On Thu, Sep 12, 2019 at 11:08 PM Joe Perches <j...@perches.com> wrote:
>
> Please name the major projects and then point to their
> .clang-format equivalents.
>
> Also note the size/scope/complexity of the major projects.

Mozilla, WebKit, LLVM and Microsoft. They have their style distributed
with the official clang-format, not sure if they enforce it.

Same for Chromium/Chrome, but it looks like they indeed enforce it:

  "A checkout should give you clang-format to automatically format C++
code. By policy, Clang's formatting of code should always be accepted
in code reviews."

I would bet other Google projects do so as well (since Chandler
Carruth has been giving talks about clang-format for 7+ years). Nick?

I hope those are major enough. There is also precedent in other
languages (e.g. Java, C#, Rust).

> I used the latest one, and quite a bit of the conversion
> was unpleasant to read.

It would be good to see particularly bad snippets to see if we can do
something about them (and, if needed, try to improve clang-format to
support whatever we need).

Did you tweak the parameters with the new ones? I am preparing an RFC
patch for an updated .clang-format configuration that improves quite a
bit the results w.r.t. to the current one (and allows for some leeway
on the developer's side, which helps prevent some cases too).

> Marking sections _no_auto_format_ isn't really a
> good solution is it?

I am thinking about special tables that are hand-crafted or very
complex macros. For those, yes, I think it is a fine solution. That is
why clang-format has that feature to begin with, and you can see an
example in Mozilla's style guide which points here:

  https://github.com/mozilla/gecko-dev/blob/master/xpcom/io/nsEscape.cpp#L22

Cheers,
Miguel

Reply via email to