Incorrect spacing in C/C++ pointer type variable definitions

[davidshepherd7]

We still can't distinguish between multiplication and pointer variables outside of function argument lists.

So for example

Thingy*my_thingy;

becomes

Thingy * my_thingy;

even though it's (probably) a pointer variable not a multiplication.

Related to #7

[andreas-roehler]

*mypointer = 20;
works for me.

++(*current);
fails.

Might a dereference be signaled in the midst of a [a-z]string?

[davidshepherd7]

Moved from #8, because I think this is where you intended to ask the question:

Spotted so far this:

electric-operator-c-pointer-type-style 'variable
"Defines how C/C++ mode pointer and reference types are spaced.
If set to 'variable' then the operator is touching the variable
name, as in int *x'. If set to 'type' then the operator is touching the type name , as in int* x'."

;;;;

Is this really a matter of style? I.e. are int *x' and int* x' semantically equal?

Yeah this is entirely a matter of style.

*mypointer = 20;
works for me.

++(*current);
fails.

Might a dereference be signaled in the midst of a [a-z]string?

Yeah, I'm using a bunch of heuristics at the moment which will make the most common cases work but they aren't always good enough.

[andreas-roehler]

Think this regexp might be changed to: space operator only if not at BOL and require being after alphanumeric and underscore (i.e. permitted symbol-character):

Result: #s(electric-operator-compiled-rule :operator "" :regex "\([ ]\)\\([ ]\)" :action electric-operator-c-mode-*)

WDYT?

[davidshepherd7]

Have a look at what it currently does, I think it's something similar to that.

…

On Wed, 17 Oct 2018, 09:28 Andreas Röhler, ***@***.***> wrote: Think this regexp might be changed to: space operator only if not at BOL and require being after alphanumeric and underscore (i.e. permitted symbol-character): Result: #s(electric-operator-compiled-rule :operator "*" :regex "\([ ]*\)\*\([ ]*\)" :action electric-operator-c-mode-*) WDYT? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#11 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAwk1OhUBbgO6d38h8JHmnuclLG8RL6sks5uluopgaJpZM4FOUf4> .