Repository for Startup Showdown; see [post][post.md].
TL;DR run ./do.sh results/all to build and run everything; install a redo if
you want to do that more than once.
digraph {
results
sut
tests
analysis
results
test_all[label="redo tests/all"]
results -> test_all
sut -> test_all
test_all -> tests
analyze[label="redo results/all"]
tests -> analyze
analysis -> analyze
analyze -> results
}sut/ has the system-under-test code. Each subdirectory holds a distinct test
case, with a sut.txtpb file describing it to the test executor.
tests/ is the output directory for the tests. redo test/all to run all
tests. This automatically rebuilds the test inputs if needed.
analysis has the code for analyzing the test outputs to extract meaningful
numbers. redo analysis/all rebuilds these tools.
results/ presents the results of the analysis from the test data. redo results/all to perform this analysis. This automatically rebuilds the
analizer if needed, and reruns the tests if needed.