[Pharo-project] compiler fixing _ -> :=
marcus math could'nt you fix the compiler to read := _ and generate := ? Stef
On 08.07.2008, at 09:19, Stéphane Ducasse wrote:
marcus math
could'nt you fix the compiler to read := _ and generate := ?
The newcompiler accepts _ and :=. The _ should be removed (but a preference might be good to be able to load old code). I will put that on the issue tracker. (this is here: http://code.google.com/p/squeaknewcompiler/issues/list) What do you mean with "generate"? When pretty-printing? -- Marcus Denker -- denker@iam.unibe.ch http://www.iam.unibe.ch/~denker
marcus math
could'nt you fix the compiler to read := _ and generate := ?
The newcompiler accepts _ and :=. The _ should be removed (but a preference might be good to be able to load old code).
I will put that on the issue tracker.
(this is here: http://code.google.com/p/squeaknewcompiler/issues/list)
What do you mean with "generate"? When pretty-printing?
The idea was can be absorb _ (and avoid to have to manually patch the code - even by invoking FixUnderscore) I mean read in _ or := but changing the text to have only have := so may be pretty printing is required. May be what I ask is too complex and we could solve that by invoking systematically FixUnderscore Stef
On 08.07.2008, at 09:54, Stéphane Ducasse wrote:
marcus math
could'nt you fix the compiler to read := _ and generate := ?
The newcompiler accepts _ and :=. The _ should be removed (but a preference might be good to be able to load old code).
I will put that on the issue tracker.
(this is here: http://code.google.com/p/squeaknewcompiler/issues/ list)
What do you mean with "generate"? When pretty-printing?
The idea was can be absorb _ (and avoid to have to manually patch the code - even by invoking FixUnderscore)
I mean read in _ or := but changing the text to have only have := so may be pretty printing is required. May be what I ask is too complex and we could solve that by invoking systematically FixUnderscore
Hmmm... I would vote against magic. Just changing the test in the image (the changes file) is not enough: then people commit this, then there is a class-side initialize among those methods that's broken ---> unloadable package. I would say as soon as there is a process that incorporated small fixes fast, it will be easy to just step-by-step fix all cases of _ Marcus -- Marcus Denker -- denker@iam.unibe.ch http://www.iam.unibe.ch/~denker
Hmmm... I would vote against magic. Just changing the test in the image (the changes file) is not enough: then people commit this, then there is a class-side initialize among those methods that's broken ---> unloadable package.
I would say as soon as there is a process that incorporated small fixes fast, it will be easy to just step-by-step fix all cases of _
No dream marcus. There is one guy merging and applying stuff. So what I can do is to add a fixunderscore call automatically to dirty packages. Stef
participants (2)
-
Marcus Denker -
Stéphane Ducasse