GNU bug report logs - #66144
29.1; eglot-shutdown request params violate JSONRPC spec

Previous Next

Package: emacs;

Reported by: Aaron Zeng <azeng <at> janestreet.com>

Date: Thu, 21 Sep 2023 21:00:01 UTC

Severity: normal

Found in version 29.1

Done: Stefan Kangas <stefankangas <at> gmail.com>

Bug is archived. No further changes may be made.

Full log


Message #53 received at 66144 <at> debbugs.gnu.org (full text, mbox):

From: João Távora <joaotavora <at> gmail.com>
To: Daniel Pettersson <daniel <at> dpettersson.net>
Cc: Javier Olaechea <pirata <at> gmail.com>, 66144 <at> debbugs.gnu.org,
 Eli Zaretskii <eliz <at> gnu.org>, Stefan Kangas <stefankangas <at> gmail.com>
Subject: Re: bug#66144: 29.1;
 eglot-shutdown request params violate JSONRPC spec
Date: Tue, 4 Feb 2025 12:08:58 +0000
I've now pushed Javier's idea to Emacs master, sending eglot-{} instead of nil
as arguments to the shutdown request.

I'll leave it up to you whether to keep the bug open for "fixing"
this in jsonrpc.el (IMO it should be closed since the original problem is
solved).

João

On Fri, Aug 9, 2024 at 11:58 PM Daniel Pettersson
<daniel <at> dpettersson.net> wrote:
>
>  Javier Olaechea <pirata <at> gmail.com> writes:
>
> > I did sit down twice to read the jsonrpc.el code and understand what
> > I need to do to make the
> > change I hoped to. But unfortunately I can't even say I understood
> > jsonrpc--async-request-1 well
> > enough.Or even, if that is indeed the function we need to change so
> > that we don't seed a params
>
> As I see it there are two candidates:
>
> Modify calls to jsonrpc-connection-send (jsonrpc--async-request-1,
> jsonrpc-notify, to keep the behavior consistent) such that
> jsonrpc-connection-send is not called with keyword :params if params is
> nil.
>
> Or make the change in jsonrpc-convert-to-endpoint which makes the
> behavior part of jsonrpc.el's API.  We are already doing some
> sanitizing here before serialization of a similar character.
>
> I think both options are fine, but I would favor the second one.
>
> > attribute when params is nil. It seems that we are passing the
> > params down in multiple arms?
>
> Both solution candidates above should account for this.
>
> This sort of change does come with some extensive regression testing, as
> João mentioned.



-- 
João Távora




This bug report was last modified 166 days ago.

Previous Next


GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.