And of course there's Strachey's Fundamental Concepts in Programming Languages (1967), which introduced the LValue/RValue distinction. It is very readable and relevant today.
b) a jupyter notebook (which also uses markdown blocks). jupyter can embed C++, so you can have interactive blocks that you can evaluate and play with.
I looked for a number of options and I didn't find anything which could provide me with automatic syntax highlighting, non-html layouts, and quick iteration. That doesn't mean that such a thing doesn't exist!
Jupyter notebooks are probably what you are looking for. I highly recommend them. Embedded REPL makes iteration very fast and results can be shared directly through a URL, exported to various formats or shared in websites that support it such as github.
Because `i` isn't an int, it's an initializer_list, and you almost never want to actually create an initializer_list instead of an int? Seems pretty understandable to me.
I think the biggest issue in C++ is that you don't have to understand this whole mess to be productive in it. This is also perhaps its greatest strength. Where you have Rust which won't even compile a basic toy program without demonstrating a full understanding of its Ownership model.
Which is it? C++ programmers are smart enough to not need the features that Rust provides? Or C++ programmers aren't smart enough to learn basic concepts that they should already know considering their compiler doesn't help them write correct code?
No, that post refers to the lifetime of temporaries held by local reference variables. My question is about temporaries created to pass reference parameters to a function. Is the lifetime of that temporary the BODY of the function? The full expression that calls the function? The enclosing scope of that full expression? (for the record, VS2017 does the latter)
If the lifetime of the temporary holding the value 3 can (implementation-defined) be limited to the body of the function, the assignment to r happens when the temporary can (implementation-defined) be destroyed already.
This is not a problem. The temporary ceases to exist after the function call expression. Copying the return value has to be done before that, otherwise you could never return the value of any local variable.
That's what I had always thought would happen; it makes sense. However, if I'm reading things correctly, the posted article disagrees, and this link also disagrees:
"It is implementation-defined whether the lifetime of a parameter ends when the function in which it is defined returns or at the end of the enclosing full-expression."
Hence why I asked for some substantive clarification or solid interpretation of the standard. "Common sense" doesn't cut it with C++. :)
I'm not sure where you see a disagreement with section 5.2.2/4 of the standard:
"It is implementation-defined whether the lifetime of a parameter ends when the function in which it is defined returns or at the end of the enclosing full-expression."
Regardless of where exactly the lifetime of the temporary ends (end of function or end of expression), your example would be correct and not UB.
I don't think it is in dispute whether or not the return expression inside the function is evaluated before the lifetime of the parameter ends, but that is when the copy of a into r is made.
Where do you see a disagreement with the posted article?
If the lifetime of the temporary ended with the function body, the assignment (which reads the temporary via the returned reference) happens afterwards and that's UB. Essentially,
int const &r = f(3);
and
int r = f(3);
would both be UB under that implementation-defined case.
The posted article, second row of "common mistakes", describes returning a const reference parameter, when called with a prvalue, as blanket UB.
"6.1 A temporary object bound to a reference parameter in a function call persists until the completion of the full-expression containing the call."
So, while the parameter's (the reference "int const &a") lifetime is implementation-defined, the temporary to which that parameter refers to is the full expression of the call. Sanity is restored.
const int& f(int parameter) { return parameter; }
const int i = f(42);
Is strictly undefined behavior under the simpler C++03 rules/wording of 5.2.2/4: "The lifetime of a parameter ends when the function in which it is defined returns.")
P0135R0 would loosen the requirements on parameter's lifetime as a means of enabling guaranteed copy elision, in such a way that an implementation could technically decide to make this to have defined behavior (not that I suggest relying on it!)
A comment on the document's design: The low contrast of light gray text on yellow background makes it difficult to read. Black text on white, or dark gray text on white is perfectly acceptable.
Another drawback of colored backgrounds is that readers wanting to print out the pdf will consume ink/toner to render the non-white background.
That site is great but it leaves out something else that is very important. Color blind people often have trouble with seeing colors that are "near" to each other in color-space.
I'm not a PhD in Colorology but choosing higher contrast colors completely eliminates that problem I think. It's an easy fix that opens up many new markets to your software.
http://eli.thegreenplace.net/2011/12/15/understanding-lvalue...