unescaped character literals

unescaped character literals

[Andreas Röhler]

When opening Emacs26-pretest, get a message:

unescaped character literals `?)' detected!

AFAIU it refers to a test which contains

(should (eq (char-before) ?))

which seems right. Should that warning be reported as bug?

Cheers,

Andreas

Re: unescaped character literals

[Emanuel Berg]

Andreas Röhler wrote:

> When opening Emacs26-pretest, get a message:
>
> unescaped character literals `?)' detected!
>
> AFAIU it refers to a test which contains
>
> (should (eq (char-before) ?))

Perhaps just escape it?

    (should (eq (char-before) ?\)) ; t

?

-- 
underground experts united
http://user.it.uu.se/~embe8573

Re: unescaped character literals

[Emanuel Berg]

> Perhaps just escape it?
>
>     (should (eq (char-before) ?\)) ; t
>
> ?

No, there should be three parens! Paren match
doesn't work in message mode with such
exotic escaping...

    (should
         (eq (char-before) ?\))
     )

-- 
underground experts united
http://user.it.uu.se/~embe8573

Re: unescaped character literals

[Emanuel Berg]

> No, there should be three parens! Paren match
> doesn't work in message mode with such exotic
> escaping...
>
>     (should
>          (eq (char-before) ?\))
>      )

I forgot to document it:

;; only after a while
;; are you expected
;; to understand this

-- 
underground experts united
http://user.it.uu.se/~embe8573

Re: unescaped character literals

[Emanuel Berg]

Why does one use `should' ?

    (should FORM)

    Evaluate FORM. If it returns nil, abort the
    current test as failed.

    Returns the value of FORM.

Isn't that what plain `or' do?

    (or 1) ; 1

-- 
underground experts united
http://user.it.uu.se/~embe8573

Re: unescaped character literals

[Emanuel Berg]

> Isn't that what plain `or' do?
>
>     (or 1) ; 1

Actually you don't even need `or'.

    1 ; 1

Ha ha h... eheh

-- 
underground experts united
http://user.it.uu.se/~embe8573

Re: unescaped character literals

[John Mastro]

Emanuel Berg <moasen@zoho.com> wrote:
> Why does one use `should' ?
>
>     (should FORM)
>
>     Evaluate FORM. If it returns nil, abort the
>     current test as failed.
>
>     Returns the value of FORM.
>
> Isn't that what plain `or' do?

`should' is part of ERT, a testing tool. That's what the docstring is
referring to when it says "abort the current test as failed".

See (info "(ERT)")

        John

Re: unescaped character literals

[Emanuel Berg]

John Mastro wrote:

> `should' is part of ERT, a testing tool.
> That's what the docstring is referring to
> when it says "abort the current test as
> failed".
>
> See (info "(ERT)")

OK.

Well, testing is actually great! It is the best
tool to find bugs in software second only to
using it. YRM V.
             ^not

Or actually they do! And that makes it
even better.

-- 
underground experts united
http://user.it.uu.se/~embe8573

Re: unescaped character literals

[Andreas Röhler]

On 08.02.2018 17:32, Andreas Röhler wrote:
> When opening Emacs26-pretest, get a message:
> 
> unescaped character literals `?)' detected!
> 
> AFAIU it refers to a test which contains
> 
> (should (eq (char-before) ?))
> 
> which seems right. Should that warning be reported as bug?
> 
> Cheers,
> 
> Andreas
> 

Okay, should escape it, because otherwise the paren-match goes wrong.