Bug 9047 – Expression requiring std.math fails with function-local import

void main() { import std.math; auto f = (double a, double b) => a ^^ b; } DMD 2.061alpha gives: test.d(3): Error: undefined identifier 'std'

I don't think this is a diagnostic bug, it's a problem where the compiler rewrites the call to: void main() { import std.math; auto f = (double a, double b) => .std.math.pow(a, b); } This is why you get a "undefined identifier std" message. Using the .dot won't work here because the import is function-local.

(In reply to comment #1) > I don't think this is a diagnostic bug, it's a problem where the compiler > rewrites the call to: > > void main() { > import std.math; > auto f = (double a, double b) => .std.math.pow(a, b); > } > > This is why you get a "undefined identifier std" message. > > Using the .dot won't work here because the import is function-local. The code the performs that rewriting should verify that the correctness conditions are satisfied, and otherwise possibly rewrite the ^^ with something that shows a good error message.

*** Issue 4845 has been marked as a duplicate of this issue. ***

(In reply to comment #3) > https://github.com/D-Programming-Language/dmd/pull/1438 Thank you Andrej. The usability of a language comes also from little forward steps like this :-) It's a small step, but hundreds of such steps make a difference.

*** Issue 8354 has been marked as a duplicate of this issue. ***

void main() { import std.stdio; int x = 3; immutable j = 2 ^^ x; } Error received: temp.d(4): Error: undefined identifier 'std' Expected error: temp.d(4): Error: must import std.math to use ^^ operator That problem causes curious cascades of errors like: void main() { import std.stdio; enum size_t N = 5; char[N][N] mat = '0'; int e = 0; foreach (immutable i, ref row; mat) { immutable j = (i + (2 ^^ e)) % N; row[j] = '1'; } writeln(mat); } temp.d(7): Error: undefined identifier 'std' ...\dmd2\src\phobos\std\range.d(611): Error: static assert "Cannot put a const(dchar) into a Appender!(char[])" ...\dmd2\src\phobos\std\format.d(2163): instantiated from here: put!(Appender!(char[]), const(dchar)) ...\dmd2\src\phobos\std\format.d(2210): instantiated from here: formatChar!(Appender!(char[])) ...\dmd2\src\phobos\std\format.d(2113): instantiated from here: formatElement!(LockingTextWriter, char[5u], char) ...\dmd2\src\phobos\std\format.d(1827): ... (5 instantiations, -v to show) ... ...\dmd2\src\phobos\std\stdio.d(1622): instantiated from here: write!(char[5u][5u],char) temp.d(10): instantiated from here: writeln!(char[5u][5u])

(In reply to comment #7) > temp.d(7): Error: undefined identifier 'std' > ...\dmd2\src\phobos\std\range.d(611): Error: static assert "Cannot put a > const(dchar) into a Appender!(char[])" Looks like another case of http://d.puremagic.com/issues/show_bug.cgi?id=9549 which was accidentally fixed by another pull. However your test-case still creates this cascade of errors, so maybe you should put your test-case into Issue 9549 and reopen it. My pull here fixes the first error message (to "must import std.math"), but the cascade of errors still appear.

(In reply to comment #8) > However your test-case still creates this cascade of errors, so maybe you > should put your test-case into Issue 9549 and reopen it. I don't understand Issue 9549 enough, and it was opened by you, so probably I think it's better for you to copy that stuff there :-) > My pull here fixes the first error message (to "must import std.math"), but the > cascade of errors still appear. OK.

*** Issue 8587 has been marked as a duplicate of this issue. ***

Froom 2.065, dmd implicitly imports std.math module in order to resolve PowExpression. So the issue does not occur anymore.