emacs-orgmode
[Top][All Lists]
Advanced

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

Re: Public API change: How to handle function signature change gracefull


From: Nicolas Goaziou
Subject: Re: Public API change: How to handle function signature change gracefully
Date: Tue, 21 Apr 2020 19:57:18 +0200
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/26.3 (gnu/linux)

Hello,

Benjamin Andresen writes:
>
> I would like to change the public API of the :face part of 
> `org-link-set-parameters':
>
>         (org-link-set-parameters "file" :face 'org-link)
>
> My ultimate goal is to have org-links be able to be have their face changed 
> based on the contents, not just the path of the link.
>
> I found the relevant code in org.el in the function `org-activate-links':
>
>    'face (pcase (org-link-get-parameter type :face)
>    ((and (pred functionp) face) (funcall face path))
>    ((and (pred facep) face) face)
>    ((and (pred consp) face) face) ;anonymous
>    (_ 'org-link))
>
> and would like to change this to 
>
>    'face (pcase (org-link-get-parameter type :face)
>    ((and (pred functionp) face) (funcall face path contents)) ;; this is the 
> change
>    ((and (pred facep) face) face)
>    ((and (pred consp) face) face) ;anonymous
>    (_ 'org-link))
>
> Now that will introduce a host of call issues because the callees don't 
> expect to the amount of arguments changed under their bottom.
>
> I would like some guidance how I could get what I think is neat: The
> contents of the bracket-style links as an additional parameter to set
> faces on and not breaking existing hookups.

Isn't the function called with point on the link? You may just need to
extract the contents from the environment.

Otherwise, a solution is to catch `wrong-number-of-arguments' error and
call again the function with the old calling convention. See, e.g.,
`org-link-open'.

> To make this more palatable I would suggest it's changed so that the
> 2nd argument will be a list of alists or keywords so to not have this
> function signature problem if someone else comes up with a reason to
> introduce yet more data.

Beware the over-engineering. At this point, someone motivated enough can
put an advice to the link fontification function.

Regards,

-- 
Nicolas Goaziou



reply via email to

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