Naming conventions for eieio slot accessors

Naming conventions for eieio slot accessors

[Fu Yuan]

[-- Attachment #1: Type: text/plain, Size: 611 bytes --]

Hi,

I'm writing a package which takes advantage of eieio.
When defining classes, I couldn't figure out how to name
the accessors.

First, should I use prefixes? Second, if I use a prefix,
should I use a double dash (--) or single dash (-)?

Say I'm writing a package called uikit.el and I have a class
`uikit-view' with a slot `name'.

Should I write

    (defclass uikit-view ()
      ((name
        :accessor uikit--name-of)))

or

    (defclass uikit-view ()
      ((name
        :accessor uikit-name-of)))

or

    (defclass uikit-view ()
      ((name
        :accessor name-of)))

Thanks,
Sincerely, Yuan

[-- Attachment #2: Type: text/html, Size: 3112 bytes --]

Re: Naming conventions for eieio slot accessors

[Stefan Monnier]

> I'm writing a package which takes advantage of eieio.
> When defining classes, I couldn't figure out how to name
> the accessors.

The accessor is a function, so it should at least follow the same naming
convention as used by other functions (e.g. use a "uikit-" or "uikit--"
prefix depending on whether you want to announce it as private or not).

After that, I don't think there are many more conventions to follow
because it's not used very often, so it seems nobody's come up with
a clear convention yet.  I guess you could follow the
"<typename>-<fieldname>" convention used in cl-defstruct.


        Stefan