[llvm-dev] changing variable naming rules in LLVM codebase

[David Greene via llvm-dev]

That's interesting because I have always thought it strange to name members differently
from other variables.  I guess in my mind if a local variable isn't easily identified as such,
it's either declared much too far away from its use (the function is too large, is lacking
proper scoping, whatever) or it is not well-named such as to denote its use.  Note that
I specifically write, "denote its use" and not, "denote its scope."  Of course the poor
naming could go the other way; naming a member "i," for example.

I don't think I've ever come across a naming convention that treats function parameters
specially.  Why?  Arguably they are as different from locals as members are, particularly
when it comes to reference parameters.

Slapping an "m_" in front of poorly-named members isn't really going to help much, any
more than slapping an "l_" in front of local variables would.

That said, I am certainly open to being convinced otherwise.

                                               -David

________________________________________
From: Nemanja Ivanovic <nemanja.i.ibm at gmail.com>
Sent: Thursday, February 14, 2019 7:02:19 AM
To: James Y Knight
Cc: Zachary Turner; David Greene; llvm-dev
Subject: Re: [llvm-dev] changing variable naming rules in LLVM codebase

I have to agree with Paul that I think it is rather useful to have a naming convention that distinguishes class members from locals, etc. I'm not sure what that would look like, whether an m prefix for data members would be something others would entertain, but something that makes it clear would probably be useful. To use Paul's example, I think that mTheStuff vs. TheStuff makes it super easy to visually identify what this is. I imagine this wasn't mentioned in this thread or previously adopted because of some good reason I am not aware of.
A more minor point about underscores vs camel case - what I like about camel case is that it generally keeps my fingers on the 3 rows of the keyboard I use the most. From an ergonomics perspective, I find typing a whole lot of underscores a bit unnatural. So since I find camel case easier to type and equally as readable, I would favour it over underscores.

On Wed, Feb 13, 2019 at 11:03 PM James Y Knight via llvm-dev <llvm-dev at lists.llvm.org<mailto:llvm-dev at lists.llvm.org>> wrote:
There is of course some amount of llvm and clang code which already uses initialLowerCaseNames for variable names too, contrary to the style guide. I don't know how to easily quantify how much.

E.g. ParseGNUAttributes in clang/include/clang/Parse/Parser.h is one I noticed.


On Wed, Feb 13, 2019 at 2:49 PM Zachary Turner via llvm-dev <llvm-dev at lists.llvm.org<mailto:llvm-dev at lists.llvm.org>> wrote:
I want to reiterate the benefit that underscore_names would bring.  To be clear it's not my favorite style, but it does have a very concrete advantage which is that we have a very large subproject already using it.  it doesn't make sense to do a purely aesthetic move that not everyone is going to agree on anyway, when we could do one with actual tangible value.

On Wed, Feb 13, 2019 at 8:52 AM <paul.robinson at sony.com<mailto:paul.robinson at sony.com>> wrote:
Chandler wrote:

> FWIW, I'm pretty strongly opposed to humbleCamelCase. We already use that
> style so something else.

Presumably you are equally opposed to RegularCamelCase, because we already
use *that* style for something else.

But really, objecting on the grounds that a given style is already used for
function names is really a very weak argument.  IME function names are
*incredibly* *hard* to confuse with anything else, because they *always* have
surrounding syntactic context. Given `TheStuff->fooBar().getThingy()` is it
even conceivable that you might not instantly get that fooBar and getThingy
are methods?  Therefore, using the same convention for some other kind of
name is Not Confusing.

OTOH, `TheStuff` comes out of nowhere with no clues to its origin, and *that*
is a barrier to code-reading IME.  Even renaming it to `stuff` would help
approximately zero percent. Parameter? Local? Class member? Global?  LLVM has
incredibly few globals for other reasons, but using the same convention for
locals and class members is a real problem for code-reading, especially code
operating in methods for classes you're not super familiar with.

I acknowledge that the current RFC doesn't propose a member naming convention
different from other variables, but IMO it really ought to.  *That* is the
distinction that would really help in reading unfamiliar code.
--paulr
_______________________________________________
LLVM Developers mailing list
llvm-dev at lists.llvm.org<mailto:llvm-dev at lists.llvm.org>
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
_______________________________________________
LLVM Developers mailing list
llvm-dev at lists.llvm.org<mailto:llvm-dev at lists.llvm.org>
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev