[m-rev.] for review: insts for type constructors

[Zoltan Somogyi]

For review by anyone. The parts I am most interested in
opinions on are the two places marked with "XXX IFTC".
However, since noone will start using the (for now undocumented)
language feature being introduced in this diff, resolving those
XXXs shouldn't be needed for a commit.

This implements most of the first step outlined in my email
on sep 9. The syntax is the simpler variant of the proposal
we discussed and agreed on the couple of days before that,
since noone argued for the more complex variant.

Zoltan.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Log.iftc
Type: application/octet-stream
Size: 3418 bytes
Desc: not available
URL: <http://lists.mercurylang.org/archives/reviews/attachments/20150929/1e467d7e/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: DIFF.iftc
Type: application/octet-stream
Size: 62027 bytes
Desc: not available
URL: <http://lists.mercurylang.org/archives/reviews/attachments/20150929/1e467d7e/attachment-0001.obj>