Functional tests

[jim22k]

This only converts 2 lowering tests to functional tests, but shows the pattern.

Before converting more, we should upgrade graphblas.print to display a sparse tensor in a consistent format. Or possibly add a new operation graphblas.print_tensor with options to control how much is displayed:

• Values only
• Pointers, Indices, Values
• Shape, Rev, Pointers, Indices, Values

[coveralls]

Pull Request Test Coverage Report for Build 1392205097

• 0 of 0 changed or added relevant lines in 0 files are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage decreased (-2.9%) to 82.055%

---

Totals
Change from base Build 1387293998:	-2.9%
Covered Lines:	3219
Relevant Lines:	3923

---

💛 - Coveralls

[seibert]

This looks like a fine approach for now, but I think we may want to make graphblas-exec an actual Python executable that we can run outside of llvm-lit at the command line. That will make it easier to debug test issues in the long run.