Check whether an object is an alist

Check whether an object is an alist

[Heime]

Is there a function that checks whether an object is an alist.
Have only found listp.

Re: Check whether an object is an alist

[Heime]

Sent with Proton Mail secure email.

On Monday, September 16th, 2024 at 10:29 PM, Heime <heimeborgia@protonmail.com> wrote:

> Is there a function that checks whether an object is an alist.
> Have only found listp.

Because I may have to come up with a solution, I have made this one.
Am I missing anything ?

(defun torium-alist-p (obj)
  "Return t if object OBJ is an association list (alist)."
  (and (listp obj)
       (not (null obj))
       (every (lambda (x)
                (and (consp x)
                     (not (null x))))
              obj)))

Re: Check whether an object is an alist

[Manuel Giraud via Users list for the GNU Emacs text editor]

Heime <heimeborgia@protonmail.com> writes:

> Sent with Proton Mail secure email.
>
> On Monday, September 16th, 2024 at 10:29 PM, Heime <heimeborgia@protonmail.com> wrote:
>
>> Is there a function that checks whether an object is an alist.
>> Have only found listp.
>
> Because I may have to come up with a solution, I have made this one.
> Am I missing anything ?
>
> (defun torium-alist-p (obj)
>   "Return t if object OBJ is an association list (alist)."
>   (and (listp obj)
>        (not (null obj))
>        (every (lambda (x)
>                 (and (consp x)
>                      (not (null x))))
>               obj)))

Looks good to me.  I think it should `cl-every' instead of `every' and
also, maybe, nil is a valid empty alist.
-- 
Manuel Giraud

Re: Check whether an object is an alist

[tomas]

[-- Attachment #1: Type: text/plain, Size: 1111 bytes --]

On Tue, Sep 17, 2024 at 11:38:46AM +0200, Manuel Giraud via Users list for the GNU Emacs text editor wrote:
> Heime <heimeborgia@protonmail.com> writes:
> 
> > Sent with Proton Mail secure email.
> >
> > On Monday, September 16th, 2024 at 10:29 PM, Heime <heimeborgia@protonmail.com> wrote:
> >
> >> Is there a function that checks whether an object is an alist.
> >> Have only found listp.
> >
> > Because I may have to come up with a solution, I have made this one.
> > Am I missing anything ?
> >
> > (defun torium-alist-p (obj)
> >   "Return t if object OBJ is an association list (alist)."
> >   (and (listp obj)
> >        (not (null obj))
> >        (every (lambda (x)
> >                 (and (consp x)
> >                      (not (null x))))
> >               obj)))
> 
> Looks good to me.  I think it should `cl-every' instead of `every' and
> also, maybe, nil is a valid empty alist.

Unless you'd like to have nil as a possible key (alists allow that,
why not?):

  (setq al '((nil . "I am not") (t . "I am")))
  (alist-get nil al)
  => "I am not"

Cheers
-- 
t

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

Re: Check whether an object is an alist

[Heime]

Sent with Proton Mail secure email.

On Tuesday, September 17th, 2024 at 9:46 PM, tomas@tuxteam.de <tomas@tuxteam.de> wrote:

> On Tue, Sep 17, 2024 at 11:38:46AM +0200, Manuel Giraud via Users list for the GNU Emacs text editor wrote:
> 
> > Heime heimeborgia@protonmail.com writes:
> > 
> > > Sent with Proton Mail secure email.
> > > 
> > > On Monday, September 16th, 2024 at 10:29 PM, Heime heimeborgia@protonmail.com wrote:
> > > 
> > > > Is there a function that checks whether an object is an alist.
> > > > Have only found listp.
> > > 
> > > Because I may have to come up with a solution, I have made this one.
> > > Am I missing anything ?
> > > 
> > > (defun torium-alist-p (obj)
> > > "Return t if object OBJ is an association list (alist)."
> > > (and (listp obj)
> > > (not (null obj))
> > > (every (lambda (x)
> > > (and (consp x)
> > > (not (null x))))
> > > obj)))
> > 
> > Looks good to me. I think it should `cl-every' instead of` every' and
> > also, maybe, nil is a valid empty alist.
> 
> 
> Unless you'd like to have nil as a possible key (alists allow that,
> why not?):
> 
> (setq al '((nil . "I am not") (t . "I am")))
> (alist-get nil al)
> => "I am not"
> 
> 
> Cheers
> --
> t

And we end up with

(defun torium-alist-p (obj)
  "Return t if object OBJ is an association list (alist)."
  (and (listp obj)
       (cl-every (lambda (x)
                   (or (null x)
                       (consp x)))
                 obj)))

I am quite sure I do not have to check anything upon tho state of (car x).
Agreed ?

Re: Check whether an object is an alist

[tomas]

[-- Attachment #1: Type: text/plain, Size: 1581 bytes --]

On Tue, Sep 17, 2024 at 11:16:24AM +0000, Heime wrote:
> 
> 
> 
> 
> 
> Sent with Proton Mail secure email.
> 
> On Tuesday, September 17th, 2024 at 9:46 PM, tomas@tuxteam.de <tomas@tuxteam.de> wrote:
> 
> > On Tue, Sep 17, 2024 at 11:38:46AM +0200, Manuel Giraud via Users list for the GNU Emacs text editor wrote:
> > 
> > > Heime heimeborgia@protonmail.com writes:
> > > 
> > > > Sent with Proton Mail secure email.
> > > > 
> > > > On Monday, September 16th, 2024 at 10:29 PM, Heime heimeborgia@protonmail.com wrote:
> > > > 
> > > > > Is there a function that checks whether an object is an alist.
> > > > > Have only found listp.
> > > > 
> > > > Because I may have to come up with a solution, I have made this one.
> > > > Am I missing anything ?
> > > > 
> > > > (defun torium-alist-p (obj)
> > > > "Return t if object OBJ is an association list (alist)."
> > > > (and (listp obj)
> > > > (not (null obj))
> > > > (every (lambda (x)
> > > > (and (consp x)
> > > > (not (null x))))
> > > > obj)))
> > > 
> > > Looks good to me. I think it should `cl-every' instead of` every' and
> > > also, maybe, nil is a valid empty alist.
> > 
> > 
> > Unless you'd like to have nil as a possible key (alists allow that,
> > why not?):
> > 
> > (setq al '((nil . "I am not") (t . "I am")))
> > (alist-get nil al)
> > => "I am not"

Oh, wait. I just see you were checking the pair to be not
null. Then it's easier: your above code seems right, just
the (not (null x)) is superfluous, since (consp x) makes
sure it is so.

Cheers
-- 
t

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

Re: Check whether an object is an alist

[Heime]

On Tuesday, September 17th, 2024 at 11:23 PM, tomas@tuxteam.de <tomas@tuxteam.de> wrote:

> On Tue, Sep 17, 2024 at 11:16:24AM +0000, Heime wrote:
> 
> > Sent with Proton Mail secure email.
> > 
> > On Tuesday, September 17th, 2024 at 9:46 PM, tomas@tuxteam.de tomas@tuxteam.de wrote:
> > 
> > > On Tue, Sep 17, 2024 at 11:38:46AM +0200, Manuel Giraud via Users list for the GNU Emacs text editor wrote:
> > > 
> > > > Heime heimeborgia@protonmail.com writes:
> > > > 
> > > > > Sent with Proton Mail secure email.
> > > > > 
> > > > > On Monday, September 16th, 2024 at 10:29 PM, Heime heimeborgia@protonmail.com wrote:
> > > > > 
> > > > > > Is there a function that checks whether an object is an alist.
> > > > > > Have only found listp.
> > > > > 
> > > > > Because I may have to come up with a solution, I have made this one.
> > > > > Am I missing anything ?
> > > > > 
> > > > > (defun torium-alist-p (obj)
> > > > > "Return t if object OBJ is an association list (alist)."
> > > > > (and (listp obj)
> > > > > (not (null obj))
> > > > > (every (lambda (x)
> > > > > (and (consp x)
> > > > > (not (null x))))
> > > > > obj)))
> > > > 
> > > > Looks good to me. I think it should `cl-every' instead of` every' and
> > > > also, maybe, nil is a valid empty alist.
> > > 
> > > Unless you'd like to have nil as a possible key (alists allow that,
> > > why not?):
> > > 
> > > (setq al '((nil . "I am not") (t . "I am")))
> > > (alist-get nil al)
> > > => "I am not"
> 
> 
> Oh, wait. I just see you were checking the pair to be not
> null. Then it's easier: your above code seems right, just
> the (not (null x)) is superfluous, since (consp x) makes
> sure it is so.
> 
> Cheers
> --
> t



Have modified the code to this

(defun torium-alist-p (obj)
  "Return t if object OBJ is an association list (alist)."
  (and (listp obj)
       (every (lambda (x)
                (or (null x)
                    (consp x)))
              obj)))

I want to allow empty lists and empty keys.  Or should I remove (null x)
as well ?

Re: Check whether an object is an alist

[tomas]

[-- Attachment #1: Type: text/plain, Size: 847 bytes --]

On Tue, Sep 17, 2024 at 01:51:54PM +0000, Heime wrote:

[...]

> Have modified the code to this
> 
> (defun torium-alist-p (obj)
>   "Return t if object OBJ is an association list (alist)."
>   (and (listp obj)
>        (every (lambda (x)
>                 (or (null x)
>                     (consp x)))
>               obj)))
> 
> I want to allow empty lists and empty keys.  Or should I remove (null x)
> as well ?

That depends on how strict you want to be. Have you read elisp's
docs on association lists?

Elisp tends to tolerate alists whose elements are not pairs. Other
Lisps are stricter. Your function is a funny in-between (nil non-pairs
are OK, but others are not).

First try to understand what you are trying to do, then do. Note that,
contrary to hashes, alists are not strictly defined.

Cheers
-- 
t

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

Re: Check whether an object is an alist

[Heime]

On Wednesday, September 18th, 2024 at 1:58 AM, tomas@tuxteam.de <tomas@tuxteam.de> wrote:

> On Tue, Sep 17, 2024 at 01:51:54PM +0000, Heime wrote:
> 
> [...]
> 
> > Have modified the code to this
> > 
> > (defun torium-alist-p (obj)
> > "Return t if object OBJ is an association list (alist)."
> > (and (listp obj)
> > (every (lambda (x)
> > (or (null x)
> > (consp x)))
> > obj)))
> > 
> > I want to allow empty lists and empty keys. Or should I remove (null x)
> > as well ?
> 
> 
> That depends on how strict you want to be. Have you read elisp's
> docs on association lists?
> 
> Elisp tends to tolerate alists whose elements are not pairs. Other
> Lisps are stricter. Your function is a funny in-between (nil non-pairs
> are OK, but others are not).
> 
> First try to understand what you are trying to do, then do. Note that,
> contrary to hashes, alists are not strictly defined.
> 
> Cheers
> --
> t

Originally I tried to find something as alistp but did not find any.  I would
have stayed to what it would have allowed.  But then I started writing my own.
For what I am doing, there is always a key and value in each cell.  

But taking in consideration what others have been saying here about alists, I 
decided to be more flexible and allow empty alists and keys with nil.  Then
(null x) should not be there because it allows nil non-pairs.


This is what I am trying to do with

(defun torium-alist-p (obj)
  "Return t if object OBJ is an association list (alist).  An alist 
can be empty and nil can act as a key."
  (and (listp obj)
       (every (lambda (x)
                (or (null x)
                (consp x)))
              obj)))

Re: Check whether an object is an alist

[tomas]

[-- Attachment #1: Type: text/plain, Size: 676 bytes --]

On Tue, Sep 17, 2024 at 02:32:33PM +0000, Heime wrote:

[...]

> But taking in consideration what others have been saying here about alists, I 
> decided to be more flexible and allow empty alists and keys with nil.  Then
> (null x) should not be there because it allows nil non-pairs.

OK.

> This is what I am trying to do with
> 
> (defun torium-alist-p (obj)
>   "Return t if object OBJ is an association list (alist).  An alist 
> can be empty and nil can act as a key."
>   (and (listp obj)
>        (every (lambda (x)
>                 (or (null x) ;; <== so this one goes out?
>                 (consp x)))
>               obj)))

Cheers
-- 
t

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

Re: Check whether an object is an alist

[Heime]

On Wednesday, September 18th, 2024 at 3:20 AM, tomas@tuxteam.de <tomas@tuxteam.de> wrote:

> On Tue, Sep 17, 2024 at 02:32:33PM +0000, Heime wrote:
> 
> [...]
> 
> > But taking in consideration what others have been saying here about alists, I
> > decided to be more flexible and allow empty alists and keys with nil. Then
> > (null x) should not be there because it allows nil non-pairs.
> 
> 
> OK.
> 
> > This is what I am trying to do with
> > 
> > (defun torium-alist-p (obj)
> > "Return t if object OBJ is an association list (alist). An alist
> > can be empty and nil can act as a key."
> > (and (listp obj)
> > (every (lambda (x)
> > (or (null x) ;; <== so this one goes out?
> > (consp x)))
> > obj)))
> 
> 
> Cheers
> --
> t

Is (null x) the part that allows nil, a non-pair object ?

Re: Check whether an object is an alist

[tomas]

[-- Attachment #1: Type: text/plain, Size: 630 bytes --]

On Tue, Sep 17, 2024 at 04:26:19PM +0000, Heime wrote:

[...]

> Is (null x) the part that allows nil, a non-pair object ?

Just ask Emacs:

  (cl-every (lambda (x)
              (or (null x)
                  (consp x)))
            '((foo . bar)
              nil
              (baz . quux)))
  
  => t
  
  (cl-every (lambda (x)
              (consp x))
            '((foo . bar)
              nil
              (baz . quux)))
  
  => nil

Now, should you decide you prefer the second variant, there is
a more concise way of writing it. This is left as an exercise
for the reader.

Cheers
-- 
t

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

Re: Check whether an object is an alist

[Heime]

On Wednesday, September 18th, 2024 at 5:57 AM, tomas@tuxteam.de <tomas@tuxteam.de> wrote:

> On Tue, Sep 17, 2024 at 04:26:19PM +0000, Heime wrote:
> 
> [...]
> 
> > Is (null x) the part that allows nil, a non-pair object ?
> 
> 
> Just ask Emacs:
> 
> (cl-every (lambda (x)
> (or (null x)
> (consp x)))
> '((foo . bar)
> nil
> (baz . quux)))
> 
> => t
> 
> 
> (cl-every (lambda (x)
> (consp x))
> '((foo . bar)
> nil
> (baz . quux)))
> 
> => nil
> 
> 
> Now, should you decide you prefer the second variant, there is
> a more concise way of writing it. This is left as an exercise
> for the reader.
> 
> Cheers
> --
> t

(cl-every #'consp '((foo . bar) nil (baz . quux)))

Re: Check whether an object is an alist

[tomas]

[-- Attachment #1: Type: text/plain, Size: 267 bytes --]

On Tue, Sep 17, 2024 at 06:04:41PM +0000, Heime wrote:
> On Wednesday, September 18th, 2024 at 5:57 AM, tomas@tuxteam.de <tomas@tuxteam.de> wrote:

> > [...] as an exercise for the reader.
> 
> (cl-every #'consp '((foo . bar) nil (baz . quux)))

Yes.
-- 
t

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

Re: Check whether an object is an alist

[Heime]

On Wednesday, September 18th, 2024 at 6:28 AM, tomas@tuxteam.de <tomas@tuxteam.de> wrote:

> On Tue, Sep 17, 2024 at 06:04:41PM +0000, Heime wrote:
> 
> > On Wednesday, September 18th, 2024 at 5:57 AM, tomas@tuxteam.de tomas@tuxteam.de wrote:
> 
> > > [...] as an exercise for the reader.
> > 
> > (cl-every #'consp '((foo . bar) nil (baz . quux)))
> 
> 
> Yes.
> --
> t

Have done this small test, and even though there is a nil (or anything 
else, e.g. jarod), assoc still works on the other keys.  This means that 
one can still be able to get correct values even if the object is not
a complete alist.  In a sense, if I am looking for keys in an object,
even though it is not an alist in a strict sense, things might still
work correctly.  Is this the reason there does not exist an alistp ? 
Or because traversing the list would be expensive ? 

But if we want to be strict that an alist is a list of cons cells 
with pairs of values, then only t1 should be accepted. 

(let ( (t1 '((foo . bar) (alice . bob) (baz . qux)))
       (t2 '((foo . bar) nil (alice . bob) (baz . qux)))
       (t3 '((foo . bar) jarod (alice . bob) (baz . qux))) )
  (message "%s" (assoc 'foo t1))
  (message "%s" (assoc 'alice t2))
  (message "%s" (assoc 'baz t3)) )

(foo . bar)
(alice . bob)
(baz . qux)

Re: Check whether an object is an alist

[tomas]

[-- Attachment #1: Type: text/plain, Size: 976 bytes --]

On Tue, Sep 17, 2024 at 07:20:22PM +0000, Heime wrote:
> On Wednesday, September 18th, 2024 at 6:28 AM, tomas@tuxteam.de <tomas@tuxteam.de> wrote:
> 
> > On Tue, Sep 17, 2024 at 06:04:41PM +0000, Heime wrote:
> > 
> > > On Wednesday, September 18th, 2024 at 5:57 AM, tomas@tuxteam.de tomas@tuxteam.de wrote:
> > 
> > > > [...] as an exercise for the reader.
> > > 
> > > (cl-every #'consp '((foo . bar) nil (baz . quux)))
> > 
> > 
> > Yes.
> > --
> > t
> 
> Have done this small test, and even though there is a nil (or anything 
> else, e.g. jarod), assoc still works on the other keys.

That's what I was saying (and what Elisp's doc, which is recommended
reading also says). There are different views on that, and Elisp
itself is one of the less strict: it just ignores list elements which
aren't assocs.

There is some historical background to that. "Classical" Lisp "types"
are just defined by the operations defined on them.

Cheers
-- 
t

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

Re: Check whether an object is an alist

[Heime]

Sent with Proton Mail secure email.

On Wednesday, September 18th, 2024 at 7:35 AM, tomas@tuxteam.de <tomas@tuxteam.de> wrote:

> On Tue, Sep 17, 2024 at 07:20:22PM +0000, Heime wrote:
> 
> > On Wednesday, September 18th, 2024 at 6:28 AM, tomas@tuxteam.de tomas@tuxteam.de wrote:
> > 
> > > On Tue, Sep 17, 2024 at 06:04:41PM +0000, Heime wrote:
> > > 
> > > > On Wednesday, September 18th, 2024 at 5:57 AM, tomas@tuxteam.de tomas@tuxteam.de wrote:
> > > 
> > > > > [...] as an exercise for the reader.
> > > > 
> > > > (cl-every #'consp '((foo . bar) nil (baz . quux)))
> > > 
> > > Yes.
> > > --
> > > t
> > 
> > Have done this small test, and even though there is a nil (or anything
> > else, e.g. jarod), assoc still works on the other keys.
> 
> 
> That's what I was saying (and what Elisp's doc, which is recommended
> reading also says). There are different views on that, and Elisp
> itself is one of the less strict: it just ignores list elements which
> aren't assocs.
> 
> There is some historical background to that. "Classical" Lisp "types"
> are just defined by the operations defined on them.
> 
> Cheers
> --
> t

After changing to the following, I am getting validation for list 
element (1 2 3), even though it is not an association. 

(defun torium-alist-p (obj)
  "Return t if object OBJ is an association list (alist)."
    (and (listp obj)
         (cl-every (lambda (x) #'consp) obj)))

Re: Check whether an object is an alist

[Heime]

Sent with Proton Mail secure email.

On Wednesday, September 18th, 2024 at 9:02 AM, Heime <heimeborgia@protonmail.com> wrote:

> 
> 
> 
> 
> 
> Sent with Proton Mail secure email.
> 
> 
> On Wednesday, September 18th, 2024 at 7:35 AM, tomas@tuxteam.de tomas@tuxteam.de wrote:
> 
> > On Tue, Sep 17, 2024 at 07:20:22PM +0000, Heime wrote:
> > 
> > > On Wednesday, September 18th, 2024 at 6:28 AM, tomas@tuxteam.de tomas@tuxteam.de wrote:
> > > 
> > > > On Tue, Sep 17, 2024 at 06:04:41PM +0000, Heime wrote:
> > > > 
> > > > > On Wednesday, September 18th, 2024 at 5:57 AM, tomas@tuxteam.de tomas@tuxteam.de wrote:
> > > > 
> > > > > > [...] as an exercise for the reader.
> > > > > 
> > > > > (cl-every #'consp '((foo . bar) nil (baz . quux)))
> > > > 
> > > > Yes.
> > > > --
> > > > t
> > > 
> > > Have done this small test, and even though there is a nil (or anything
> > > else, e.g. jarod), assoc still works on the other keys.
> > 
> > That's what I was saying (and what Elisp's doc, which is recommended
> > reading also says). There are different views on that, and Elisp
> > itself is one of the less strict: it just ignores list elements which
> > aren't assocs.
> > 
> > There is some historical background to that. "Classical" Lisp "types"
> > are just defined by the operations defined on them.
> > 
> > Cheers
> > --
> > t
> 
> 
> After changing to the following, I am getting validation for list
> element (1 2 3), even though it is not an association.
> 
> (defun torium-alist-p (obj)
> "Return t if object OBJ is an association list (alist)."
> (and (listp obj)
> (cl-every (lambda (x) #'consp) obj)))

I remember.  No requirements of lambda to run #'consp on obj

Thus, again

(defun torium-alist-p (obj)
  "Return t if object OBJ is an association list (alist)."
  (and (listp obj)
       (cl-every #'consp obj)))

Re: Check whether an object is an alist

[tomas]

[-- Attachment #1: Type: text/plain, Size: 792 bytes --]

On Tue, Sep 17, 2024 at 09:25:44PM +0000, Heime wrote:

[...]

> > After changing to the following, I am getting validation for list
> > element (1 2 3), even though it is not an association.
> > 
> > (defun torium-alist-p (obj)
> > "Return t if object OBJ is an association list (alist)."
> > (and (listp obj)
> > (cl-every (lambda (x) #'consp) obj)))
> 
> I remember.  No requirements of lambda to run #'consp on obj

Much more important: /why/ are you "getting a validation" for list '(1 2 3)
with the above form and...

> (defun torium-alist-p (obj)
>   "Return t if object OBJ is an association list (alist)."
>   (and (listp obj)
>        (cl-every #'consp obj)))

...(probably, you didn't tell us) not with this one?

Why are they different?

Cheers
-- 
t

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

Re: Check whether an object is an alist

[Heime]

On Wednesday, September 18th, 2024 at 4:21 PM, tomas@tuxteam.de <tomas@tuxteam.de> wrote:

> On Tue, Sep 17, 2024 at 09:25:44PM +0000, Heime wrote:
> 
> [...]
> 
> > > After changing to the following, I am getting validation for list
> > > element (1 2 3), even though it is not an association.
> > > 
> > > (defun torium-alist-p (obj)
> > > "Return t if object OBJ is an association list (alist)."
> > > (and (listp obj)
> > > (cl-every (lambda (x) #'consp) obj)))
> > 
> > I remember. No requirements of lambda to run #'consp on obj
> 
> 
> Much more important: /why/ are you "getting a validation" for list '(1 2 3)
> with the above form and...
> 
> > (defun torium-alist-p (obj)
> > "Return t if object OBJ is an association list (alist)."
> > (and (listp obj)
> > (cl-every #'consp obj)))
> 
> 
> ...(probably, you didn't tell us) not with this one?
> 
> Why are they different?
> 
> Cheers
> --
> t

Have solved the problem.  Was not with that one, but that one solved the 
problem.


Felicitations

Re: Check whether an object is an alist

[Stephen Berman]

On Tue, 17 Sep 2024 21:02:03 +0000 Heime <heimeborgia@protonmail.com> wrote:

[...]
> After changing to the following, I am getting validation for list 
> element (1 2 3), even though it is not an association. 

It is according to (info "(elisp) Association Lists"):

     Both the values and the keys in an alist may be any Lisp objects.
  For example, in the following alist, the symbol ‘a’ is associated with
  the number ‘1’, and the string ‘"b"’ is associated with the _list_ ‘(2
  3)’, which is the CDR of the alist element:
  
       ((a . 1) ("b" 2 3))

> (defun torium-alist-p (obj)
>   "Return t if object OBJ is an association list (alist)."
>     (and (listp obj)
>          (cl-every (lambda (x) #'consp) obj)))

Steve Berman

RE: [External] : Re: Check whether an object is an alist

[Drew Adams]

> (defun torium-alist-p (obj)
>   (and (listp obj)
>        (every (lambda (x)
>                 (or (null x)
>                 (consp x)))
>               obj)))

`listp' does exactly this: (or (null x) (consp x)).

So your code just does this:

 (and (listp obj) (every #'listp obj))