Home » Modeling » Epsilon » Error message displayed when using Ant, but not with launch configuration
| Error message displayed when using Ant, but not with launch configuration [message #6872] |
Thu, 11 December 2008 10:43  |
Eclipse User |
|
|
|
Hia there,
I was just wondering whether the following behaviour I get is due to my
configuration/me being new to Epsilon or something:
I have an ETL script that had the following line:
var coll : Collection := Set.createInstance(); -- #1
When I ran the translation via a launch script, I got no error in the
Epsilon console. But when I moved to running it via an Ant build file, I
got the following error message (although, the translation succeeds to the
end):
------------------------
java.lang.InstantiationException
at
sun.reflect.InstantiationExceptionConstructorAccessorImpl.ne wInstance(InstantiationExceptionConstructorAccessorImpl.java :30)
at java.lang.reflect.Constructor.newInstance(Constructor.java:5 13)
at java.lang.Class.newInstance0(Class.java:355)
at java.lang.Class.newInstance(Class.java:308)
at
org.eclipse.epsilon.eol.types.EolCollectionType.createInstan ce(EolCollectionType.java:43)
at
org.eclipse.epsilon.eol.execute.VarStatementExecutor.execute (VarStatementExecutor.java:47)
at
org.eclipse.epsilon.eol.execute.ExecutorFactory.executeAST(E xecutorFactory.java:184)
...
...
------------------------
If I change the statement to:
var coll : Set := Set.createInstance(); -- #2
.. then running via either the launch configuration or Ant file, the error
does not get thrown (and again, the translation runs fine to the end).
The problem isn't a major one as it's easy to get round, but I was just
curious as to whether statements like #1 are permitted.
Also, I tend to develop an etl translation in combination with a launch
configuration, to begin with, and so wasn't aware of the problem until it
was run via Ant.
I may be doing something really silly here, so apologies in advance :)
Edd
|
|
|
| Re: Error message displayed when using Ant, but not with launch configuration [message #6894 is a reply to message #6872] |
Thu, 11 December 2008 11:04  |
| Eclipse User |
|
|
|
Hi Edd,
var coll : Set also creates a new Set (EOL does this implicitly for
primitive types and collection types) so the Set.createInstance() part
is not really necessary. I'll have a look into the cause of this
inconsistent behaviour and get back to you shortly.
Cheers,
Dimitris
Edward Turner wrote:
> Hia there,
>
> I was just wondering whether the following behaviour I get is due to my
> configuration/me being new to Epsilon or something:
>
> I have an ETL script that had the following line:
>
> var coll : Collection := Set.createInstance(); -- #1
>
> When I ran the translation via a launch script, I got no error in the
> Epsilon console. But when I moved to running it via an Ant build file, I
> got the following error message (although, the translation succeeds to
> the end):
>
> ------------------------
> java.lang.InstantiationException
> at
> sun.reflect.InstantiationExceptionConstructorAccessorImpl.ne wInstance(InstantiationExceptionConstructorAccessorImpl.java :30)
>
> at java.lang.reflect.Constructor.newInstance(Constructor.java:5 13)
> at java.lang.Class.newInstance0(Class.java:355)
> at java.lang.Class.newInstance(Class.java:308)
> at
> org.eclipse.epsilon.eol.types.EolCollectionType.createInstan ce(EolCollectionType.java:43)
>
> at
> org.eclipse.epsilon.eol.execute.VarStatementExecutor.execute (VarStatementExecutor.java:47)
>
> at
> org.eclipse.epsilon.eol.execute.ExecutorFactory.executeAST(E xecutorFactory.java:184)
>
> ..
> ..
> ------------------------
>
> If I change the statement to:
>
> var coll : Set := Set.createInstance(); -- #2
>
> . then running via either the launch configuration or Ant file, the
> error does not get thrown (and again, the translation runs fine to the
> end).
>
> The problem isn't a major one as it's easy to get round, but I was just
> curious as to whether statements like #1 are permitted.
>
> Also, I tend to develop an etl translation in combination with a launch
> configuration, to begin with, and so wasn't aware of the problem until
> it was run via Ant.
>
> I may be doing something really silly here, so apologies in advance :)
>
> Edd
>
|
|
|
| Re: Error message displayed when using Ant, but not with launch configuration [message #562978 is a reply to message #6872] |
Thu, 11 December 2008 11:04 |
| Eclipse User |
|
|
|
Hi Edd,
var coll : Set also creates a new Set (EOL does this implicitly for
primitive types and collection types) so the Set.createInstance() part
is not really necessary. I'll have a look into the cause of this
inconsistent behaviour and get back to you shortly.
Cheers,
Dimitris
Edward Turner wrote:
> Hia there,
>
> I was just wondering whether the following behaviour I get is due to my
> configuration/me being new to Epsilon or something:
>
> I have an ETL script that had the following line:
>
> var coll : Collection := Set.createInstance(); -- #1
>
> When I ran the translation via a launch script, I got no error in the
> Epsilon console. But when I moved to running it via an Ant build file, I
> got the following error message (although, the translation succeeds to
> the end):
>
> ------------------------
> java.lang.InstantiationException
> at
> sun.reflect.InstantiationExceptionConstructorAccessorImpl.ne wInstance(InstantiationExceptionConstructorAccessorImpl.java :30)
>
> at java.lang.reflect.Constructor.newInstance(Constructor.java:5 13)
> at java.lang.Class.newInstance0(Class.java:355)
> at java.lang.Class.newInstance(Class.java:308)
> at
> org.eclipse.epsilon.eol.types.EolCollectionType.createInstan ce(EolCollectionType.java:43)
>
> at
> org.eclipse.epsilon.eol.execute.VarStatementExecutor.execute (VarStatementExecutor.java:47)
>
> at
> org.eclipse.epsilon.eol.execute.ExecutorFactory.executeAST(E xecutorFactory.java:184)
>
> ..
> ..
> ------------------------
>
> If I change the statement to:
>
> var coll : Set := Set.createInstance(); -- #2
>
> . then running via either the launch configuration or Ant file, the
> error does not get thrown (and again, the translation runs fine to the
> end).
>
> The problem isn't a major one as it's easy to get round, but I was just
> curious as to whether statements like #1 are permitted.
>
> Also, I tend to develop an etl translation in combination with a launch
> configuration, to begin with, and so wasn't aware of the problem until
> it was run via Ant.
>
> I may be doing something really silly here, so apologies in advance :)
>
> Edd
>
|
|
|
Goto Forum:
Current Time: Thu Oct 30 22:47:07 EDT 2025
Powered by FUDForum. Page generated in 0.03090 seconds |
Home