StridedSliceHelper

Return a strided slice from `input`.

The goal of this op is to produce a new tensor with a subset of the elements from the `n` dimensional `input` tensor. The subset is chosen using a sequence of `m` sparse range specifications encoded into the arguments of this function. Note, in some cases `m` could be equal to `n`, but this need not be the case. Each range specification entry can be one of the following:

- An ellipsis (...) using ellipsis(). Ellipses are used to imply zero or more dimensions of full-dimension selection. For example, stridedSlice(foo, Indices.ellipsis() is the identity slice.

- A new axis using newAxis(). This is used to insert a new shape=1 dimension. For example, `stridedSlice(foo, Indices.newAxis()) where foo is shape (3, 4) produces a (1, 3, 4) tensor.

- A range begin:end:stride using slice(Long, Long, long) Index.slice()} or all(). This is used to specify how much to choose from a given dimension. stride can be any integer but 0. begin is an integer which represents the index of the first value to select while end represents the index of the last value to select (exclusive). Begin and end can be null, in which case the index begins or ends at the beginning or end of the dimension, respectively (reversed if stride is negative). When both are null, slice() is the same as all(). The number of values selected in each dimension is end - begin if stride > 0 and begin - end if stride < 0. begin and end can be negative where -1 is the last element, -2 is the second to last. For example, given a shape (3,) tensor stridedSlice(foo, Indices.all()), the effective begin and end are 0 and 3. Do not assume this is equivalent to stridedSlice(foo, Indices.slice(0, -1)) which has an effective begin and end of 0 and 2. Another example is stridedSlice(foo, Indices.slice(-2, null, -1)) which reverses the first dimension of a tensor while dropping the last two (in the original order elements). For example foo = [1,2,3,4]; stridedSlice(foo, Indices.slice(-2, null, -1) is [4,3].

- A single index using at(long). This is used to keep only elements that have a given index. For example (stridedSlice(foo, Indices.at(2)) on a shape (5,6) tensor produces a shape (6,) tensor. The dimension can be kept with size one using at(long, boolean).

These semantics generally follow NumPy's indexing semantics, which can be found here: https://numpy.org/doc/stable/reference/arrays.indexing.html

Requirements: `0 != strides[i] for i in [0, m)` Only one ellipsis.

Assign `value` to the sliced l-value reference of `ref`.

The values of `value` are assigned to the positions in the variable `ref` that are selected by the slice parameters. The slice parameters `begin`, `end`, `strides`, etc. work exactly as in `StridedSlice`.

NOTE this op currently does not support broadcasting and so `value`'s shape must be exactly the shape produced by the slice of `ref`.