|
|
Re: type is not exposed to the weaver??? [message #42026 is a reply to message #41995] |
Mon, 04 October 2004 15:34 |
Eclipse User |
|
|
|
Originally posted by: adrian_colyer.uk.ibm.com
Michael Moser wrote:
> Maybe I should add here, that one of the three classes listed is the
> Copyright-interface, which currently is still empty. Here I can
> understand, that the "weaver" has nothing to add to.
> But the other two classes are "normal" classes, that both contain
> quite a number of methods. One of them extends the other, but other
> than that, they don't appear to be in any sense special (no "final
> class ..." or "static class ..." or whatever).
> So these two really puzzle me, why they should not be "exposed to the
> weaver"...
> Michael
"Type not exposed to weaver" messages can occur for a number of reasons.
One common cause is having needed types on the *classpath*, but not on the
*inpath*. Only types compiled from source, or class files in directories
and jars placed on the inpath are available to the weaver for linking with
aspects. Classes simply on the classpath are unchanged by the weaving
process. A second common reason for this message is that one of the types
*referred to* by a class or aspect has not been provided (in general, the
AspectJ compiler likes a complete 'world' in order to do its work). We're
working on improving the 'type not exposed' messages to give more
accurative and informative messages in such situations...
|
|
|
|
Re: type is not exposed to the weaver??? [message #583753 is a reply to message #41995] |
Mon, 04 October 2004 15:34 |
Adrian Colyer Messages: 61 Registered: July 2009 |
Member |
|
|
Michael Moser wrote:
> Maybe I should add here, that one of the three classes listed is the
> Copyright-interface, which currently is still empty. Here I can
> understand, that the "weaver" has nothing to add to.
> But the other two classes are "normal" classes, that both contain
> quite a number of methods. One of them extends the other, but other
> than that, they don't appear to be in any sense special (no "final
> class ..." or "static class ..." or whatever).
> So these two really puzzle me, why they should not be "exposed to the
> weaver"...
> Michael
"Type not exposed to weaver" messages can occur for a number of reasons.
One common cause is having needed types on the *classpath*, but not on the
*inpath*. Only types compiled from source, or class files in directories
and jars placed on the inpath are available to the weaver for linking with
aspects. Classes simply on the classpath are unchanged by the weaving
process. A second common reason for this message is that one of the types
*referred to* by a class or aspect has not been provided (in general, the
AspectJ compiler likes a complete 'world' in order to do its work). We're
working on improving the 'type not exposed' messages to give more
accurative and informative messages in such situations...
|
|
|
Powered by
FUDForum. Page generated in 0.03739 seconds