[squeak-dev] Rewriting 'ReadStream on: smth'

Hi,

I'm interested in enhancing the code of Pharo (and maybe Squeak if the
community is interested).
Currently, I'm thinking about replacing 'ReadStream on: expr' by 'expr
readStream'.

I can see the following advantages of this change:

- no more direct reference on ReadStream where it is not necessary
- allows different implementation of #readStream depending on the receiver class
- clarify the code

Can you detect any problem with this approach?

The RefactoringBrowser allows this change to be done simply in all the
method of the image.

--
Damien Cassou
Peter von der Ahé: «I'm beginning to see why Gilad wished us good
luck». (http://blogs.sun.com/ahe/entry/override_snafu)

Damien Cassou wrote:
>
> Can you detect any problem with this approach?
>

That transformation should be safe. I can think of some situations that
would cause problems (for example an external client which implements
#readStream incorrectly) but broken code is broken code so I wouldn't
worry too much about it.

Cheers,
   - Andreas

2008/6/4 Damien Cassou <[hidden email]>:
Yeah, but it also pretty prints the code.

Cheers
Philippe

On Wed, Jun 4, 2008 at 6:30 PM, Philippe Marschall
<[hidden email]> wrote:
Another advantage (thank you Philippe):

- the source code is prettier :-D

(the only problem is with the strange meaning RB assigns to 'pretty' :-)

--
Damien Cassou
Peter von der Ahé: «I'm beginning to see why Gilad wished us good
luck». (http://blogs.sun.com/ahe/entry/override_snafu)

On Wed, 4 Jun 2008 17:54:51 +0200, "Damien Cassou"
<[hidden email]> wrote:

> Currently, I'm thinking about replacing 'ReadStream on: expr' by 'expr
> readStream'.

As far as I can tell, this method has been around in SequenceableCollection
since sometime in 2000, at least in "normal" Squeak...

I've got a virgin Squeak 3.1 image that has it...

Later,
Jon

--------------------------------------------------------------
   Jon Hylands      [hidden email]      http://www.huv.com/jon

  Project: Micro Raptor (Small Biped Velociraptor Robot)
           http://www.huv.com/blog

On Wed, Jun 4, 2008 at 6:47 PM, Jon Hylands <[hidden email]> wrote:
What do you imply?

I don't want to remove ReadStream>>on:, only use another level of
indirection by using the already existing method #readStream. I think
this method has been made to avoid the use of ReadStream>>on: that's
why I believe this change is interesting.


--
Damien Cassou
Peter von der Ahé: «I'm beginning to see why Gilad wished us good
luck». (http://blogs.sun.com/ahe/entry/override_snafu)

Damien Cassou wrote:
>>> The RefactoringBrowser allows this change to be done simply in all the
>>> method of the image.
>> Yeah, but it also pretty prints the code.

Eeeek!

> Another advantage (thank you Philippe):
>
> - the source code is prettier :-D
>
> (the only problem is with the strange meaning RB assigns to 'pretty' :-)

Indeed. Don't do that. Reformatting code that ain't yours should be done
with great care if at all. Letting some so-called "pretty" printer do it
on a handful of methods at random is just evil.

Cheers,
   - Andreas

On Wed, 4 Jun 2008 18:52:37 +0200, "Damien Cassou"
<[hidden email]> wrote:

> What do you imply?
>
> I don't want to remove ReadStream>>on:, only use another level of
> indirection by using the already existing method #readStream. I think
> this method has been made to avoid the use of ReadStream>>on: that's
> why I believe this change is interesting.

Oh, I'm sorry - I was thinking you wanted to implement #readStream, but
you're saying you want to fix all senders of #on: to use #readStream...

Sorry, my misunderstanding...

Later,
Jon

--------------------------------------------------------------
   Jon Hylands      [hidden email]      http://www.huv.com/jon

  Project: Micro Raptor (Small Biped Velociraptor Robot)
           http://www.huv.com/blog

Damien Cassou wrote:
> Hi,
>
> I'm interested in enhancing the code of Pharo (and maybe Squeak if the
> community is interested).
> Currently, I'm thinking about replacing 'ReadStream on: expr' by 'expr
> readStream'.

And #writeStream of course.

> I can see the following advantages of this change:
>
> - no more direct reference on ReadStream where it is not necessary
> - allows different implementation of #readStream depending on the receiver class
> - clarify the code

+1.  In GNU Smalltalk I have added WriteStream>>#readStream and this not
only adds polymorphism, it makes code faster by removing the copy made
by #contents.

> Can you detect any problem with this approach?
>
> The RefactoringBrowser allows this change to be done simply in all the
> method of the image.

If you're willing to review the changes by hand, you can more or less
easily modify the Refactoring Browser's ParseTreeRewriter so that it
prettyprints only the replacement text, not everything.  This solution
would be pretty ad hoc; and as I said, I think it would require manual
review.  But in my opinion it is worth the effort for this kind of
refactoring.

Paolo