Writing Julia Objects to pickle and load from python

[chengchingwen]

To make Julia Object loadable from python, there are three issues:

1. For simple Julia struct (no functions): utilities the stack machine to build a similar python class and load data into it. Will need a rule-base translation for the struct definition to python class definition string.

2. For Julia Arrays: use the array element type to determine whether to write as python list or numpy array.

3. For complex Julia Object (like Function and struct with functions): currently there is no simple solution, so these will be unsupported.

[findmyway]

For Julia Arrays: use the array element type to determine whether to write as python list or numpy array.

Could it be more specific here? I guess you mean Array{<:Any} -> list, Array{<:Number} -> numpy array?

[chengchingwen]

Yes, I think we can use some rule to determine whether the Array is compatible with numpy array, and if not, fallback to python list.

[chengchingwen]

numpy array also have python object dtype, we probably can use it for Array{Any, N}, but I haven't try it.