From: Florian Haftmann <florian.haftmann@informatik.tu-muenchen.de>
Dear all,
I stumbled over a syntax issue with Bit-Not.
("NOT _" [70] 71)
("~~ _" [70] 71)
This non-regular syntax leads to ambiguities:
term ‹NOT (2 * a)› \<comment>‹Prints as NOT 2 * a›
term ‹(NOT 2) * a› \<comment>‹Prints as NOT 2 * a›
term ‹NOT 2 * a› \<comment>‹Does not parse›
Since NOT is a regular prefix operation, it would be much better to
avoid any tinkering with priorities:
("NOT")
("~~")
The change would require some precedence clarification in the
distribution and AFP, but to me it seems worth the effort.
Cheers,
Florian
signature.asc
From: "Klein, Gerwin (Data61, Kensington NSW)" <Gerwin.Klein@data61.csiro.au>
I’ll probably regret saying this, but I agree, better to clarify this, even if it will be annoying to fix the fallout.
Cheers,
Gerwin
From: Florian Haftmann <florian.haftmann@informatik.tu-muenchen.de>
Hi Gerwin,
I’ll probably regret saying this, but I agree, better to clarify this, even if it will be annoying to fix the fallout.
the problem mainly manifests if NOT is applied to regular prefix operations:
NOT f x ~> NOT (f x)
~~ f x ~> ~~ (f x)
The examples I found so far are all for unary f. Hence there is a good
chance that most required changes can be done using sth like
find -type f -exec perl -i -pe 's/(NOT|~~)\s+([^(]+)\s+([^(]+)/$1 ($2 $3)/g' {} +
(not tested).
Anyway I tend to keep this issue »on hold« for a few more days in case
there is further feedback.
Cheers,
Florian
Cheers,
GerwinOn 9 Nov 2019, at 21:04, Florian Haftmann <florian.haftmann@informatik.tu-muenchen.de> wrote:
Dear all,
I stumbled over a syntax issue with Bit-Not.
("NOT _" [70] 71)
("~~ _" [70] 71)This non-regular syntax leads to ambiguities:
term ‹NOT (2 * a)› \<comment>‹Prints as NOT 2 * a›
term ‹(NOT 2) * a› \<comment>‹Prints as NOT 2 * a›
term ‹NOT 2 * a› \<comment>‹Does not parse›Since NOT is a regular prefix operation, it would be much better to
avoid any tinkering with priorities:("NOT")
("~~")The change would require some precedence clarification in the
distribution and AFP, but to me it seems worth the effort.Cheers,
Florian
From: Florian Haftmann <florian.haftmann@informatik.tu-muenchen.de>
See now rev. a7d1fb0c9e16.
Florian
signature.asc
Last updated: Nov 21 2024 at 12:39 UTC