Bug 6798 – Integrate overloadings for multidimentional indexing and slicing

The following replacement has been proposed by issue 3474. x[$-2, y[$-6, $-9], $-2] // is translated to // x.opIndex(x.opDollar!0 - 2, // y.opIndex(y.opDollar!0 - 6, y.opDollar!1 - 9), // x.opDollar!2 - 2) Similarly, I propose the following replacement for the overloading mixture of indexing and slicing. x[2..$, $-1, y[$-6, 0..$]] // is translated to // x.opIndex(x.opSlice!0(2, x.opDollar!0), // x.opDollar!1 - 1, // y.opIndex(y.opDollar!0 - 6, y.opSlice!1(0, y.opDollar!1))) And additionally, I propose to merge slicing into indexing. x[] // is translated to // x.opIndex() After all, the indexing and slicing getter will be translated to x.opIndex(...), and the setter will be translated to x.opIndexAssign(value, ...). And also, opSlice and opSliceAssign would be unnecessary.

So is this: y[$-6, 0..$:2] getting translated like this? y.opIndex(y.opDollar!0 - 6, y.opSlice!1(0, y.opDollar!1, 2)) Strides/steps are sometimes useful in numerics code.

Posted pull request. https://github.com/D-Programming-Language/dmd/pull/443 (This does not contains stride syntax/overloading.)

There was a discussion thread about the idea of supporting a syntax like: y[$-6, 0..$:2] Don: http://www.digitalmars.com/webnews/newsgroups.php?art_group=digitalmars.D&article_id=146464 And in my opinion one of the last messages in that thread, by Robert Jacques, seals the topic. I think nice stride syntax will be useful for user-defined types: http://www.digitalmars.com/webnews/newsgroups.php?art_group=digitalmars.D&article_id=146533 Do you want me to open another enhancement request that asks just for the stride syntax?

What is the state of this issue and the pull?

https://github.com/D-Programming-Language/dmd/pull/443 has been merged finally. Hooray!