Bug 129427 – Some small modernizing tweaks to the coding style guide

• Component: New Bugs
• Status: RESOLVED FIXED

[Darin Adler]

Change coding style guide to specify nullptr instead of 0, and other small modernizing tweaks

[Darin Adler]

Created attachment 225374 [details]
Patch

[Alexey Proskuryakov]

Comment on attachment 225374 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=225374&action=review

> Websites/webkit.org/ChangeLog:14
> +        without non-WebKit-style abbreviations and such. Use auto& instead of const auto& in for loop
> +        example, since that's normally preferred. Changed section title to say "zero" instead of "0".

This is one part I'm not sure of. We have a lot of "const auto&" loop variables currently in the tree, and I personally like those.

[Darin Adler]

(In reply to comment #2)
> This is one part I'm not sure of. We have a lot of "const auto&" loop variables currently in the tree, and I personally like those.

The compiler will automatically propagate the const from the collection if begin/end are overloaded suitably.

How critical is it to narrow the reference to read-only in cases where the access to the collection is not already read-only? Is it worth an extra keyword on every one of these for loops?

[Darin Adler]

Comment on attachment 225374 [details]
Patch

For now I will land with "auto&". We can change back later if we confirm that "const auto&" is preferred.

[Darin Adler]

Committed r164930: <http://trac.webkit.org/changeset/164930>

[Joseph Pecoraro]

Comment on attachment 225374 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=225374&action=review

> Websites/webkit.org/ChangeLog:17
> +        over index iteration in vector iteration example and also showed use of unsigned rather than
> +        size_t since that is almost always what we want. Eliminated the use of PassOwnPtr in the "return

Why is "unsigned rather than size_t" preferred? Shouldn't we match the type that the method / object produces?