Re: [AUCTeX-devel] Re: safe-local-variable expressions

[David Kastrup]

Reiner Steib <address@hidden> writes:

> On Wed, May 10 2006, David Kastrup wrote:
>
>> Reiner Steib <address@hidden> writes:
>>> | This variable is safe to use as a file local variable only if its value
>>> | satisfies the predicate `#[(x) [BYTE-CODE] [x (t nil shared dwim)] 2]'.
>>> | [...]
>>> `----
>>>
>>> Richard advised to quote the lambda expression in this case.  [...]
>>
>> Hm, maybe just make the function non-anonymous?
>
> You mean to introduce a function like `TeX-master-safe-p' to have...
>
> | This variable is safe to use as a file local variable only if its value
> | satisfies the predicate `TeX-master-safe-p'.
>
> ...  instead of the following?
>
> | This variable is safe to use as a file local variable only if its value
> | satisfies the predicate `(lambda (x) (or (stringp x) (member x (quote (t 
> nil shared dwim)))))'.
>
> [ Maybe we should discuss this on emacs-pretest-bug or emacs-devel
> because `TeX-master' is just one example? ]

Well, you have been discussing this already, as I have seen.
Personally, I find it nonsensical to print any anonymous form in this
context.  It would be more helpful to offer a button which will take
you to the point in the source code where the form has been defined.
But we probably don't have that kind of information for anonymous
byte-compiled functions, right?  Introducing a named function would
provide the link to the source.  If the function is anonymous, I don't
think we should print diagnostics short of "not considered safe by the
package author" or something.

I am directing this thread to emacs-devel (at least I try to).  It
seems more appropriate there.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum