[basic.lookup.unqual] Clarify conversion-function-id components lookup #5191
Conversation
|
@opensdh, your territory. |
languagelawyer
force-pushed
the
conversion-function-id-lookup
branch
from
54b0561 to
860c700
Compare
|
@opensdh: What's your verdict -- is this an improvement, and, importantly, is this still editorial? |
| If $T$ appears in a \grammarterm{conversion-function-id} that is a qualified name | ||
| whose lookup context\iref{basic.lookup.qual} is a class $S$, | ||
| lookup for $U$ performs a search in the scope associated with $S$; | ||
| if that lookup finds nothing, $U$ undergoes unqualified name lookup. |
There was a problem hiding this comment.
I think it would be clearer to say "[…]; otherwise, or if that lookup finds nothing, […]". It's true that the general "Unless otherwise specified" from the previous paragraph ought to apply, but it tends to sound like the alternative is that the name is not looked up at all.
There was a problem hiding this comment.
It's true that the general "Unless otherwise specified" from the previous paragraph ought to apply, but it tends to sound like the alternative is that the name is not looked up at all.
To me, the previous paragraph clearly applies to unqualified \grammarterm{conversion-function-id} :/
May I not apply the proposed change, please?
There was a problem hiding this comment.
This paragraph is about an unqualified U, looked up separately from the conversion-function-id that contains it. I want to avoid the (nonsensical) suggestion that, if T does not appear in the given context, U is not looked up at all.
There was a problem hiding this comment.
This paragraph is about an unqualified U, looked up separately from the conversion-function-id that contains it.
Yeah, right, but whatever, the previous paragraph applies to other unqualified names (which are not mentioned in this paragraphs) (and in any other paragraphs-exceptions).
I want to avoid the (nonsensical) suggestion that, if T does not appear in the given context, U is not looked up at all.
I don't see how is it possible. I mean, not only the previous paragraph, but also the previous sentence suggest that such U s are looked up:
Lookup for an unqualified name U that is ... of a conversion-type-id T considers only names that ... If T appears in a conversion-function-id ...
If U is not looked up when T doesn't appear in conversion-function-id, there is no reason these sentences are in this and not the opposite order
There was a problem hiding this comment.
I don't insist on a change; it just strikes me as a trip hazard for the reader to say "If X, then Y; if Y does nothing, Z." and have Z also happen when X doesn't obtain.
|
I didn't know I issued verdicts. I think this is definitely an improvement (and I made my one further comment about it); it's certainly not strictly editorial, although one could argue that it merely stipulates the obvious thing in a place where the existing text does not specify any behavior. It might fall into the "CWG glances at a merge request during a teleconference" category. |
|
Thanks, that's precisely what I needed to hear! @jensmaurer: could you kindly ask the CWG chair to clear this proposed change with his people? |
|
There is still a request for further adjustment by @opensdh. Will forward to the CWG chair after that is fixed. |
jensmaurer
added
the
changes requested
What does
looked up in the same fashion as the \grammarterm{conversion-function-id}mean when the\grammarterm{conversion-function-id}is not looked up? Like in declarations.