[squeak-dev] The Trunk: Collections-ar.123.mcz

Andreas Raab uploaded a new version of Collections to project The Trunk:
http://source.squeak.org/trunk/Collections-ar.123.mcz

==================== Summary ====================

Name: Collections-ar.123
Author: ar
Time: 1 September 2009, 12:38:06 pm
UUID: e7432274-e18a-1e43-a002-f3ab261bd465
Ancestors: Collections-tfel.122

Some fixes for RWBinaryOrTextStream which was too agressive optimizing some paths. Also fixes Character>>codePoint: which has no reason to raise an exception for values > 256.

=============== Diff against Collections-tfel.122 ===============

Item was added:
+ ----- Method: RWBinaryOrTextStream>>upTo: (in category 'accessing') -----
+ upTo: anObject
+ "Answer a subcollection from the current access position to the
+ occurrence (if any, but not inclusive) of anObject in the receiver. If
+ anObject is not in the collection, answer the entire rest of the receiver."
+ | newStream element species |
+ species := isBinary ifTrue:[ByteArray] ifFalse:[String].
+ newStream := WriteStream on: (species new: 100).
+ [self atEnd or: [(element := self next) = anObject]]
+ whileFalse: [newStream nextPut: element].
+ ^newStream contents!

Item was changed:
  ----- Method: RWBinaryOrTextStream>>next:putAll:startingAt: (in category 'writing') -----
  next: anInteger putAll: aCollection startingAt: startIndex
+ "Optimized for ByteArrays"
+ aCollection class == ByteArray
+ ifTrue:[^super next: anInteger putAll: aCollection asString startingAt: startIndex].
+ ^super next: anInteger putAll: aCollection startingAt: startIndex!
- ^super next: anInteger putAll: aCollection asString startingAt: startIndex!

Item was changed:
  ----- Method: RWBinaryOrTextStream>>nextPutAll: (in category 'writing') -----
  nextPutAll: aCollection
+ "Optimized for ByteArrays"
+ aCollection class == ByteArray
+ ifTrue:[^super nextPutAll: aCollection asString].
+ ^super nextPutAll: aCollection!
- ^super nextPutAll: aCollection asString!

Item was changed:
  ----- Method: Character class>>codePoint: (in category 'instance creation') -----
  codePoint: integer
+ "Return a character whose encoding value is integer.
+ For ansi compability."
+ ^self value: integer!
- "Return a character whose encoding value is integer."
- #Fundmntl.
- (0 > integer or: [255 < integer])
- ifTrue: [self error: 'parameter out of range 0..255'].
- ^ CharacterTable at: integer + 1!

2009/9/1  <[hidden email]>:
Don't know about the reason for such a limitation by just reading the diff...
However, there was some insurance that a Character of value < 256
would be unique.
Doesn't this matter?

Nicolas

On 01.09.2009, at 11:37, Nicolas Cellier wrote:

#value: does this.

- Bert -

2009/9/1 Bert Freudenberg <[hidden email]>:
Does it mean #codePoint: is a low level (basic) creation message,
private and not to be used except in special circumstances like
creating the CharacterTable, while #value is the public API?

Or does it mean any code using the ANSI protocol #codePoint: wont'
share the uniqueness property of #value:?

The category and comment made me think it was the latter, thus my question.

Nicolas

On 01.09.2009, at 13:51, Nicolas Cellier wrote:

I don't understand. #codePoint: uses #value: so there is no difference.

- Bert -

Nicolas Cellier wrote:
> Does it mean #codePoint: is a low level (basic) creation message,
> private and not to be used except in special circumstances like
> creating the CharacterTable, while #value is the public API?
>
> Or does it mean any code using the ANSI protocol #codePoint: wont'
> share the uniqueness property of #value:?

Neither. #codePoint: is simply the ansi name for #value:. It shares all
the properties and just hadn't been updated with our switch to  Unicode.

Cheers,
   - Andreas

2009/9/1 Bert Freudenberg <[hidden email]>:
Thanks for your patience, I just realized it...

Nicoals