assignment | Assignments and Graders for Discrete Optimization | Machine Learning library

This is currently a known bug in Rakudo. The intended behavior is for has to support list assignment, which would make syntax very much like that shown in the question work.

I am not sure if the supported syntax will be:

This section discusses the specific statements from your sample functions.
See the bottom section for background information.

• $n = 1 and $n++ are assignments and therefore do not produce output.
• $n is an expression whose value is output
• $null - ditto, but even though it is output, it doesn't display by default
• ($n++) - due to enclosure in (...) - turns the assignment into an expression and therefore does output the assigned value (too).
  • However, because you've used the post-increment form of the assignment, it is the old value that is output, not the now incremented value; to increment first (pre-increment) and then output, use (++$n)
• [System.Console]::WriteLine("Hello") prints directly to the console, which bypasses PowerShell's system of output streams.
  • This means you cannot capture or redirect such output from inside PowerShell (see the next section).

^{Tip of the hat to iRon for his help.}

PowerShell, following the model of traditional shells, is organized around streams - see the conceptual about_Redirection help topic for an overview of all 6 streams that PowerShell supports.^[1]

That is, any statement - and therefore potentially multiple ones - in a script or function can write to any of the output streams.

The primary output stream, meant to convey data, is the success output stream (whose number is 1), and only it is sent through the pipeline by default, and therefore by default only it is captured in a variable, suppressed, or redirected to a file.

There are two ways to write to the success output stream, i.e. to produce data output:

• Explicitly, with a Write-Output call - although that is rarely needed.

• Typically implicitly, by neither capturing, suppressing, nor redirecting output produced by a statement.

  • In other words: Output from any command (e.g., Get-ChildItem *.txt) or expression (e.g, 1 + 2 or (42).ToString('x')) is sent to the success output stream by default.

  • Unlike in traditional programming languages, return is not needed to produce output - in fact, its primary purpose is to exit the enclosing scope independently of any output the scope produces, though as a syntactic convenience you can combine the two aspects:

    • return is in effect the same as the following two statements, the first of which (potentially) produces output, the second of which exits the scope: ; return

  • This implicit output behavior is convenient and and allows for concise, expressive code, but can also be a pitfall: it is easy to accidentally produce output - typically from a .NET method whose return value isn't needed (see this question for an example).

    • iRon's GitHub feature request #15781 discusses one potential way to remedy this problem: introduction of an opt-in strict mode that only permits using explicit output statements (Write-Output, return) in order to produce output.

    • This answer shows troubleshooting techniques you can use with the currently available features.

As for assignments - e.g. $n = 1; $n += 1; ++$n; $n--:

• By default they do not produce output.
  • A hybrid case is the chaining form of a multi-assignment, e.g. $a = $b = 1, which assigns 1 to both variables: statement-internally the assignment value is passed through, but the statement as a whole has no output.
• However, as an opt-in you can make them pass the value(s) being assigned through via (...), the grouping operator; e.g. ($n = 1) both assigns 1 to variable $n and outputs 1, which allows it to participate in larger expressions, such as ($n = 1) -gt 0
  • Note that the related $(...) (subexpression operator) and @(...) (array-subexpression operator) do not have that effect - they wrap one or more entire statement(s), without affecting the enclosed statements' intrinsic output behavior; e.g. $($n = 1) does not produce output, because $n = 1 by itself doesn't produce output; however, $(($n = 1)) does, because ($n = 1) by itself does.

As for output enumeration behavior:

• By default, PowerShell enumerates collections that are being output, in the spirit of streaming output: That is, it sends a collection's elements to the pipeline, one by one.

• In the rare event that you do need to output a collection as a whole - which in general should be avoided, so as not to confound other commands participating in a pipeline, which usually do expect object-by-object input - you have two options:

  • , $collection (sic; uses an aux. one-element wrapper array)
  • More explicitly, but less efficiently: Write-Output -NoEnumerate $collection
  • See this answer for more information.

As for outputting $null:

• $null is output to the pipeline, but by default doesn't show.

  • $null by itself produces no visible output,

  • but the following returns $true, demonstrating that the value was sent:

The evaluation order of A = B was not specified before c++17, after c++17 B is guaranteed to be evaluated before A, see https://en.cppreference.com/w/cpp/language/eval_order rule 20.

The behaviour of valMap[val] = valMap.size(); is therefore unspecified in c++14, you should use:

there could be at least one difference - case when you need to pass variable to some other function, for example:

To the internal content of an optional, doesn't the optional require placement new in order to reconstruct the internal in place storage or union?

For assignment, yes it does.

But while we still cannot do actual placement new during constexpr time, we did get a workaround for its absence: std::construct_at (from P0784). This is a very limited form of placement new, but it's enough to get optional assignment working.

The other change was that we also needed to be able to actually change the active member of a union - since it wouldn't matter if we could construct the new object if we couldn't actually switch. That also happened in C++20 (P1330).

Put those together, and you get a functional implementation of optional assignment: P2231. An abbreviated implementation would look like this:

The object() class is like a fundamental particle of the python universe, and is the base class (or building block) for all objects (read everything) in Python. As such, the stated behavior is logical, for not all objects can (or should) have arbitrary attributes set. For example, it wouldn't make sense if a NoneType object could have attributes set, and, just like object(), a None object also does not have a __dict__ attribute. In fact, the only difference in the two is that a None object has a __bool__ attribute. For example:

GCC and Clang report that S is trivially copyable in C++11 through C++23 standard modes. MSVC reports that S is not trivially copyable in C++14 through C++20 standard modes.

N3337 (~ C++11) and N4140 (~ C++14) say:

A trivially copyable class is a class that:

• has no non-trivial copy constructors,
• has no non-trivial move constructors,
• has no non-trivial copy assignment operators,
• has no non-trivial move assignment operators, and
• has a trivial destructor.

By this definition, S is trivially copyable.

N4659 (~ C++17) says:

A trivially copyable class is a class:

• where each copy constructor, move constructor, copy assignment operator, and move assignment operator is either deleted or trivial,
• that has at least one non-deleted copy constructor, move constructor, copy assignment operator, or move assignment operator, and
• that has a trivial, non-deleted destructor

By this definition, S is not trivially copyable.

N4860 (~ C++20) says:

A trivially copyable class is a class:

• that has at least one eligible copy constructor, move constructor, copy assignment operator, or move assignment operator,
• where each eligible copy constructor, move constructor, copy assignment operator, and move assignment operator is trivial, and
• that has a trivial, non-deleted destructor.

By this definition, S is not trivially copyable.

Thus, as published, S was trivally copyable in C++11 and C++14, but not in C++17 and C++20.

The change was adopted from DR 1734 in February 2016. Implementors generally treat DRs as though they apply to all prior language standards by convention. Thus, by the published standard for C++11 and C++14, S was trivially copyable, and by convention, newer compiler versions might choose to treat S as not trivially copyable in C++11 and C++14 modes. Thus, all compilers could be said to be correct for C++11 and C++14.

For C++17 and beyond, S is unambiguously not trivially copyable so GCC and Clang are incorrect. This is GCC bug #96288 and LLVM bug #39050

It looks like strictly speaking, this doesn't conform with the grammar as specified. But at least, it does make sense for a compiler to allow it: the reason PrimaryNoNewArray is used instead of Primary here is so that an expression like new int[1][0] cannot be ambiguously parsed as either a 2D array or as an array access like (new int[1])[0]. If the array has an initializer like new int[]{0}[0] then the syntax is not ambiguous because this cannot be parsed as creating a 2D array.

That said, since the JLS doesn't specify that it's allowed, it could be treated as either an implementation detail or a bug in the compiler which allows it.

Somehow, answering the questions parts in the opposite order felt more natural to me. :-)

Second, does auto-flattening allow any behavior that would be impossible if the left hand side were non-flattening?

It's relatively common to want to assign the first (or first few) items of a list into scalars and have the rest placed into an array. List assignment descending into iterables on the left is what makes this work: