Microsoft.Build.Prediction

This library runs predictors against evaluated MSBuild ProjectInstance to predict file and directory inputs that will be read, and output directories that will be written, by the project.

Predictors are implementations of the IProjectPredictor interface. Execution logic in this library applies the predictors in parallel to a given Project. The library aggregates results from all predictors into a final set of predicted inputs and outputs for a Project.

Input and output predictions produced here can be used, for instance, for Project build caching and sandboxing. Predicted inputs are added to the project file itself and known global files and folders from SDKs and tools to produce a set of files and folders that can be hashed and summarized to produce an inputs hash that can be used to look up the results of previous build executions. The more accurate and complete the predicted set of inputs, the narrower the set of cached output results, and the better the cache performance. Predicted build output directories are used to guide static construction and analysis of build sandboxes.

Usage

Basic usage:

List<IProjectPredictor> predictors = new List<IProjectPredictor>();
predictors.AddRange(ProjectPredictors.AllPredictors);
// Add any custom IProjectPredictor implementations

var predictionExecutor = new ProjectPredictionExecutor(predictors);

ProjectInstance projectInstance = /* Create an MSBuild ProjectInstance */;
ProjectPredictions predictions = predictionExecutor.PredictInputsAndOutputs(projectInstance);

Providing a custom IProjectPredictionCollector:

IReadOnlyCollection<IProjectPredictor> predictors = ProjectPredictors.AllPredictors;

var predictionExecutor = new ProjectPredictionExecutor(predictors);
var predictionCollector = new CustomProjectPredictionCollector();

ProjectInstance projectInstance = /* Create an MSBuild ProjectInstance */;
predictionExecutor.PredictInputsAndOutputs(projectInstance, predictionCollector);

Using alongside MSBuild's ProjectGraph:

string projectFile = /* Your entry project file */;
ProjectCollection projectCollection = ProjectCollection.GlobalProjectCollection;

// Use a shared evaluation context for all projects.
EvaluationContext evaluationContext = EvaluationContext.Create(EvaluationContext.SharingPolicy.Shared);

var projectGraph = new ProjectGraph(
    projectFile,
    projectCollection
    (string projectPath, Dictionary<string, string> globalProperties, ProjectCollection projCollection) =>
    {
        var projectOptions = new ProjectOptions
        {
            GlobalProperties = globalProperties,
            ProjectCollection = projCollection,
            EvaluationContext = evaluationContext,
        };

        return ProjectInstance.FromFile(projectPath, projectOptions);
    });


ProjectInstance[] projectInstances = projectGraph.ProjectNodes.Select(node => node.ProjectInstance).ToArray();

IReadOnlyCollection<IProjectPredictor> predictors = ProjectPredictors.AllPredictors;

// Using single-threaded prediction since we're parallelizing on project instances instead.
var predictionExecutor = new ProjectPredictionExecutor(predictors, new ProjectPredictionOptions { MaxDegreeOfParallelism = 1 });

var predictionCollector = new CustomProjectPredictionCollector();

Parallel.ForEach(
    projectInstances,
    new ParallelOptions() { MaxDegreeOfParallelism = Environment.ProcessorCount },
    project => predictionExecutor.PredictInputsAndOutputs(project, predictionCollector));

Design

See Design.

Contributing

This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit https://cla.microsoft.com.

When you submit a pull request, a CLA-bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., label, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repos using our CLA.

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.