> If string literals were immutable you would have to copy them yourself,
> too. So in any case you need to write
>
> plain := 'squeaker' copy
>
> if you intend to modify that string in-place.
>