import org.gradle.internal.service.scopes.ServiceScope;

import javax.annotation.Nullable;
import java.util.Map;
import java.util.Set;

/**
 * Mediates access across project boundaries.
 */
@ServiceScope(Scope.Build.class)
public interface CrossProjectModelAccess {
    /**
     * Locates the given project relative to some project.
     *

// defined in `native_proto_caster.h` during the resolution process. This is
// because the type casters in `native_proto_caster.h`, which allow seamlessly
// exchanging protobuf messages across c++-python boundaries, potentially
// without serialization, fail in the open-source environment.
// Explicitly-specialized type casters for serialized protobufs are added on an
// on-demand basis for quantization library.

# ADR-0004 - Use a platform-oriented architecture for Gradle

## Date

2024-02-07

## Context

The Gradle code base is essentially a large monolith, without strong internal boundaries.
This has a number of negative effects on productivity, including:

- Unclear ownership of code.
- Difficult to focus on one particular area.
- Unintended coupling between areas of the code, including tests.

## Decision

        then:
        def failure = fails 'classes'
        failure.assertHasDescription "Execution failed for task ':${language.compileTaskName}'."
    }

    def "recompiles dependent classes across project boundaries"() {
        given:
        file("lib/src/main/${language.name}/IPerson.${language.name}") << basicInterface
        file("app/src/main/${language.name}/Person.${language.name}") << classImplementingBasicInterface

     * <p>This lock allows both access to the project state and the right to execute as a task. Acquiring this lock also acquires the lock returned by {@link #getAccessLock()}.
     *
     * <p>When a task reaches across project boundaries, the project state lock is released but the task execution lock is not. This prevents other tasks from the same project from starting but

	configSizeBytes = monitoring.NewDistribution(
		"pilot_xds_config_size_bytes",
		"Distribution of configuration sizes pushed to clients",
		// Important boundaries: 10K, 1M, 4M, 10M, 40M
		// 4M default limit for gRPC, 10M config will start to strain system,
		// 40M is likely upper-bound on config sizes supported.
		[]float64{1, 10000, 1000000, 4000000, 10000000, 40000000},

	r         io.ReadCloser
	decoder   *json.Decoder
	remaining []byte
}

// NewJSONFramedReader returns an io.Reader that will decode individual JSON objects off
// of a wire.
//
// The boundaries between each frame are valid JSON objects. A JSON parsing error will terminate
// the read.
func NewJSONFramedReader(r io.ReadCloser) io.ReadCloser {
	return &jsonFrameReader{
		r:       r,

	// waits up to 5 minutes for the child goroutine to return.
	resultCh := make(chan *result, 1)

	go func() {
		result := &result{}

		// panics don't cross goroutine boundaries, so we have to handle ourselves
		defer func() {
			reason := recover()
			if reason != nil {
				// do not wrap the sentinel ErrAbortHandler panic value
				if reason != http.ErrAbortHandler {

For example:

----
$ gradle build -Dorg.gradle.unsafe.isolated-projects=true
----

When enabled, Gradle will fail the build whenever build logic attempts to cross project boundaries and access the model of another project.
Gradle collects all of these access problems in the configuration cache report, in the same way as other problems.

        expect:
        compare("1.0", "1_0") == 0
        compare("1_0", "1-0") == 0
        compare("1-0", "1+0") == 0
        compare("1.a.2", "1a2") == 0 // number-word and word-number boundaries are considered separators
    }

    def "leading zeros in numeric parts are ignored"() {
        expect:
        compare("01.0", "1.0") == 0
        compare("1.0", "01.0") == 0