engine

type Config struct {
	// Number of concurrent scanner workers,
	// also serves as a multiplier for other worker types (e.g., detector workers, notifier workers)
	Concurrency int

	ConfiguredSources             []sources.ConfiguredSource
	Decoders                      []decoders.Decoder
	Detectors                     []detectors.Detector
	DetectorVerificationOverrides map[config.DetectorID]bool
	IncludeDetectors              string
	ExcludeDetectors              string
	CustomVerifiersOnly           bool
	VerifierEndpoints             map[string]string

	// Verify determines whether the scanner will verify candidate secrets.
	Verify bool

	// Defines which results will be notified by the engine
	// (e.g., verified, unverified, unknown)
	Results               map[string]struct{}
	LogFilteredUnverified bool

	// FilterEntropy filters out unverified results using Shannon entropy.
	FilterEntropy float64
	// FilterUnverified sets the filterUnverified flag on the engine. If set to
	// true, the engine will only return the first unverified result for a chunk for a detector.
	FilterUnverified      bool
	ShouldScanEntireChunk bool

	Dispatcher ResultsDispatcher

	// SourceManager is used to manage the sources and units.
	// TODO (ahrav): Update this comment, i'm dumb and don't really know what else it does.
	SourceManager *sources.SourceManager

	// PrintAvgDetectorTime sets the printAvgDetectorTime flag on the engine. If set to
	// true, the engine will print the average time taken by each detector.
	// This option allows us to measure the time taken for each detector ONLY if
	// the engine is configured to print the results.
	// Calculating the average time taken by each detector is an expensive operation
	// and should be avoided unless specified by the user.
	PrintAvgDetectorTime bool

	// VerificationOverlap determines whether the scanner will attempt to verify candidate secrets
	// that have been detected by multiple detectors.
	// By default, it is set to true.
	VerificationOverlap bool

	// DetectorWorkerMultiplier is used to determine the number of detector workers to spawn.
	DetectorWorkerMultiplier int

	// NotificationWorkerMultiplier is used to determine the number of notification workers to spawn.
	NotificationWorkerMultiplier int

	// VerificationOverlapWorkerMultiplier is used to determine the number of verification overlap workers to spawn.
	VerificationOverlapWorkerMultiplier int

	VerificationResultCache  verificationcache.ResultCache
	VerificationCacheMetrics verificationcache.MetricsReporter
}

type Engine struct {

	// ahoCorasickHandler manages the Aho-Corasick trie and related keyword lookups.
	AhoCorasickCore *ahocorasick.Core

	WgNotifier sync.WaitGroup
	// contains filtered or unexported fields
}

type HuggingfaceConfig struct {
	Endpoint           string
	Models             []string
	Spaces             []string
	Datasets           []string
	Organizations      []string
	Users              []string
	IncludeModels      []string
	IgnoreModels       []string
	IncludeSpaces      []string
	IgnoreSpaces       []string
	IncludeDatasets    []string
	IgnoreDatasets     []string
	SkipAllModels      bool
	SkipAllSpaces      bool
	SkipAllDatasets    bool
	IncludeDiscussions bool
	IncludePrs         bool
	Token              string
	Concurrency        int
}

type JenkinsConfig struct {
	Endpoint              string
	Username              string
	Password              string
	Header                string
	InsecureSkipVerifyTLS bool
}

type Metrics struct {
	BytesScanned           uint64
	ChunksScanned          uint64
	VerifiedSecretsFound   uint64
	UnverifiedSecretsFound uint64
	AvgDetectorTime        map[string]time.Duration

	ScanDuration time.Duration
	// contains filtered or unexported fields
}

type Printer interface {
	Print(ctx context.Context, r *detectors.ResultWithMetadata) error
}

type PrinterDispatcher struct {
	// contains filtered or unexported fields
}

type ResultsDispatcher interface {
	Dispatch(ctx context.Context, result detectors.ResultWithMetadata) error
}