From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tim Visher Subject: Re: File Scoped Properties? Date: Fri, 6 Mar 2020 09:09:07 -0500 Message-ID: References: <87mu8ugemp.fsf@ucl.ac.uk> Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="0000000000002a5a9e05a0303534" Return-path: Received: from eggs.gnu.org ([2001:470:142:3::10]:49089) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1jADfq-00040G-CT for emacs-orgmode@gnu.org; Fri, 06 Mar 2020 09:10:00 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1jADfo-0003HE-L9 for emacs-orgmode@gnu.org; Fri, 06 Mar 2020 09:09:58 -0500 Received: from mail-wr1-x430.google.com ([2a00:1450:4864:20::430]:43417) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1jADfo-00033J-Dg for emacs-orgmode@gnu.org; Fri, 06 Mar 2020 09:09:56 -0500 Received: by mail-wr1-x430.google.com with SMTP id v9so2505874wrf.10 for ; Fri, 06 Mar 2020 06:09:55 -0800 (PST) In-Reply-To: <87mu8ugemp.fsf@ucl.ac.uk> List-Id: "General discussions about Org-mode." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: emacs-orgmode-bounces+geo-emacs-orgmode=m.gmane-mx.org@gnu.org Sender: "Emacs-orgmode" To: "Fraga, Eric" Cc: Emacs Org Mode mailing list --0000000000002a5a9e05a0303534 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Thanks for the response, Eric. :) On Fri, Mar 6, 2020 at 1:51 AM Fraga, Eric wrote: > On Thursday, 5 Mar 2020 at 10:28, Tim Visher wrote: > > I _am_ able to get it to work by adding a file local variable like > > > > ``` > > # Local Variables: > > # org-attach-id-dir: "~/.foo/data" > > # End: > > ``` > > > > but then whenever I open the file it tells me it's possibly not safe to > set > > that. > > You've already received a more org-ish response but I'll give you an > Emacs response to this part of your post: Emacs is simply making sure > you are aware that a variable is being set when visiting a file. > Yep. That's fully understood. I'm less clear on why certain variables are considered safe and some are not but that doesn't seem relevant. I've done enough with file local and directory local variables in the past that I have a pretty clear understanding of what they do. My question was more around why I had to do it at all since based on my reading of the manual it seems like I should've been able to do this with some kind of file-wide property. Specifically, I've never been able to wrap my head around `(info "(org) Property Syntax")`. "Properties are key=E2=80=93value pairs. When they are associated with a si= ngle entry or with a tree=E2=80=A6," for instance, seems to imply by "When they = are associated with a single entry=E2=80=A6" that they can be associated with a= ll the entries (or a particular tree or node). Anyway, I think between you and Gustav I finally have this sorted. 1. To set properties at the top level of a file you need to use the `#+PROPERTY: ` syntax. I've been trying = to figure out how I misinterpreted that in the past and I _think_ it was because I assumed that the `#+PROPERTY` was actually `#+` as i= n `#+DIR` rather than `#+PROPERTY: DIR`. It looks like to set a file local property in an org file you _must_ (at least on 9.3 or earlier) use the `#+PROPERTY: ` syntax. 2. Even then by default org-attach property inheritance is set to `'selective` and `org-use-property-inheritance` is set to off. I've now customized `org-use-property-inheritance` to `'("DIR")` which I believe says that I consider the `DIR` property to be a possible candidate for inheritance and no others. I'm a little concerned about performance implications as the manual gives me all kinds of scary warnings but we'l= l see about that. Anyway this appears to work as I expect it to. I don't have to set a file-local variable to anything. I'm using org properties. And `org-attach-dir` now returns the proper directory for this file. Thanks all! If you think I'm still misunderstanding something please correct me. :) -- In Christ, Timmy V. https://blog.twonegatives.com https://five.sentenc.es --0000000000002a5a9e05a0303534 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable
Thanks for the response, Eric. :)

On Fri, Mar 6, 2020 at 1:51 AM Fraga, Eric <e.fraga@ucl.ac.uk> wrot= e:
On Thursday,=C2=A0 5 Mar 2020 at 10:28, Tim Visher wrote:
> I _am_ able to get it to work by adding a file local variable like
>
> ```
> # Local Variables:
> # org-attach-id-dir: "~/.foo/data"
> # End:
> ```
>
> but then whenever I open the file it tells me it's possibly not sa= fe to set
> that.

You've already received a more org-ish response but I'll give you a= n
Emacs response to this part of your post: Emacs is simply making sure
you are aware that a variable is being set when visiting a file.

Yep. That's fully understood. I'm less cl= ear on why certain variables are considered safe and some are not but that = doesn't seem relevant. I've done enough with file local and directo= ry local variables in the past that I have a pretty clear understanding of = what they do.

My question was more around why I ha= d to do it at all since based on my reading of the manual it seems like I s= hould've been able to do this with some kind of file-wide property. Spe= cifically, I've never been able to wrap my head around `(info "(or= g) Property Syntax")`.

"Properties are k= ey=E2=80=93value pairs. When they are associated with a single entry or wit= h a tree=E2=80=A6," for instance, seems to imply by "When they ar= e associated with a single entry=E2=80=A6" that they can be associated= with all the entries (or a particular tree or node).

<= div>Anyway, I think between you and Gustav I finally have this sorted.
  1. To set properties at the top level of a file you need to use = the `#+PROPERTY: <PROPERTY_NAME> <PROPERTY_VALUE>` syntax. I= 9;ve been trying to figure out how I misinterpreted that in the past and I = _think_ it was because I assumed that the `#+PROPERTY` was actually `#+<= PROPERTY>` as in `#+DIR` rather than `#+PROPERTY: DIR`. It looks like to= set a file local property in an org file you _must_ (at least on 9.3 or ea= rlier) use the `#+PROPERTY: <PROPERTY_NAME> <PROPERTY_VALUE>` s= yntax.

  2. Even then by default org-attach property inheritance= is set to `'selective` and `org-use-property-inheritance` is set to of= f. I've now customized `org-use-property-inheritance` to `'("D= IR")` which I believe says that I consider the `DIR` property to be a = possible candidate for inheritance and no others. I'm a little concerne= d about performance implications as the manual gives me all kinds of scary = warnings but we'll see about that.
Anyway this appears to= work as I expect it to. I don't have to set a file-local variable to a= nything. I'm using org properties. And `org-attach-dir` now returns the= proper directory for this file.

Thanks all!= If you think I'm still misunderstanding something please correct me. := )

--0000000000002a5a9e05a0303534--