lambdas and non-captured, unevaluated operands
in [C++]
|
Prev: [expr.prim.lambda] lambdas and non-captured, unevaluated operands
Next: move only types in standard containers
From: Daniel Krügler on 24 Apr 2010 09:05 On 23 Apr., 15:16, Paul Bibbings <paul.bibbi...(a)gmail.com> wrote: > From the above, my reading is that g5 is OK since the presence of i does > not, in this context, constitute a 'use'. Exactly. > Thus, from [basic.def.odr], > �3.2/2: > > "An expression is potentially evaluated unless it is an unevaluated > operand (Clause 5) or a subexpression thereof. A variable or > non-overloaded function whose name appears as a potentially-evaluated > expression is used ..." > > and, of course, this is the case for the operand to sizeof in the > example above: > > [expr.sizeof] �5.3.3/1: > > "The sizeof operator yields the number of bytes in the object > representation of its operand. The operand is either an expression, > which is an unevaluated operand (Clause 5), or a parenthesized > type-id." > > What appears to be missing in [expr.prim.lambda] (unless it is me that > is missing something) is some statement indicating how it happens - in > the absence of explicit or implicit capture - that i is nevertheless > `available' in the context of the compound-statement of the > lambda-expression in g5, above; in effect, by what mechanism it is > available outside of being captured. What you are missing is [expr.prim.lambda]/7: "The lambda-expression's compound-statement yields the function-body (8.4) of the function call operator, but for purposes of name lookup (3.4), [..], the compound-statement is considered in the context of the lambda- expression." So this defines the name lookup of any entity named in the lambda-expression. It means that the name 'i' within g5 is resolved in the same way as this name would be resolved at the textual position where the lambda expression occurred, i.e. void f2() { int i = 1; // ... void g5(int = sizeof i); } This is also well-formed, because i is not used, since it appears in an unevaluated operand (I'm aware that several compilers don't get that right). > I am wanting to get a better sense of this since gcc-4.5.0 does not > accept the above code, failing it with: > > 11:53:03 Paul Bibbings(a)JIJOU > /cygdrive/d/CPPProjects/nano/gcc_bugs $cat _5_1_2_13.cpp > // file: _5_1_2_13.cpp > > void f2() { > int i = 1; > // ... > void g5(int = ([]{ return sizeof i; })()); > } > > 11:53:06 Paul Bibbings(a)JIJOU > /cygdrive/d/CPPProjects/nano/gcc_bugs $i686-pc-cygwin-gcc-4.5.0 -Wall > -Wno-unused -std=c++0x -pedantic -c _5_1_2_13.cpp > _5_1_2_13.cpp: In lambda function: > _5_1_2_13.cpp:6:37: error: local variable �i� may not appear in this > context > _5_1_2_13.cpp:6:40: error: return-statement with a value, in function > returning 'void' > _5_1_2_13.cpp: In function �void f2()�: > _5_1_2_13.cpp:6:44: error: default argument for �int <anonymous>� has > type �void� This is a compiler bug. HTH & Greetings from Bremen, Daniel Kr�gler -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] |