Skip to content

local Cross Tree Constraints #3

Open
@gzoumix

Description

@gzoumix

I work on a industrial tool simulating fluid flow (e.g., air flow on an supersonic airplane).
For this tool, I developed the pydop library, and I would really like to be able to directly use the UVL as input syntax.
However, UVL faces a couple of mostly-cosmetics limitations that my physicists colleagues do no like.
The first one is the fact that the cross-tree constraints are defined globally at the end of the file.
This makes writing and reading the feature model less natural, since the information related to some features are not close by.
Syntactically, local ctc only requires adding an optional new "constraints" block within a feature, at the same level of groups:

features
    root
        optional
            f1
                optional
                    f2
                    f3
                constraints
                    f3 => f2
            f4
                ...

The drawback of such local ctc, as noted by Sebastian Krieter is that it requires some scope analysis to ensures that the features mentioned in a local ctc are declared in the local fm subtree. Then the semantics of such a ctc is simply "if the including feature is selected, then the ctc must be validated", which is quite intuitive, as demonstrated by my physicists colleagues who didn't have any prior knowledge of SPL before our collaboration.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions