Eclipse Community Forums: M2T (model-to-text transformation) » [Acceleo] proper signature of template

Home

Home » Modeling » M2T (model-to-text transformation) » [Acceleo] proper signature of template

[Acceleo] proper signature of template [message #1804893]

Tue, 02 April 2019 16:27

Eclipse User

Hi,
tell me, please what exactly means the following template call:

[xx('txt')/]

assuming that self is of type String.

xx(a1: String, a2: String) where a1=self (of the caller), a2='txt',
xx(a1: String) where a1='txt',
first of the above which is defined
nothing of the above, I completely misunderstand it?

The experiments show that the third option is true.
I'm confused because when none is defined, the OCL in Acceleo parser generate only one error mark with the signature (1) so only this version can be generated as quick fix. The mark is not (and it shouldn't be) generated if the second variant is defined as it would be more general case.

Tom

Re: [Acceleo] proper signature of template [message #1804899 is a reply to message #1804893]

Wed, 03 April 2019 01:14

Eclipse User

Hi

My recollection (sorry cannot find the Bugzilla) is that it is indeed confusing; the OMG MOFM2T specification is inadequate. Acceleo, as the only implementation of MOFM2T, had to pioneer a solution which should have been reflected in a MOFM2T 1.1. As such you may find some clues in more recent Acceleo documentation.

The principle is that a new helper X::Y(...) may be declared for use by templates as anX.Y(...) without needing to change the metamodel for X. Use of the new helper effectively requires arguments to be shifted along once.

Complexities arise from polymorphic overload and use of non-EObject datatypes that are not traceable.

If you want to use these facilities you are probably better off figuring it out yourself rather than relying on my recollection.

Regards

Ed Willink

Re: [Acceleo] proper signature of template [message #1804906 is a reply to message #1804899]

Wed, 03 April 2019 03:08

Eclipse User

Hi

Ed Willink wrote on Wed, 03 April 2019 01:14

My recollection (sorry cannot find the Bugzilla

It was https://www.eclipse.org/forums/index.php?t=msg&th=208238&goto=784647&#msg_784647

The Acceleo team chose to stick to the letter of the MOFM2T specification rather than pioneer a more OOP approach that might be more in keeping with the spirit of MOFM2T.

https://wiki.eclipse.org/Acceleo/FAQ#How_does_the_overriding_system_works_in_Acceleo_.3F provides better documentation.

Regards

Ed Willink

Previous Topic:	Does the Acceleo model-to-text transformation support different kind of input models in .mtl file?
Next Topic:	[xpand + xtend]: how to access MWE workflow properties in Xpand templates, with xtend 2.17.x?

Goto Forum:

-=] Back to Top [=-

Current Time: Mon Jun 23 18:56:27 EDT 2025

.:: Contact :: Home ::.

Breadcrumbs

Sign up to our Newsletter