|
Re: Modelling XBRL Taxonomy [message #636498 is a reply to message #636468] |
Mon, 01 November 2010 15:11 |
Ed Merks Messages: 33141 Registered: July 2009 |
Senior Member |
|
|
Paul,
Comments below.
Paul Fraser wrote:
> I am not sure if I am expecting too much of the EMF, but in attempting
> to create an EMF project based on an xbrl taxonomy I am experiencing
> some errors.
> The xsd files are available
> http://xbrl.iasb.org/repository/2010-04-30/ifrs_20100430.zip and the
> file I am playing with initially is "ifrs-cor_2010-04-30.xsd".
> The ecore model seems to be produced ok but when trying to create the
> edit code I have problems.
> When attempting on a Ubuntu machine I get errors similar to:
> "DocumentationTypeItemProvider cannot be resolved to a type"
Is that class missing? (You can find it with Ctrl-Shift-T.)
> On a windows vista machine I get error similar to:
> "...code of method collectNewChildDescriptors.....is exceeding the
> 65535 bytes limit"
It sounds like you have an incredibly massive model; it's not atypical
that schema designers produce unwieldy things. Even if this works (the
method body wasn't too big), the menu for the set of choices will exceed
the human capacity for choosing from a long menu. You can manually
split it into multiple methods to work around it, but as I said, the UI
will be unwieldy.
> I suppose the question should be, is EMF suitable for this type of
> project?
Yes, but Java has limitations, such as method body size. But worse
still, human minds have limitations as well, which suggests that a menu
with hundreds (or perhaps thousands) of child creation choices won't be
suitable.
>
> Regards
> Paul Fraser
Ed Merks
Professional Support: https://www.macromodeling.com/
|
|
|
Powered by
FUDForum. Page generated in 0.04472 seconds