Home » Archived » Visual Editor (VE) » Too complicated to be evaluated
Too complicated to be evaluated [message #15812] |
Tue, 20 January 2004 10:38  |
Eclipse User |
|
|
|
I have my own beans which have complex customizable properties.
I have screens built from these beans which were created using JBuilders
visual
GUI editor.
When attempting to view these in the customizer for a bean in one of these
beans, I get the following error message logged
(stack trace is attached).
"new ColumnsDataSpec(
new DataSpec[]{
new DataSpec("Arguments", "Access Method Table",
new String[]{"getSelector","putSelector","type","forceReadOnly ",}, new
String[]{"Fixed","Fixed","Fixed","Fixed",}, new
String[]{"displayString","","","true",}),})"
is too complicated to be evaluated.
|
|
|
Re: Too complicated to be evaluated [message #15853 is a reply to message #15812] |
Tue, 20 January 2004 12:01   |
Eclipse User |
|
|
|
Originally posted by: richkulp.NO.SPAM.us.ibm.com
Darren Hurt wrote:
>
> "new ColumnsDataSpec(
> new DataSpec[]{
> new DataSpec("Arguments", "Access Method Table",
> new String[]{"getSelector","putSelector","type","forceReadOnly ",}, new
> String[]{"Fixed","Fixed","Fixed","Fixed",}, new
> String[]{"displayString","","","true",}),})"
>
>
> is too complicated to be evaluated.
Currently we don't use a complete syntax parser for inspecting the
initialization string (we are working on that at this very moment
actually). Until then we have a simplified parser.
However, that said, your string doesn't look too complicated, except for
maybe the ColumnsDataSpec and DataSpec are not fully-qualified. We need
them to be fully-qualified for now. Try putting the full package name on
the classes and see if that helps (it shouldn't be necessary for
String, we recognize that as a special case).
Though with all of the nesting of constructors and array
initializations, it may actually be too complex at the moment.
--
Thanks, Rich Kulp
|
|
|
Re: Too complicated to be evaluated [message #15922 is a reply to message #15853] |
Wed, 21 January 2004 04:24   |
Eclipse User |
|
|
|
Actually, I only stripped out the package prefixes to make it more
readable to you guys, they are there in the actual initialization strings.
We have many cases where the complexity of the expression is
comparable to this. However, if you are currently working on this problem,
then that is great, and I look forward to the next release.
One small point:
The extra commas at the end of each array initializer, before the end
bracket
,} are not in the java initialization string, so I guess your code is
inserting them.
I know it makes no difference to the validity of the code, but maybe this
confuses
your parser? I'm probably talking rubbish, but it is just a thought.
Darren Hurt
"Rich Kulp" <richkulp@NO.SPAM.us.ibm.com> wrote in message
news:bujmpj$4oq$1@eclipse.org...
>
>
> Darren Hurt wrote:
>
> >
> > "new ColumnsDataSpec(
> > new DataSpec[]{
> > new DataSpec("Arguments", "Access Method
Table",
> > new String[]{"getSelector","putSelector","type","forceReadOnly ",}, new
> > String[]{"Fixed","Fixed","Fixed","Fixed",}, new
> > String[]{"displayString","","","true",}),})"
> >
> >
> > is too complicated to be evaluated.
>
> Currently we don't use a complete syntax parser for inspecting the
> initialization string (we are working on that at this very moment
> actually). Until then we have a simplified parser.
>
> However, that said, your string doesn't look too complicated, except for
> maybe the ColumnsDataSpec and DataSpec are not fully-qualified. We need
> them to be fully-qualified for now. Try putting the full package name on
> the classes and see if that helps (it shouldn't be necessary for
> String, we recognize that as a special case).
>
> Though with all of the nesting of constructors and array
> initializations, it may actually be too complex at the moment.
>
> --
> Thanks, Rich Kulp
>
>
|
|
| | |
Re: Too complicated to be evaluated [message #16567 is a reply to message #16536] |
Fri, 23 January 2004 11:41  |
Eclipse User |
|
|
|
Originally posted by: richkulp.NO.SPAM.us.ibm.com
It will be in a stable 1.0.0 drop soon, but the actual "release" of
1.0.0 will not be for a while. So if you are doing production level
code, you probably should not be relying on the stable or even milestone
builds because those will still have bugs and missing features. Maybe
you should wait for the release, date of which is still open but should
be early 2nd Half of 2004.
--
Thanks, Rich Kulp
|
|
|
Re: Too complicated to be evaluated [message #578696 is a reply to message #15812] |
Tue, 20 January 2004 12:01  |
Eclipse User |
|
|
|
Originally posted by: richkulp.NO.SPAM.us.ibm.com
Darren Hurt wrote:
>
> "new ColumnsDataSpec(
> new DataSpec[]{
> new DataSpec("Arguments", "Access Method Table",
> new String[]{"getSelector","putSelector","type","forceReadOnly ",}, new
> String[]{"Fixed","Fixed","Fixed","Fixed",}, new
> String[]{"displayString","","","true",}),})"
>
>
> is too complicated to be evaluated.
Currently we don't use a complete syntax parser for inspecting the
initialization string (we are working on that at this very moment
actually). Until then we have a simplified parser.
However, that said, your string doesn't look too complicated, except for
maybe the ColumnsDataSpec and DataSpec are not fully-qualified. We need
them to be fully-qualified for now. Try putting the full package name on
the classes and see if that helps (it shouldn't be necessary for
String, we recognize that as a special case).
Though with all of the nesting of constructors and array
initializations, it may actually be too complex at the moment.
--
Thanks, Rich Kulp
|
|
|
Re: Too complicated to be evaluated [message #578751 is a reply to message #15853] |
Wed, 21 January 2004 04:24  |
Eclipse User |
|
|
|
Actually, I only stripped out the package prefixes to make it more
readable to you guys, they are there in the actual initialization strings.
We have many cases where the complexity of the expression is
comparable to this. However, if you are currently working on this problem,
then that is great, and I look forward to the next release.
One small point:
The extra commas at the end of each array initializer, before the end
bracket
,} are not in the java initialization string, so I guess your code is
inserting them.
I know it makes no difference to the validity of the code, but maybe this
confuses
your parser? I'm probably talking rubbish, but it is just a thought.
Darren Hurt
"Rich Kulp" <richkulp@NO.SPAM.us.ibm.com> wrote in message
news:bujmpj$4oq$1@eclipse.org...
>
>
> Darren Hurt wrote:
>
> >
> > "new ColumnsDataSpec(
> > new DataSpec[]{
> > new DataSpec("Arguments", "Access Method
Table",
> > new String[]{"getSelector","putSelector","type","forceReadOnly ",}, new
> > String[]{"Fixed","Fixed","Fixed","Fixed",}, new
> > String[]{"displayString","","","true",}),})"
> >
> >
> > is too complicated to be evaluated.
>
> Currently we don't use a complete syntax parser for inspecting the
> initialization string (we are working on that at this very moment
> actually). Until then we have a simplified parser.
>
> However, that said, your string doesn't look too complicated, except for
> maybe the ColumnsDataSpec and DataSpec are not fully-qualified. We need
> them to be fully-qualified for now. Try putting the full package name on
> the classes and see if that helps (it shouldn't be necessary for
> String, we recognize that as a special case).
>
> Though with all of the nesting of constructors and array
> initializations, it may actually be too complex at the moment.
>
> --
> Thanks, Rich Kulp
>
>
|
|
|
Re: Too complicated to be evaluated [message #578850 is a reply to message #15922] |
Wed, 21 January 2004 09:59  |
Eclipse User |
|
|
|
Originally posted by: richkulp.NO.SPAM.us.ibm.com
About the extra commas. Good question. They are actually added
internally by the JDT's parser. But we should be able to handle it. I'll
take a look and if we can't, then I'll ask you to open a bugzilla about
that.
Thanks,
Rich
--
Thanks, Rich Kulp
|
|
|
Re: Too complicated to be evaluated [message #579205 is a reply to message #15991] |
Fri, 23 January 2004 06:01  |
Eclipse User |
|
|
|
On the general subject of the work that is being done on expression
evaluation (i've encountered other problems when my expressions refer
to instance variables etc), do you have any idea when a release with a
more comprehensive mechanism will be available?
We are hoping to be able to drop JBuilder in favour of the VE plug-in as
a tool for building screens based on our Java beans. However, the defective
expression evaluation mechanism seems to be the main barrier stopping us
from doing this.
"Rich Kulp" <richkulp@NO.SPAM.us.ibm.com> wrote in message
news:bum416$ss7$1@eclipse.org...
> About the extra commas. Good question. They are actually added
> internally by the JDT's parser. But we should be able to handle it. I'll
> take a look and if we can't, then I'll ask you to open a bugzilla about
> that.
>
> Thanks,
> Rich
>
>
> --
> Thanks, Rich Kulp
>
>
|
|
|
Re: Too complicated to be evaluated [message #579246 is a reply to message #16536] |
Fri, 23 January 2004 11:41  |
Eclipse User |
|
|
|
Originally posted by: richkulp.NO.SPAM.us.ibm.com
It will be in a stable 1.0.0 drop soon, but the actual "release" of
1.0.0 will not be for a while. So if you are doing production level
code, you probably should not be relying on the stable or even milestone
builds because those will still have bugs and missing features. Maybe
you should wait for the release, date of which is still open but should
be early 2nd Half of 2004.
--
Thanks, Rich Kulp
|
|
|
Goto Forum:
Current Time: Sat Jul 12 23:52:52 EDT 2025
Powered by FUDForum. Page generated in 0.08417 seconds
|