P2223 Trimming whitespaces before line splicing

[wg21bot]

P2223R0 Trimming whitespaces before line splicing (Corentin Jabot)

[jfbastien]

Seen in today's EWG telecon.

P2223R0 make trailing whitespace after an end-of-line backslash non-significant (note: this leaves handling of raw string literals untouched).

SF	F	N	A	SA
3	9	1	0	0

P2223R0 fix CWG1698 as proposed in the paper.

SF	F	N	A	SA
0	4	3	1	2

Corentin will update the paper, and then we'll vote on sending it to CWG.

[wg21bot]

P2223R1 Trimming whitespaces before line splicing (Corentin Jabot)

[jfbastien]

EWG February 2021 polls:

Poll: Forward P2223R1 "Trimming whitespaces before line splicing" to Core, thereby also fixing [CWG1698].

SF	F	N	A	SA
11	15	4	0	0

Poll outcome: ✅ consensus.

Salient comments:

• Strongly in favor as this aligns the standard with common practice, avoids contorting standard wording by implementations in order to claim conformance while producing the behavior programmers want, and avoids needless surprised and implementation divergence.
• We should not have implementation divergence on this, and this direction to resolve it is good.
• I’m not particularly excited about this paper because it only partially solves the problem of implementation divergence in this space and does not outline a play to solve all the existing issues in a consistent way. This will not likely impact my users.
• Invisible differences in the behavior of code are bad.
• Reducing divergence in compilers is good.
• I find this somewhat unimportant.