[ghc-steering-committee] Proposal: Or patterns #43 (Round 2)

[Simon Peyton Jones]

The Haskell report does some things, including typing rules, by giving the desugaring, and saying “the original typechecks if the deguared version does).  As it happens that approach is a very nice, simple way to say what or-patterns should typecheck; it deals with all the issues of existentials and locally-captured dictionaries (by ruling them out) at one blow.

So I’d be content with a deguaring rule for or-patterns, along the lines I gave, saying that this governs the typing too.  That should only take Omer a few minutes to add; indeed we could accept subject to that condition.

Simon

From: ghc-steering-committee <ghc-steering-committee-bounces at haskell.org> On Behalf Of Manuel M T Chakravarty
Sent: 19 June 2018 03:17
To: ghc-steering-committee at haskell.org
Subject: Re: [ghc-steering-committee] Proposal: Or patterns #43 (Round 2)

Let’s consolidate the discussion of Proposal #43 in this current thread about the revised proposal. And for reference, here is SimonPJ’s recent message posted in the old thread:


SimonPJ’s message got support from Iavor and Richard, and SimonM previously expressed support on this Round 2 thread.

In the meantime, the proposal has also been changed to use ’;’ as a separator (instead of the initially proposed ’|’). However, this still leaves us with

  https://github.com/ghc-proposals/ghc-proposals/pull/43#issuecomment-398086013<https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fghc-proposals%2Fghc-proposals%2Fpull%2F43%23issuecomment-398086013&data=02%7C01%7Csimonpj%40microsoft.com%7C745885965631491de68f08d5d58ace2a%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636649715262476300&sdata=%2FiUxySHtyiQnjcu0XpbJpINCOHpZqAraSqmlCYXyIi4%3D&reserved=0>

where the proposal author writes that he will update the proposal with new typing rules along SimonPJ’s suggestions within a few weeks.

Hence, it seems to me that we need to wait with a formal decision until that update of the typing rules has happened. Do you agree, Simon?

In any case, is anybody unhappy with any aspect other than ”what would and would not typecheck” (which will hopefully be addressed in those forthcoming typing rules)? If so, please speak now.

Cheers,
Manuel


01.06.2018 17:30 Manuel M T Chakravarty <chak at justtesting.org<mailto:chak at justtesting.org>>:

We now have a revised version of the ”Or pattern” proposal #43 to consider:

 (formatted) https://github.com/osa1/ghc-proposals/blob/or_patterns/proposals/0000-or-<https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fosa1%2Fghc-proposals%2Fblob%2For_patterns%2Fproposals%2F0000-or-&data=02%7C01%7Csimonpj%40microsoft.com%7C745885965631491de68f08d5d58ace2a%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636649715262486309&sdata=RjKikfeuhOIrd6n3czFjcXppEl%2BUY5OKCW6wAm7FXkw%3D&reserved=0>patterns.rst
 (PR thread) https://github.com/ghc-proposals/ghc-proposals/pull/43<https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fghc-proposals%2Fghc-proposals%2Fpull%2F43&data=02%7C01%7Csimonpj%40microsoft.com%7C745885965631491de68f08d5d58ace2a%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636649715262486309&sdata=RJvsbA%2BV5%2F3fl4JCo0JPLRuLviJMRAKD5%2Fo59D5TBcI%3D&reserved=0>

You may remember that we discuss this a while back

  https://mail.haskell.org/pipermail/ghc-steering-committee/2017-November/000231.html

and finally decided to bounce it back to the authors with the following recommendation:

  https://github.com/ghc-proposals/ghc-proposals/pull/43#issuecomment-358189327<https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fghc-proposals%2Fghc-proposals%2Fpull%2F43%23issuecomment-358189327&data=02%7C01%7Csimonpj%40microsoft.com%7C745885965631491de68f08d5d58ace2a%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636649715262496314&sdata=Dnk0UiIiTlNksRLggy6inQKYshcq8qlV7132PZV6xi0%3D&reserved=0>

We previously decided that we are inclined to accept this proposal if the authors change the points that we highlighted. Judging from the conversation of SimonPJ and Richard with the author on the GitHub thread, it appears to me that most of the issues have been addressed. However, Simon’s point

• "Patterns that bind existentials, dictionaries, or equalities are rejected by the type checker". I disagree. Earlier I suggested "no variable bound by an or-pattern can have a type that mentions an existential variable bound by the or-pattern". That is much more friendly because you can still match against an existential constructor; you just can’t bind an existential variable.

from <https://github.com/ghc-proposals/ghc-proposals/pull/43#issuecomment-368933401<https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fghc-proposals%2Fghc-proposals%2Fpull%2F43%23issuecomment-368933401&data=02%7C01%7Csimonpj%40microsoft.com%7C745885965631491de68f08d5d58ace2a%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636649715262506323&sdata=wkaSDSkWycoydAgIlMYOZpEa3fjrQ%2Fax6Zp8iNbhuDY%3D&reserved=0>> still doesn’t seem to be addressed, or am I misunderstanding?

Overall, I propose to accept the proposal, possibly requiring that the above point be addressed.

What do you all think?

Cheers,
Manuel


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.haskell.org/pipermail/ghc-steering-committee/attachments/20180619/3d5a4863/attachment-0001.html>