* bug#19670: 24.4; pcase-let bug?
@ 2015-01-24 2:42 Leo Liu
2015-01-24 5:07 ` Stefan Monnier
0 siblings, 1 reply; 4+ messages in thread
From: Leo Liu @ 2015-01-24 2:42 UTC (permalink / raw)
To: 19670
Eval the following two expressions:
1.
(pcase-let ((`(,x) '(1 2 3)))
(list x))
2.
(pcase '(1 2 3)
(`(,x) (list x)))
1 produces (1) and 2 nil. Bug?
Leo
^ permalink raw reply [flat|nested] 4+ messages in thread
* bug#19670: 24.4; pcase-let bug?
2015-01-24 2:42 bug#19670: 24.4; pcase-let bug? Leo Liu
@ 2015-01-24 5:07 ` Stefan Monnier
2015-01-26 14:27 ` Leo Liu
0 siblings, 1 reply; 4+ messages in thread
From: Stefan Monnier @ 2015-01-24 5:07 UTC (permalink / raw)
To: Leo Liu; +Cc: 19670
> 1 produces (1) and 2 nil. Bug?
Feature: the `pcase' form can have several branches (and defaults to
nil if no branch matches), whereas the `pcase-let' form doesn't have the
luxury of a "default case if it doesn't match", so we have to choose
between "signal an error if it doesn't match" or "pretend it matched".
I opted for the second choice.
Stefan
^ permalink raw reply [flat|nested] 4+ messages in thread
* bug#19670: 24.4; pcase-let bug?
2015-01-24 5:07 ` Stefan Monnier
@ 2015-01-26 14:27 ` Leo Liu
2015-01-26 22:59 ` Stefan Monnier
0 siblings, 1 reply; 4+ messages in thread
From: Leo Liu @ 2015-01-26 14:27 UTC (permalink / raw)
To: Stefan Monnier; +Cc: 19670
On 2015-01-24 13:07 +0800, Stefan Monnier wrote:
> Feature: the `pcase' form can have several branches (and defaults to
> nil if no branch matches), whereas the `pcase-let' form doesn't have the
> luxury of a "default case if it doesn't match", so we have to choose
> between "signal an error if it doesn't match" or "pretend it matched".
> I opted for the second choice.
I still find this inconsistent. Another similar-looking example:
(pcase 'whatever
((let `(,x) '(1 2 3)) (list x))) ;; => nil
It seems to make more sense for pcase-let to signal an error or bind X
to nil. WDYT?
Leo
^ permalink raw reply [flat|nested] 4+ messages in thread
* bug#19670: 24.4; pcase-let bug?
2015-01-26 14:27 ` Leo Liu
@ 2015-01-26 22:59 ` Stefan Monnier
0 siblings, 0 replies; 4+ messages in thread
From: Stefan Monnier @ 2015-01-26 22:59 UTC (permalink / raw)
To: Leo Liu; +Cc: 19670
> It seems to make more sense for pcase-let to signal an error or bind X
> to nil. WDYT?
Signaling an error would be fine. But it significantly increases the
macro-expanded code size (i.e. it's a lot less efficient).
Defaulting to binding all vars to nil in case of match failure seems
like an even more surprising behavior than what we have now.
The current pcase-let (and pcase-dolist) is meant as a "destructuring"
operation: it doesn't *check* anything, it just accesses subelements of
a structure.
Stefan
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2015-01-26 22:59 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2015-01-24 2:42 bug#19670: 24.4; pcase-let bug? Leo Liu
2015-01-24 5:07 ` Stefan Monnier
2015-01-26 14:27 ` Leo Liu
2015-01-26 22:59 ` Stefan Monnier
Code repositories for project(s) associated with this public inbox
https://git.savannah.gnu.org/cgit/emacs.git
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).