KerML 1.0b4 FTF Avatar
  1. OMG Issue

KERML_ — InvocationExpression::result specialization

  • Key: KERML_-117
  • Status: closed  
  • Source: Dassault Systemes ( Mr. Tomas Juknevicius)
  • Summary:

    Pilot implementation creates implied Subsetting for SelectExpression::result, but there is nothing in the spec for this situation. There is no information in semantic chapter “8.3.4.8.15 SelectExpression”, no formal constraint. Is that a miss in the spec?

  • Reported: KerML 1.0b2 — Mon, 28 Oct 2024 07:39 GMT
  • Disposition: Resolved — KerML 1.0b4
  • Disposition Summary:

    Add appropriate semantic constraints

    • Add a semantic constraint for SelectExpression capturing the expectation that the type of the result should be the same as the type of the result of the first argument expression (since the result values are necessarily a subset of the argument values), as suggested in the issue.
    • Also add a new IndexExpression metaclass with a similar semantic constraint for index expressions using the using the operator #. Even though not mentioned in th issue, it is also expected that the type of the result of of an index expression is the same as the type of the result its first argument, at least if the result of the first argument is not an array (because, for an array, the index function '#' is specialized to return an element of the array, which will general be of a different type than the array itself).
  • Updated: Sat, 19 Jul 2025 18:59 GMT
  • Attachments: