My library guile-newra (1) has quite general multidimensional array slicing. The indices can be linear ranges or arbitrary integer arrays and they can have any rank. You can also use the indexed array as write target. If all the indices are linear ranges then the operation is done without copies. It also has placeholders, so you can skip axes, like : and ... in numpy. guile-newra's arrays are applicative, so you can do (thearray firstindex secondindex ...), without special brackets.
You need wrappers to use it with Guile's native array types, so it may not be as convenient. Or maybe you don't need to handle arbitrary rank, then it might not be the best option. It's also far from finished, although the stuff I mentioned above does work. I didn't adopt Pythons -1 = end convention because in Guile arrays can have any base index, not necessarily 0 (whether that's a good idea or not). (1) https://lloda.github.io/guile-newra/#Slicing <https://lloda.github.io/guile-newra/#Slicing> https://github.com/lloda/guile-newra/ <https://github.com/lloda/guile-newra/> https://notabug.org/lloda/guile-newra <https://notabug.org/lloda/guile-newra>
