bug-gnu-emacs
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

bug#12689: 24.2; Eshell ${cmd} substitution


From: Jim Porter
Subject: bug#12689: 24.2; Eshell ${cmd} substitution
Date: Wed, 9 Feb 2022 11:01:05 -0800

On 2/9/2022 1:12 AM, Lars Ingebrigtsen wrote:
Aidan Gauland <aidalgol@no8wireless.co.nz> writes:

In Eshell, ${cmd} expands to nil.
e.g.
$ echo ${/bin/echo -e "foo\nbar"}
$
Note the absence of any output from `echo'.

$ echo ${/bin/echo -e "foo\nbar"}-foo
nil-foo
n$

(I'm going through old bug reports that unfortunately weren't resolved
at the time.)

Looks like this has been fixed in the decade since it was reported:

/ $ echo ${/bin/echo -e "foo\nbar"}
("foo" "bar")
/ $ echo ${/bin/echo -e "foo\nbar"}-foo
("foo" "bar")-foo

So I'm therefore closing this bug report.  If there's more to be done
here, please respond to the debbugs address and we'll reopen.

I think the fix is from bug#30725. I'd meant to post an update to this bug as well, but I haven't had a chance to investigate the other half of this bug yet. Specifically, I think there's at least an argument that the result *should* look like this:

  / $ echo ${/bin/echo -e "foo\nbar"}
  foo
  bar
  / $ echo ${/bin/echo -e "foo\nbar"}-foo
  foo
  bar-foo

I'm still not 100% sure what's going on here, but I think the ${} subcommand evaluation splits the output on "\n" to make a list (which seems reasonable to me). However, when it gets converted back to a string (either to print directly, as in the first case, or to concatenate, as in the second case), it ends up looking like '("foo" "bar")'. You can see a little more clearly how these cases are different by calling `message' instead of `echo':

  ~ $ message %S ${*echo -e "foo\nbar"}
  ("foo" "bar")
  ~ $ message %S ${*echo -e "foo\nbar"}-foo
  "(\"foo\" \"bar\")-foo"

So the result of "${...}" is a list, and the result of "${...}-foo" is a string.

I think it would make sense to do something to improve how "foo\nbar" gets round-tripped here so that the result is more consistent with regular shells. I have a couple of ideas, but I'll need a bit of time to tinker with things to see how they work, and to find a solution that keeps incompatible changes to a minimum.





reply via email to

[Prev in Thread] Current Thread [Next in Thread]