[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
|
Re: [imp-dev] API change in pdb.values Type
|
Hi Samantha,
Type is in "org.eclipse.imp.pdb.facts.type" which is located at "/org.eclipse.imp.pdb.values/src/org/eclipse/imp/pdb/facts/type/Type.java"
Code that is likely to be affected would be related to static analysis features or refactoring support for a particular language.
This API is used to represent, store and share facts about programs that may later be used to implement program analyses, refactorings, or visualizations. This is all part of the IMP analysis feature, which is a part of IMP outside of its core features for IDE run-time support and meta programming. Mainstream use of IMP's language features/extensions and wizards are not affected by the current API change.
The API change is targeted for the next pre-release of the IMP analysis feature. I have already committed the change.
Hope this helps! Otherwise, please don't hesitate to ask for more details.
Cheers,
Jurgen
On Thu, Jul 8, 2010 at 8:26 PM, Samantha Chan
<chanskw@xxxxxxxxxx> wrote:
Hi,
I cannot find the API described here.
Can you please let me know what package
Type belongs to? What version are you targeting this API change for?
and when do you plan on committing the change?
I would like to know more to assess
impact to our product. Thanks.
Thanks...
Samantha
Given that the change would help to avoid errors that
I could easily
see new PDB users committing unknowingly, and particularly errors that
can be hard to diagnose, I'd say the API change is warranted.
On Jul 7, 2010, at 9:31 AM, Jurgen Vinju wrote:
> Hi!
>
> I'm about to commit an API change, removing an argument from the
> method Type.instantiate(TypeStore store, Type type) in pdb.values.
>
> The store argument was a design flaw. A TypeStore should hold only
> declarations of (type parametrized) abstract data types, not
> instances, so instantiating a type with actual type parameters should
> not need a TypeStore from anywhere.
> Removing the argument obviously makes sure nobody accidentally passes
> in a TypeStore.
> TypeStores that accidentally hold type instances instead of type
> definitions eventually lead to subtle bugs, that's what I'm trying
to
> prevent here.
>
> I'm not sure this is clear for anybody that has not read or used the
> pdb.values code, but I wanted to put out this alert anyway. If you
> have this method in use, simply removing the parameter will fix your
> code.
>
> Cheers,
>
> Jurgen
Cheers,
- Bob
_______________________________________________
imp-dev mailing list
imp-dev@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/imp-dev
_______________________________________________
imp-dev mailing list
imp-dev@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/imp-dev
--
Jurgen Vinju
- Centrum Wiskunde & Informatica - SEN1
- INRIA Lille - ATEAMS
- Universiteit van Amsterdam
www:
http://jurgen.vinju.org,
http://www.rascal-mpl.nl,
http://twitter.com/jurgenvinju
skype: jurgen.vinju
