You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
As of now, GIPSL implicitly multiplies the mapping's binary ILP variable to the terms given in, for example, a constraint with the mapping context. Example:
mapping m with someGtRule;
constraint -> mapping::m {
<some constraint>
}
The problem arises that because of this, no other mappings can be accessed in a constraint with a mapping context because this would result in a quadratic equation. Hence, the following example is currently not possible:
mapping n with someOtherGtRule;
constraint -> mapping::m {
<some constraint on self> <=> mappings.n->[...]
}
The current workaround is to use the pattern/rule of the mapping as context and overlap it completely within the filter expression as such:
This results in a very "wordy" specification because a large amount of filter specification code is just used to get the one mapping that corresponds to the match of the pattern context.
As a result, this issue proposes to remove the implicit variable multiplication for the mapping context.
I will open another issue for another solution to the problem of "wordiness", but the implicit multiplication should nevertheless be remove IMO.
The text was updated successfully, but these errors were encountered:
As of now, GIPSL implicitly multiplies the mapping's binary ILP variable to the terms given in, for example, a constraint with the mapping context. Example:
The problem arises that because of this, no other mappings can be accessed in a constraint with a mapping context because this would result in a quadratic equation. Hence, the following example is currently not possible:
The current workaround is to use the pattern/rule of the mapping as context and overlap it completely within the filter expression as such:
This results in a very "wordy" specification because a large amount of filter specification code is just used to get the one mapping that corresponds to the match of the pattern context.
As a result, this issue proposes to remove the implicit variable multiplication for the mapping context.
I will open another issue for another solution to the problem of "wordiness", but the implicit multiplication should nevertheless be remove IMO.
The text was updated successfully, but these errors were encountered: