Misleading error message when datamanager fails to load refg

[SophiaPerzan-DG]

When a datamanager fails to resolve the datafile path for an assembly or annotation, the following error message is printed:
"Assembly is missing its chrom.sizes or annotation is missing its cfg file:

This is misleading because the stack trace from the datamanager failure isn't printed.

[SophiaPerzan-DG]

After discussion with @ovesh the strategy for resolving this is as follows:

When an error occurs in getting the reference genome for an assembly or annotation, there are 2 failure cases.

1. The assembly or annotation is malformed (missing a chrom.sizes or .cfg).
   • The data manager completes its request to find these files successfully in this case, the files are just not found.
2. The data manager attempts to look for a chrom.sizes or .cfg and fails (permissions errors, network errors, etc.)

Currently, GK treats both of these failures as the 1st case. What we want to do is explicitly handle the 2nd case by printing the python stack trace when an unexpected error is thrown by the data manager.

To do this, the data manager needs to be altered to raise a new/custom FileNotFound exception (or something similar). This exception would denote the request to find the file failed, but did so not as the result of some exception (a successful failure essentially).

Using that exception, we could differentiate the cases, and handle the errors appropriately.

[SophiaPerzan-DG]

Part of this change would involve changing how the resolve_datafile_path_with_python function within py_init.cpp is handeled. We don't want to always raise a value error if the path fails to be found.