Skip to Content.
Sympa Menu

coq-club - Re: [Coq-Club] Missing a detail on type checking -- bad error reporting?

coq-club AT inria.fr

Subject: The Coq mailing list

List archive

Re: [Coq-Club] Missing a detail on type checking -- bad error reporting?


Chronological Thread 
  • From: Pierre Courtieu <pierre.courtieu AT gmail.com>
  • To: Coq Club <coq-club AT inria.fr>
  • Cc: Kevin Sullivan <sullivan.kevinj AT gmail.com>
  • Subject: Re: [Coq-Club] Missing a detail on type checking -- bad error reporting?
  • Date: Tue, 28 Nov 2017 11:16:54 +0100
  • Authentication-results: mail2-smtp-roc.national.inria.fr; spf=None smtp.pra=pierre.courtieu AT gmail.com; spf=Pass smtp.mailfrom=pierre.courtieu AT gmail.com; spf=None smtp.helo=postmaster AT mail-ot0-f174.google.com
  • Ironport-phdr: 9a23:VnMsoBX/Pgh4GdwYXe0d1zsdtXzV8LGtZVwlr6E/grcLSJyIuqrYZR2At8tkgFKBZ4jH8fUM07OQ6P+wHzFYqb+681k8M7V0HycfjssXmwFySOWkMmbcaMDQUiohAc5ZX0Vk9XzoeWJcGcL5ekGA6ibqtW1aMlzFOAF0PuX4HJLJx4Tyjrjqus6bXwIdrz2kKZh2MR/++Q7Wr4wdhZZoAqc30BrA5HVSLbd432RtcGqSkgzm64+b+4N57yVdprp1789NS7/3Oa8/UKZEDTk7G28w7czv8xLESF3ctTMnTmwKn08QUED+5xbgU8K063Oiuw==

Note that Clément Pit-claudel implemented something to help in these
cases in his company-coq mode for emacs/proofgeneral: it uses diff
facility on error messages to show what differs in two big types.

See the tutorial M-x company-coq-tutorial.

Best,
Pierre


2017-11-28 10:33 GMT+01:00 Ralf Jung
<jung AT mpi-sws.org>:
> Hi,
>
>> The term
>> "{|
>> field_final := isFinal;
>> field_value := value;
>> field_rwt_ok := pf |}"
>> has type
>> "Field
>> mTipe_string
>> "io.compression.codecs"
>> io_compression_codecs_rwt"
>> while it is expected to have type
>> "Field
>> mTipe_string
>> "io_compression_codecs"
>> io_compression_codecs_rwt".
>
> Would it be possible to improve that error message? For example, when
> similar errors are shown in Rust, it shows both the two overall types
> that do not match, but also the concrete place in both of them that
> fails to match -- so here, it would additionally show that it has type
> "io.compression.codecs" while it is expected to have
> "io_compression_codecs". This is really useful, even more so if the
> only difference is in implicit arguments that are otherwise not even
> printed.
>
> ; Ralf



Archive powered by MHonArc 2.6.18.

Top of Page