Why may some Org regexps be risky when set as buffer-local variables?

Why may some Org regexps be risky when set as buffer-local variables?

[Thorsten Jolitz]

Hi List, 

there are dozens of regexps that are set as buffer-local variables when
org-mode is turned on as major-mode, but other regexps are defined and
set only once as global variables when e.g. org.el is loaded, and the
docstring of these variables often reads like this:

,---------------------------------------------------------------
| org-block-regexp is a variable defined in `org.el'.
| Its value is [...]
| 
|   This variable may be risky if used as a file-local variable.
| 
| Documentation:
| Regular expression for hiding blocks.
`---------------------------------------------------------------


Why are these variables considered risky when used as a file-local
variables? Was this "risk" the prime reason to make them global - or are
they global simply because there was no special reason to make them
buffer-local? 

-- 
cheers,
Thorsten

Re: Why may some Org regexps be risky when set as buffer-local variables?

[Nick Dokos]

Thorsten Jolitz <tjolitz@gmail.com> writes:

> Hi List, 
>
> there are dozens of regexps that are set as buffer-local variables when
> org-mode is turned on as major-mode, but other regexps are defined and
> set only once as global variables when e.g. org.el is loaded, and the
> docstring of these variables often reads like this:
>
> ,---------------------------------------------------------------
> | org-block-regexp is a variable defined in `org.el'.
> | Its value is [...]
> | 
> |   This variable may be risky if used as a file-local variable.
> | 
> | Documentation:
> | Regular expression for hiding blocks.
> `---------------------------------------------------------------
>
>
> Why are these variables considered risky when used as a file-local
> variables? Was this "risk" the prime reason to make them global - or are
> they global simply because there was no special reason to make them
> buffer-local? 

All it means is that nobody has gone to the trouble of certifying them
as safe. By default, file local variable are considered risky. See

  (info "(emacs) Safe File Variables")

for more.
-- 
Nick

Re: Why may some Org regexps be risky when set as buffer-local variables?

[Thorsten Jolitz]

Nick Dokos <ndokos@gmail.com> writes:

> Thorsten Jolitz <tjolitz@gmail.com> writes:
>
>> Hi List, 
>>
>> there are dozens of regexps that are set as buffer-local variables when
>> org-mode is turned on as major-mode, but other regexps are defined and
>> set only once as global variables when e.g. org.el is loaded, and the
>> docstring of these variables often reads like this:
>>
>> ,---------------------------------------------------------------
>> | org-block-regexp is a variable defined in `org.el'.
>> | Its value is [...]
>> | 
>> |   This variable may be risky if used as a file-local variable.
>> | 
>> | Documentation:
>> | Regular expression for hiding blocks.
>> `---------------------------------------------------------------
>>
>>
>> Why are these variables considered risky when used as a file-local
>> variables? Was this "risk" the prime reason to make them global - or are
>> they global simply because there was no special reason to make them
>> buffer-local? 
>
> All it means is that nobody has gone to the trouble of certifying them
> as safe. By default, file local variable are considered risky. See
>
>   (info "(emacs) Safe File Variables")
>
> for more.

Ok, then as long as these variables hold strings there is no risk anyway
since e.g. (eval "(+ 2 2)") gives "(+ 2 2)" and not 4. Thanks.

-- 
cheers,
Thorsten