[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
bug#57884: [PATCH] Flymake backend using the shellcheck program
From: |
Philip Kaludercic |
Subject: |
bug#57884: [PATCH] Flymake backend using the shellcheck program |
Date: |
Sun, 18 Sep 2022 20:22:33 +0000 |
Augusto Stoffel <arstoffel@gmail.com> writes:
> On Sun, 18 Sep 2022 at 13:46, Philip Kaludercic wrote:
>
>>> I like in your backend that you read a JSON output, which presumably
>>> provides the start _and end_ of the diagnostic region. How did you
>>> convert from line/column to buffer position?
>>
>> I didn't do that, Flymake takes care of that if you give it the buffer
>> position (unless I misunderstood something, but it appears to be working
>> for me).
>
> Okay, perhaps the docstring of `flymake-make-diagnostic' is easy to
> misinterpret, but apparently it works fine if one passes a (LINE . COL)
> cons for BEG and END, even if LOCUS is a buffer. This would also mean
> that the call to `flymake-diag-region` in the example backend of the
> Flymake manual is unnecessary, so I'm a bit suspicious here.
The example backend in the manual has had issues before so I wouldn't be
surprised if this were the case.
> Anyway, I rewrote the backend to use the JSON output of shellcheck,
> which has the advantage that it provides the end position of each
> diagnostic, so Flymake doesn't have to guess it (which is by nature
> sometimes inaccurate). Let me know what you think.
LGTM, but I haven't tested it yet.
> PS: What is the best practice w.r.t. the fact that json-serialize might
> not be available? It seems that every library that needs to parse JSON
> has to define a new alias similar to sh--json-read here. This is rather
> suboptimal...
All I can add is this: I've re-implemented the native-json functions in
Elisp for Compat[0], and it turns out to be more complicated than I had
initially expected -- though not impossible. If anyone has Compat
installed these definitions will also be loaded even if the user is
running Emacs 27.1+, since the test case it uses to check if the
definition should be installed is:
(not (condition-case nil
(equal (json-serialize '()) "{}")
(:success t)
(void-function nil)
(json-unavailable nil)))
As Compat is part of GNU ELPA, we could just copy these definitions as
fallback into json.el and replace the tests with (fboundp ...) checks.
[0]
https://git.sr.ht/~pkal/compat/tree/e827a9f3e2ba585de8b07cd246534a15ec414a8b/item/compat-27.el#L137
- bug#57884: [PATCH] Flymake backend using the shellcheck program, (continued)
- bug#57884: [PATCH] Flymake backend using the shellcheck program, Philip Kaludercic, 2022/09/18
- bug#57884: [PATCH] Flymake backend using the shellcheck program, Augusto Stoffel, 2022/09/18
- bug#57884: [PATCH] Flymake backend using the shellcheck program, Philip Kaludercic, 2022/09/18
- bug#57884: [PATCH] Flymake backend using the shellcheck program, Augusto Stoffel, 2022/09/18
- bug#57884: [PATCH] Flymake backend using the shellcheck program, Philip Kaludercic, 2022/09/18
- bug#57884: [PATCH] Flymake backend using the shellcheck program, Augusto Stoffel, 2022/09/18
- bug#57884: [PATCH] Flymake backend using the shellcheck program,
Philip Kaludercic <=
- bug#57884: [PATCH] Flymake backend using the shellcheck program, Philip Kaludercic, 2022/09/18
- bug#57884: [PATCH] Flymake backend using the shellcheck program, Augusto Stoffel, 2022/09/19
- bug#57884: [PATCH] Flymake backend using the shellcheck program, Philip Kaludercic, 2022/09/19
- bug#57884: [PATCH] Flymake backend using the shellcheck program, Augusto Stoffel, 2022/09/24
- bug#57884: [PATCH] Flymake backend using the shellcheck program, Philip Kaludercic, 2022/09/24
- bug#57884: [PATCH] Flymake backend using the shellcheck program, Philip Kaludercic, 2022/09/18