Bug 185135 – Use some C++17 features

• Status: RESOLVED
• Resolution: FIXED
• Component: Tools / Tests

[JF Bastien]

As discussed here [0] let's move WebKit to a subset of C++17. We now require GCC 6 [1] which means that, according to [2] we can use the following C++17 language features (I removed some uninteresting ones):

 - New auto rules for direct-list-initialization		
 - static_assert with no message
 - typename in a template template parameter
 - Nested namespace definition
 - Attributes for namespaces and enumerators
 - u8 character literals
 - Allow constant evaluation for all non-type template arguments
 - Fold Expressions
 - Unary fold expressions and empty parameter packs
 - __has_include in preprocessor conditional
 - Differing begin and end types in range-based for
 - Improving std::pair and std::tuple

Consult the Tony Tables [3] to see before / after examples.

Of course we can use any library feature if we're willing to import them to WTF (and they don't require language support).


  [0]: https://lists.webkit.org/pipermail/webkit-dev/2018-March/029922.html
  [1]: https://trac.webkit.org/changeset/231152/webkit
  [2]: https://en.cppreference.com/w/cpp/compiler_support
  [3]: https://github.com/tvaneerd/cpp17_in_TTs/blob/master/ALL_IN_ONE.md

[JF Bastien]

Created attachment 339119 [details]
patch

[JF Bastien]

This built in debug without any troubles for me. I must have done something wrong, let's see what the bots say.

[JF Bastien]

Created attachment 339122 [details]
patch

Fix libstdc++ build.

[JF Bastien]

Created attachment 339124 [details]
patch

Fix feature test macro for libstdc++.

[JF Bastien]

Created attachment 339129 [details]
patch

[Darin Adler]

Comment on attachment 339129 [details]
patch

View in context: https://bugs.webkit.org/attachment.cgi?id=339129&action=review

> Source/WebCore/platform/URLParser.cpp:2500
> +    URLParser::LCharBuffer output;

Why do we need a URLParser:: prefix in a URLParser member function template?

[JF Bastien]

Created attachment 339144 [details]
patch

I'm building for ARM locally, I'm not sure I get why this error occurs, seems unrelated.


(In reply to Darin Adler from comment #6)
> Comment on attachment 339129 [details]
> patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=339129&action=review
> 
> > Source/WebCore/platform/URLParser.cpp:2500
> > +    URLParser::LCharBuffer output;
> 
> Why do we need a URLParser:: prefix in a URLParser member function template?

We don't, I was over-enthusiastic in find-and-replace. It's now gone.

[JF Bastien]

(In reply to JF Bastien from comment #7)
> Created attachment 339144 [details]
> patch
> 
> I'm building for ARM locally, I'm not sure I get why this error occurs,
> seems unrelated.

My local build works just fine, and looking at the code I just don't see anything wrong. I'll assume that ToT was borked for a short bit.

[WebKit Commit Bot]

Comment on attachment 339144 [details]
patch

Clearing flags on attachment: 339144

Committed r231170: <https://trac.webkit.org/changeset/231170>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[Radar WebKit Bug Importer]

<rdar://problem/39848434>

[JF Bastien]

I'll move the main Xcode code configurations to C++17 in a follow-up, it'll require a bit more code change it looks like.