bug#24381: 25.0.95; eval-after-load for features

bug#24381: 25.0.95; eval-after-load for features

[Roland Winkler]

cat > foo.el <<EOF
(eval-after-load 'foo
  (message "feature foo loaded: %s" (featurep 'foo)))
EOF

emacs -Q -l foo.el

According to the doc string of eval-after-load:

  Alternatively, FILE can be a feature (i.e. a symbol), in which case FORM
  is evaluated at the end of any file that ‘provide’s this feature.

So I would expect that the above form is not run because the file
foo.el does not provide the feature foo.  Yet the algorithm
underlying eval-after-load compares the symbol foo with the string
"foo.el" and this is sufficient to evaluate the form.

I think that the behavior described in the docstring would be
useful: my init.el is broken into smaller files which have the same
names as the features they refer to.  So I suggest to change the
behavior of eval-after-load instead of fixing its docstring.  But
there might also be other issues that I am not aware of.




In GNU Emacs 25.0.95.1 (x86_64-unknown-linux-gnu, GTK+ Version 3.10.8)
 of 2016-07-15 built on lukas
Windowing system distributor 'The X.Org Foundation', version 11.0.11600000
System Description:	Ubuntu 14.04.5 LTS

bug#24381: 25.0.95; eval-after-load for features

[Clément Pit--Claudel]

[-- Attachment #1.1: Type: text/plain, Size: 1359 bytes --]

Did you mean to use with-eval-after-load?  eval-after-load takes a quoted form IIRC.

Clément.

On 2016-09-06 11:11, Roland Winkler wrote:
> 
> cat > foo.el <<EOF
> (eval-after-load 'foo
>   (message "feature foo loaded: %s" (featurep 'foo)))
> EOF
> 
> emacs -Q -l foo.el
> 
> According to the doc string of eval-after-load:
> 
>   Alternatively, FILE can be a feature (i.e. a symbol), in which case FORM
>   is evaluated at the end of any file that ‘provide’s this feature.
> 
> So I would expect that the above form is not run because the file
> foo.el does not provide the feature foo.  Yet the algorithm
> underlying eval-after-load compares the symbol foo with the string
> "foo.el" and this is sufficient to evaluate the form.
> 
> I think that the behavior described in the docstring would be
> useful: my init.el is broken into smaller files which have the same
> names as the features they refer to.  So I suggest to change the
> behavior of eval-after-load instead of fixing its docstring.  But
> there might also be other issues that I am not aware of.
> 
> 
> 
> 
> In GNU Emacs 25.0.95.1 (x86_64-unknown-linux-gnu, GTK+ Version 3.10.8)
>  of 2016-07-15 built on lukas
> Windowing system distributor 'The X.Org Foundation', version 11.0.11600000
> System Description:	Ubuntu 14.04.5 LTS
> 
> 
> 
> 


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

bug#24381: 25.0.95; eval-after-load for features

[Roland Winkler]

On Tue Sep 6 2016 Clément Pit--Claudel wrote:
> Did you mean to use with-eval-after-load?  eval-after-load takes a
> quoted form IIRC.

You are right, it must have been too early in the morning for me.
I am sorry for the noise.

Closing.