[cpp.predefined] Use of "an" broadens condition too much

[JohelEGP]

Originating comment: #5670 (comment).

• STDCPP_BFLOAT16_T
  Defined as the integer literal 1 if and only if the implementation supports an extended floating-point type with the properties described in [basic.extended.fp].

The use of "an" suggests that if any other "extended floating-point type with the properties described in [basic.extended.fp]" other than std::float16_t is defined by the implementation, then __STDCPP_BFLOAT16_T__ should be defined to 1 even if std::float16_t isn't.

[dkolsen-pgi]

You are right. I'm not sure how I messed that one up. There must have been some context that got lost when the wording was moved to [cpp.predefined].

I suggest:

Defined as the integer literal 1 if and only if the implementation supports an extended floating-point type with the properties of the typedef-name std::bfloat16_t as described in [basic.extended.fp].

[tkoeppe]

@jensmaurer Do you consider fixing that editorial or would you prefer CWG/EWG oversight?

[jensmaurer]

This looks editorial to me.