public void DistinctInputsAndOutputsAreAggregated()
{
var predictors = new IProjectStaticPredictor[]
{
new MockPredictor(new StaticPredictions(
new[] { new BuildInput(@"foo\bar1", false) },
new[] { new BuildOutputDirectory(@"blah\boo1") })),
new MockPredictor2(new StaticPredictions(
new[] { new BuildInput(@"foo\bar2", false) },
new[] { new BuildOutputDirectory(@"blah\boo2") })),
};
var executor = new ProjectStaticPredictionExecutor(@"c:\repo", predictors);
var project = TestHelpers.CreateProjectFromRootElement(ProjectRootElement.Create());
StaticPredictions predictions = executor.PredictInputsAndOutputs(project);
BuildInput[] expectedInputs =
{
new BuildInput(@"foo\bar1", false, "MockPredictor"),
new BuildInput(@"foo\bar2", false, "MockPredictor2"),
};
BuildOutputDirectory[] expectedBuildOutputDirectories =
{
new BuildOutputDirectory(@"blah\boo1", "MockPredictor"),
new BuildOutputDirectory(@"blah\boo2", "MockPredictor2"),
};
predictions.AssertPredictions(expectedInputs, expectedBuildOutputDirectories);
}
public void PredictorSparsenessStressTest()
{
int[] numPredictorCases = { 40 }; // Set to 1000 or more for reduced average noise when using tickResults for measurements.
int[] sparsenessPercentages = { 0, 25, 50, 75, 100 };
var tickResults = new long[numPredictorCases.Length][];
var proj = TestHelpers.CreateProjectFromRootElement(ProjectRootElement.Create());
// Run through twice and keep the second round only - first round affected by JIT overhead.
for (int iter = 0; iter < 2; iter++)
{
for (int p = 0; p < numPredictorCases.Length; p++)
{
int numPredictors = numPredictorCases[p];
tickResults[p] = new long[sparsenessPercentages.Length];
var predictors = new IProjectStaticPredictor[numPredictors];
int sparseIndex = 0;
for (int s = 0; s < sparsenessPercentages.Length; s++)
{
int sparsenessPercentage = sparsenessPercentages[s];
for (int i = 0; i < numPredictors; i++)
{
if (sparseIndex < sparsenessPercentage)
{
predictors[i] = new MockPredictor(null);
}
else
{
predictors[i] = new MockPredictor(new StaticPredictions(null, null));
}
sparseIndex++;
if (sparseIndex > sparsenessPercentage)
{
sparseIndex = 0;
}
}
var executor = new ProjectStaticPredictionExecutor(@"c:\repo", predictors);
Stopwatch sw = Stopwatch.StartNew();
executor.PredictInputsAndOutputs(proj);
sw.Stop();
tickResults[p][s] = sw.ElapsedTicks;
Console.WriteLine($"{numPredictors} @ {sparsenessPercentage}%: {sw.ElapsedTicks} ticks");
}
}
}
}
public void EmptyPredictionsResultInEmptyAggregateResult()
{
var predictors = new IProjectStaticPredictor[]
{
new MockPredictor(new StaticPredictions(null, null)),
new MockPredictor(new StaticPredictions(null, null)),
};
var executor = new ProjectStaticPredictionExecutor(@"c:\repo", predictors);
var project = TestHelpers.CreateProjectFromRootElement(ProjectRootElement.Create());
StaticPredictions predictions = executor.PredictInputsAndOutputs(project);
Assert.NotNull(predictions);
Assert.Equal(0, predictions.BuildInputs.Count);
Assert.Equal(0, predictions.BuildOutputDirectories.Count);
}
public void DuplicateInputsAndOutputsMergePredictedBys()
{
var predictors = new IProjectStaticPredictor[]
{
new MockPredictor(new StaticPredictions(
new[] { new BuildInput(@"foo\bar", false) },
new[] { new BuildOutputDirectory(@"blah\boo") })),
new MockPredictor2(new StaticPredictions(
new[] { new BuildInput(@"foo\bar", false) },
new[] { new BuildOutputDirectory(@"blah\boo") })),
};
var executor = new ProjectStaticPredictionExecutor(@"c:\repo", predictors);
var project = TestHelpers.CreateProjectFromRootElement(ProjectRootElement.Create());
StaticPredictions predictions = executor.PredictInputsAndOutputs(project);
predictions.AssertPredictions(
new[] { new BuildInput(@"foo\bar", false, "MockPredictor", "MockPredictor2") },
new[] { new BuildOutputDirectory(@"blah\boo", "MockPredictor", "MockPredictor2") });
}