[Issue 6278] Regression(2.054 beta): 'in' contract inheritance doesn't work with safe code

[d-bugmail at puremagic.com]

http://d.puremagic.com/issues/show_bug.cgi?id=6278


Walter Bright <bugzilla at digitalmars.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |bugzilla at digitalmars.com


--- Comment #8 from Walter Bright <bugzilla at digitalmars.com> 2012-02-01 00:03:28 PST ---
The concept here is that, if a base function and its override both have an IN
contract, then only one of them needs to succeed. This is done by generating:

  void derived.in()
  {
    try
    {
      base.in();
    }
    catch ()
    {
      ... body of derived.in() ...
    }
  }

So if base.in() doesn't throw, derived.in() need not be executed, and the
contract is valid. If base.in() throws, then derived.in()'s body is executed.

As you can see, it swallows any exception generated by the contract. Whether
this is correct or not is certainly debatable.

I think a reasonable solution which will target this particular thing would be
to mark that try/catch as being valid even in safe mode. That'll change it from
a regression to an enhancement request to figure out a better way, as at least
it'll work as documented.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------