Skip to content

Find, verify, and analyze leaked credentials

27.2 KB
Raw Blame
package sources

import (
	"encoding/json"
	"errors"
	"runtime"
	"sync"

	"google.golang.org/protobuf/types/known/anypb"

	"github.com/trufflesecurity/trufflehog/v3/pkg/common"
	"github.com/trufflesecurity/trufflehog/v3/pkg/context"
	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/source_metadatapb"
	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/sourcespb"
)

type (
	SourceID int64
	JobID    int64
)

// Chunk contains data to be decoded and scanned along with context on where it came from.
type Chunk struct {
	// Data is the data to decode and scan.
	Data []byte
	// OriginalData holds the pre-decode source data, preserved for secret
	// storage. Set before iterative decoding so it retains the original
	// content even after Data is replaced with decoded forms.
	OriginalData []byte

	// SourceName is the name of the Source that produced the chunk.
	SourceName string
	// SourceID is the ID of the source that the Chunk originated from.
	SourceID SourceID
	// JobID is the ID of the job that the Chunk originated from.
	JobID JobID
	// SecretID is the ID of the secret, if it exists.
	// Only secrets that are being reverified will have a SecretID.
	SecretID int64

	// SourceMetadata holds the context of where the Chunk was found.
	SourceMetadata *source_metadatapb.MetaData
	// SourceType is the type of Source that produced the chunk.
	SourceType sourcespb.SourceType

	// SourceVerify specifies whether this chunk was generated by a source that has verification enabled in its config.
	SourceVerify bool
}

// ChunkingTarget specifies criteria for a targeted chunking process.
// Instead of collecting data indiscriminately, this struct allows the caller
// to specify particular subsets of data they're interested in. This becomes
// especially useful when one needs to verify or recheck specific data points
// without processing the entire dataset.
type ChunkingTarget struct {
	// QueryCriteria represents specific parameters or conditions to target the chunking process.
	QueryCriteria *source_metadatapb.MetaData
	// SecretID is the ID of the secret.
	SecretID int64
}

// Source defines the interface required to implement a source chunker.
type Source interface {
	// Type returns the source type, used for matching against configuration and jobs.
	Type() sourcespb.SourceType
	// SourceID returns the initialized source ID used for tracking relationships in the DB.
	SourceID() SourceID
	// JobID returns the initialized job ID used for tracking relationships in the DB.
	JobID() JobID
	// Init initializes the source. Calling this method more than once is undefined behavior.
	Init(aCtx context.Context, name string, jobId JobID, sourceId SourceID, verify bool, connection *anypb.Any, concurrency int) error
	// Chunks emits data over a channel which is then decoded and scanned for secrets.
	// By default, data is obtained indiscriminately. However, by providing one or more
	// ChunkingTarget parameters, the caller can direct the function to retrieve
	// specific chunks of data. This targeted approach allows for efficient and
	// intentional data processing, beneficial when verifying or rechecking specific data points.
	Chunks(ctx context.Context, chunksChan chan *Chunk, targets ...ChunkingTarget) error
	// GetProgress is the completion progress (percentage) for Scanned Source.
	GetProgress() *Progress
}

// SourceUnitEnumChunker are the two required interfaces to support enumerating
// and chunking of units.
type SourceUnitEnumChunker interface {
	SourceUnitEnumerator
	SourceUnitChunker
}

// SourceUnitUnmarshaller defines an optional interface a Source can implement
// to support units coming from an external source.
type SourceUnitUnmarshaller interface {
	UnmarshalSourceUnit(data []byte) (SourceUnit, error)
}

// SourceUnitEnumerator defines an optional interface a Source can implement to
// support enumerating an initialized Source into SourceUnits.
type SourceUnitEnumerator interface {
	// Enumerate creates 0 or more units from an initialized source,
	// reporting them or any errors to the UnitReporter. This method is
	// synchronous but can be called in a goroutine to support concurrent
	// enumeration and chunking. An error should only be returned from this
	// method in the case of context cancellation, fatal source errors, or
	// errors returned by the reporter. All other errors related to unit
	// enumeration are tracked by the UnitReporter.
	Enumerate(ctx context.Context, reporter UnitReporter) error
}

// ConfiguredSource is a Source with most of its initialization values
// pre-configured from a [sourcespb.LocalSource] configuration struct. It
// exposes a simplified Init() method and can be only initialized once. This
// struct is not necessary for running sources, but it helps simplify gathering
// all of the necessary information to call the [Source.Init] method.
type ConfiguredSource struct {
	Name       string
	Debug      bool // turns on debug logging for this source only
	source     Source
	initParams struct {
		verify      bool
		conn        *anypb.Any
		concurrency int
	}
}

// NewConfiguredSource pre-configures an instantiated Source object with the
// provided protobuf configuration.
func NewConfiguredSource(s Source, config *sourcespb.LocalSource) ConfiguredSource {
	return ConfiguredSource{
		Name:   config.GetName(),
		Debug:  config.GetDebug(),
		source: s,
		initParams: struct {
			verify      bool
			conn        *anypb.Any
			concurrency int
		}{
			verify:      config.GetVerify(),
			conn:        config.GetConnection(),
			concurrency: runtime.NumCPU(),
		},
	}
}

// SourceType exposes the underlying source type.
func (c *ConfiguredSource) SourceType() sourcespb.SourceType {
	return c.source.Type()
}

// Init returns the initialized Source. The ConfiguredSource is unusable after
// calling this method because initializing a [Source] more than once is undefined.
func (c *ConfiguredSource) Init(ctx context.Context, sourceID SourceID, jobID JobID) (Source, error) {
	if c.source == nil {
		return nil, errors.New("source already initialized")
	}
	src := c.source
	err := src.Init(ctx, c.Name, jobID, sourceID, c.initParams.verify, c.initParams.conn, c.initParams.concurrency)
	c.source = nil
	return src, err
}

// BaseUnitReporter is a helper struct that implements the UnitReporter interface
// and includes a JobProgress reference.
type baseUnitReporter struct {
	child    UnitReporter
	progress *JobProgress
}

func (b baseUnitReporter) UnitOk(ctx context.Context, unit SourceUnit) error {
	b.progress.ReportUnit(unit)
	if b.child != nil {
		return b.child.UnitOk(ctx, unit)
	}
	return nil
}

func (b baseUnitReporter) UnitErr(ctx context.Context, err error) error {
	b.progress.ReportError(err)
	if b.child != nil {
		return b.child.UnitErr(ctx, err)
	}
	return nil
}

// UnitReporter defines the interface a source will use to report whether a
// unit was found during enumeration. Either method may be called any number of
// times. Implementors of this interface should allow for concurrent calls.
type UnitReporter interface {
	UnitOk(ctx context.Context, unit SourceUnit) error
	UnitErr(ctx context.Context, err error) error
}

// SourceUnitChunker defines an optional interface a Source can implement to
// support chunking a single SourceUnit.
type SourceUnitChunker interface {
	// ChunkUnit creates 0 or more chunks from a unit, reporting them or
	// any errors to the ChunkReporter. An error should only be returned
	// from this method in the case of context cancellation, fatal source
	// errors, or errors returned by the reporter. All other errors related
	// to unit chunking are tracked by the ChunkReporter.
	ChunkUnit(ctx context.Context, unit SourceUnit, reporter ChunkReporter) error
}

// ChunkReporter defines the interface a source will use to report whether a
// chunk was found during unit chunking. Either method may be called any number
// of times. Implementors of this interface should allow for concurrent calls.
type ChunkReporter interface {
	ChunkOk(ctx context.Context, chunk Chunk) error
	ChunkErr(ctx context.Context, err error) error
}

type SourceUnitKind string

// SourceUnit is an object that represents a Source's unit of work. This is
// used as the output of enumeration, progress reporting, and job distribution.
type SourceUnit interface {
	// SourceUnitID uniquely identifies a source unit. It does not need to
	// be human readable or two-way, however, it should be canonical and
	// stable across runs.
	SourceUnitID() (string, SourceUnitKind)

	// Display is the human readable representation of the SourceUnit.
	Display() string
}

// DockerConfig defines the optional configuration for a Docker source.
type DockerConfig struct {
	// Images is the list of images to scan.
	Images []string
	// BearerToken is the token to use to authenticate with the source.
	BearerToken string
	// UseDockerKeychain determines whether to use the Docker keychain.
	UseDockerKeychain bool
	// ExcludePaths is a list of paths to exclude from scanning.
	ExcludePaths []string
	// Namespace represents a Docker organization or user account.
	Namespace string
	// RegistryToken is an optional authentication token used to access private images within the namespace.
	RegistryToken string
}

// GCSConfig defines the optional configuration for a GCS source.
type GCSConfig struct {
	// CloudCred determines whether to use cloud credentials.
	// This can NOT be used with a secret.
	CloudCred,
	// WithoutAuth is a flag to indicate whether to use authentication.
	WithoutAuth bool
	// ApiKey is the API key to use to authenticate with the source.
	ApiKey,
	// ProjectID is the project ID to use to authenticate with the source.
	ProjectID,
	// ServiceAccount is the service account to use to authenticate with the source.
	ServiceAccount string
	// MaxObjectSize is the maximum object size to scan.
	MaxObjectSize int64
	// Concurrency is the number of concurrent workers to use to scan the source.
	Concurrency int
	// IncludeBuckets is a list of buckets to include in the scan.
	IncludeBuckets,
	// ExcludeBuckets is a list of buckets to exclude from the scan.
	ExcludeBuckets,
	// IncludeObjects is a list of objects to include in the scan.
	IncludeObjects,
	// ExcludeObjects is a list of objects to exclude from the scan.
	ExcludeObjects []string
}

// GitConfig defines the optional configuration for a git source.
type GitConfig struct {
	// HeadRef is the head reference to use to scan from.
	HeadRef string
	// BaseRef is the base reference to use to scan from.
	BaseRef string
	// MaxDepth is the maximum depth to scan the source.
	MaxDepth int
	// Bare is an indicator to handle bare repositories properly.
	Bare bool
	// URI is the URI of the repository to scan. file://, http://, https:// and ssh:// are supported.
	URI string
	// IncludePathsFile is the path to a file containing a list of regexps to include in the scan.
	IncludePathsFile string
	// ExcludePathsFile is the path to a file containing a list of regexps to exclude from the scan.
	ExcludePathsFile string
	// ExcludeGlobs is a list of comma separated globs to exclude from the scan.
	// This differs from the Filter exclusions as ExcludeGlobs is applied at the `git log -p` level
	ExcludeGlobs string
	// SkipBinaries allows skipping binary files from the scan.
	SkipBinaries bool
	// ClonePath is the local path used to clone repositories before scanning.
	ClonePath string
	// NoCleanup allows to keeps cloned repositories in ClonePath after scanning instead of removing them.
	NoCleanup bool
	// PrintLegacyJSON indicates whether to print legacy JSON output format for this source.
	PrintLegacyJSON bool
	// TrustLocalGitConfig allows to trust the local git config.
	TrustLocalGitConfig bool
}

// GithubConfig defines the optional configuration for a github source.
type GithubConfig struct {
	// Endpoint is the endpoint of the source.
	Endpoint string
	// Token is the token to use to authenticate with the source.
	Token string
	// IncludeForks indicates whether to include forks in the scan.
	IncludeForks bool
	// IncludeMembers indicates whether to include members in the scan.
	IncludeMembers bool
	// Concurrency is the number of concurrent workers to use to scan the source.
	Concurrency int
	// Repos is the list of repositories to scan.
	Repos []string
	// Orgs is the list of organizations to scan.
	Orgs []string
	// ExcludeRepos is a list of repositories to exclude from an org/user scan.
	// It only filters repos enumerated via Orgs; it has no effect when Repos is set
	// explicitly, since explicit repos bypass the filtering step entirely.
	ExcludeRepos []string
	// IncludeRepos is a list of repositories to include in an org/user scan.
	// It only filters repos enumerated via Orgs; it has no effect when Repos is set
	// explicitly, since explicit repos bypass the filtering step entirely.
	IncludeRepos []string
	// Filter is the filter to use to scan the source.
	Filter *common.Filter
	// IncludeIssueComments indicates whether to include GitHub issue comments in the scan.
	IncludeIssueComments bool
	// IncludePullRequestComments indicates whether to include GitHub pull request comments in the scan.
	IncludePullRequestComments bool
	// IncludeGistComments indicates whether to include GitHub gist comments in the scan.
	IncludeGistComments bool
	// SkipBinaries allows skipping binary files from the scan.
	SkipBinaries bool
	// IncludeWikis indicates whether to include repository wikis in the scan.
	IncludeWikis bool
	// CommentsTimeframeDays indicates how many days of comments to include in the scan.
	CommentsTimeframeDays uint32
	// AuthInUrl determines wether to use authentication token in repository url or in header.
	AuthInUrl bool
	// ClonePath is the local path used to clone repositories before scanning.
	ClonePath string
	// NoCleanup allows to keeps cloned repositories in ClonePath after scanning instead of removing them.
	NoCleanup   bool
	IgnoreGists bool
	// ExcludeArchived indicates whether to exclude archived repositories from the scan.
	ExcludeArchived bool
	// PrintLegacyJSON indicates whether to print legacy JSON output format for this source.
	PrintLegacyJSON bool
}

// GitHubExperimentalConfig defines the optional configuration for an experimental GitHub source.
type GitHubExperimentalConfig struct {
	// Repository is the repository to scan.
	Repository string
	// Token is the token to use to authenticate with the source.
	Token string
	// ObjectDiscovery indicates whether to discover all commit objects (CFOR) in the repository.
	ObjectDiscovery bool
	// CollisionThreshold is the number of short-sha collisions tolerated during hidden data enumeration. Default is 1.
	CollisionThreshold int
	// DeleteCachedData indicates whether to delete cached data.
	DeleteCachedData bool
}

// GitlabConfig defines the optional configuration for a gitlab source.
type GitlabConfig struct {
	// Endpoint is the endpoint of the source.
	Endpoint string
	// Token is the token to use to authenticate with the source.
	Token string
	// Repos is the list of repositories to scan.
	Repos []string
	// GroupIds is the list of groups to scan.
	GroupIds []string
	// Filter is the filter to use to scan the source.
	Filter *common.Filter
	// SkipBinaries allows skipping binary files from the scan.
	SkipBinaries bool
	// IncludeRepos is a list of repositories to include in the scan.
	IncludeRepos []string
	// ExcludeRepos is a list of repositories to exclude from the scan.
	ExcludeRepos []string
	// AuthInUrl determines wether to use authentication token in repository url or in header.
	AuthInUrl bool
	// ClonePath is the local path used to clone repositories before scanning
	ClonePath string
	// NoCleanup allows to keeps cloned repositories in ClonePath after scanning instead of removing them.
	NoCleanup bool
	// PrintLegacyJSON indicates whether to print legacy JSON output format for this source.
	PrintLegacyJSON bool
}

// FilesystemConfig defines the optional configuration for a filesystem source.
type FilesystemConfig struct {
	// Paths is the list of files and directories to scan.
	Paths []string
	// IncludePathsFile is the path to a file containing a list of regexps to include in the scan.
	IncludePathsFile string
	// ExcludePathsFile is the path to a file containing a list of regexps to exclude from the scan.
	ExcludePathsFile string
	// MaxSymlinkDepth enables following symlink upto the depth specified with max depth of 40
	MaxSymlinkDepth int32
}

// S3Config defines the optional configuration for an S3 source.
type S3Config struct {
	// CloudCred determines whether to use cloud credentials.
	// This can NOT be used with a secret.
	CloudCred bool
	// Key is any key to use to authenticate with the source.
	Key,
	// Secret is any secret to use to authenticate with the source.
	Secret,
	// Temporary session token associated with a temporary access key id and secret key.
	SessionToken string
	// Buckets is the list of buckets to scan.
	Buckets []string
	// IgnoreBuckets is the list buckets to ignore.
	IgnoreBuckets []string
	// Roles is the list of Roles to use.
	Roles []string
	// MaxObjectSize is the maximum object size to scan.
	MaxObjectSize int64
	// Endpoint targets an S3-compatible service other than AWS S3.
	Endpoint string
	// Region is the region used to sign requests.
	Region string
	// IncludePrefixes limits the scan to objects under these key prefixes.
	IncludePrefixes []string
	// ExcludePrefixes skips objects under these key prefixes.
	ExcludePrefixes []string
	// IncludeExtensions limits the scan to objects with these file extensions.
	IncludeExtensions []string
	// ExcludeExtensions skips objects with these file extensions.
	ExcludeExtensions []string
}

// SyslogConfig defines the optional configuration for a syslog source.
type SyslogConfig struct {
	// Address used to connect to the source.
	Address,
	// Protocol used to connect to the source.
	Protocol,
	// CertPath is the path to the certificate to use to connect to the source.
	CertPath,
	// Format is the format used to connect to the source.
	Format,
	// KeyPath is the path to the key to use to connect to the source.
	KeyPath string
	// Concurrency is the number of concurrent workers to use to scan the source.
	Concurrency int
}

// PostmanConfig defines the optional configuration for a Postman source.
type PostmanConfig struct {
	// Workspace UUID(s) or file path(s) to Postman workspace (.zip)
	Workspaces []string
	// Collection ID(s) or file path(s) to Postman collection (.json)
	Collections []string
	// Environment ID(s) or file path(s) to Postman environment (.json)
	Environments []string
	// Token is the token to use to authenticate with the API.
	Token string
	// IncludeCollections is a list of Collections to include in the scan.
	IncludeCollections []string
	// IncludeEnvironment is a list of Environments to include in the scan.
	IncludeEnvironments []string
	// ExcludeCollections is a list of Collections to exclude in the scan.
	ExcludeCollections []string
	// ExcludeEnvironment is a list of Environments to exclude in the scan.
	ExcludeEnvironments []string
	// Concurrency is the number of concurrent workers to use to scan the source.
	Concurrency int
	// CollectionPaths is the list of paths to Postman collections.
	CollectionPaths []string
	// WorkspacePaths is the list of paths to Postman workspaces.
	WorkspacePaths []string
	// EnvironmentPaths is the list of paths to Postman environments.
	EnvironmentPaths []string
	// Filter is the filter to use to scan the source.
	Filter *common.Filter
}

// ElasticsearchConfig defines the configuration for an ElasticSearch source.
type ElasticsearchConfig struct {
	Nodes          []string
	Username       string
	Password       string
	CloudID        string
	APIKey         string
	ServiceToken   string
	IndexPattern   string
	QueryJSON      string
	SinceTimestamp string
	BestEffortScan bool
}

// StdinConfig defines the configuration for a stdin source.
type StdinConfig struct{}

// JSONEnumeratorConfig defines the configuration for a JSON enumerator source.
type JSONEnumeratorConfig struct {
	// Paths is the list of JSON enumerator files to scan.
	Paths []string
}

// Progress is used to update job completion progress across sources.
type Progress struct {
	mut sync.Mutex
	// encodedResumeInfoByID is used for sub-unit resumption (see below)
	encodedResumeInfoByID map[string]string
	PercentComplete       int64
	Message               string
	EncodedResumeInfo     string
	SectionsCompleted     int32
	SectionsRemaining     int32
	// unitProgressByID is used for progress inside a unit (see below)
	unitProgressByID map[string]UnitProgress
}

// Validator is an interface for validating a source. Sources can optionally implement this interface to validate
// their configuration.
type Validator interface {
	Validate(ctx context.Context) []error
}

// SetProgressComplete sets job progress information for a running job based on the highest level objects in the source.
// i is the current iteration in the loop of target scope
// scope should be the len(scopedItems)
// message is the public facing user information about the current progress
// encodedResumeInfo is an optional string representing any information necessary to resume the job if interrupted
//
//	NOTE: SetProgressOngoing should be used when source does not yet know how many items it is scanning (scope)
//	and does not want to display a percentage complete
func (p *Progress) SetProgressComplete(i, scope int, message, encodedResumeInfo string) {
	p.mut.Lock()
	defer p.mut.Unlock()

	p.Message = message
	p.EncodedResumeInfo = encodedResumeInfo
	p.SectionsCompleted = int32(i)
	p.SectionsRemaining = int32(scope)

	// If the iteration and scope are both 0, completion is 100%.
	if i == 0 && scope == 0 {
		p.PercentComplete = 100
		return
	}

	p.PercentComplete = int64((float64(i) / float64(scope)) * 100)
}

// SetProgressOngoing sets information about the current running job based on
// the highest level objects in the source.
// message is the public facing user information about the current progress
// encodedResumeInfo is an optional string representing any information necessary to resume the job if interrupted
//
//	NOTE: This method should be used over SetProgressComplete when the source does
//	not yet know how many items it is scanning and does not want to display a percentage complete.
func (p *Progress) SetProgressOngoing(message string, encodedResumeInfo string) {
	p.mut.Lock()
	defer p.mut.Unlock()

	p.Message = message
	p.EncodedResumeInfo = encodedResumeInfo
	// Explicitly set SectionsRemaining to 0 so the frontend does not display a percent.
	p.SectionsRemaining = 0
}

// GetProgress gets job completion percentage for metrics reporting.
func (p *Progress) GetProgress() *Progress {
	p.mut.Lock()
	defer p.mut.Unlock()
	return p
}

// -sub-unit-resumption------------------------------------------------------------
//
// The following collection of methods are intended to provide a thread-safe
// way to access the EncodedResumeInfo for Sources to enable saving and
// resuming progress mid SourceUnit scan.
//
// This level of synchronization is only necessary when multiple concurrent
// invocations of ChunkUnit consume/mutate the same Progress object. The source
// manager executes scans this way under certain circumstances.
//
// Usage:
//  - id should be the SourceUnit ID
//  - value is opaque data each Source uses
//

// GetEncodedResumeInfoFor gets the encoded resume information for the provided
// ID, usually a unit ID.
func (p *Progress) GetEncodedResumeInfoFor(id string) string {
	p.mut.Lock()
	defer p.mut.Unlock()
	p.ensureEncodedResumeInfoByID()
	return p.encodedResumeInfoByID[id]
}

// SetEncodedResumeInfoFor sets the encoded resume information for the provided
// ID, usually a unit ID.
func (p *Progress) SetEncodedResumeInfoFor(id, value string) {
	p.mut.Lock()
	defer p.mut.Unlock()
	p.ensureEncodedResumeInfoByID()
	p.encodedResumeInfoByID[id] = value
	p.EncodedResumeInfo = marshalEncodedResumeInfo(p.encodedResumeInfoByID)
}

// ClearEncodedResumeInfoFor removes the encoded resume information from being
// tracked.
func (p *Progress) ClearEncodedResumeInfoFor(id string) {
	p.mut.Lock()
	defer p.mut.Unlock()
	p.ensureEncodedResumeInfoByID()
	delete(p.encodedResumeInfoByID, id)
	p.EncodedResumeInfo = marshalEncodedResumeInfo(p.encodedResumeInfoByID)
}

// ensureEncodedResumeInfoByID ensures the encodedResumeInfoByID attribute is a
// non-nil map. The mutex must be held when calling this function.
func (p *Progress) ensureEncodedResumeInfoByID() {
	if p.encodedResumeInfoByID != nil {
		return
	}
	p.encodedResumeInfoByID = unmarshalEncodedResumeInfo(p.EncodedResumeInfo)
}

// marshalEncodedResumeInfo converts a map of values into a serialized string.
func marshalEncodedResumeInfo(values map[string]string) string {
	marshalled, _ := json.Marshal(values)
	return string(marshalled)
}

// unmarshalEncodedResumeInfo converts a serialized string into a map of values.
func unmarshalEncodedResumeInfo(data string) map[string]string {
	resumeInfo := make(map[string]string)
	_ = json.Unmarshal([]byte(data), &resumeInfo)
	return resumeInfo
}

// -/sub-unit-resumption-----------------------------------------------------------

// -unit-progress------------------------------------------------------------------
//
// The following methods let a Source report how far it has got inside a unit,
// for units large enough that counting finished units reads as no progress.
// They are kept apart from EncodedResumeInfo, so reporting progress never
// changes where a resumed scan starts.
//
// Usage:
//  - id should be the SourceUnit ID
//

// UnitProgressState says whether the totals in a UnitProgress are final.
type UnitProgressState string

const (
	// UnitProgressCounting means the source is still finding out how much the
	// unit holds, so the totals are only a lower bound.
	UnitProgressCounting UnitProgressState = "counting"
	// UnitProgressReady means the totals are final.
	UnitProgressReady UnitProgressState = "ready"
	// UnitProgressUnavailable means the source could not find out how much the
	// unit holds, so only the done counts mean anything.
	UnitProgressUnavailable UnitProgressState = "unavailable"
)

// UnitProgress is how far a Source has got through one unit. Items are
// whatever the source scans one at a time, such as objects or files.
type UnitProgress struct {
	State      UnitProgressState `json:"state"`
	ItemsDone  uint64            `json:"items_done"`
	ItemsTotal uint64            `json:"items_total"`
	BytesDone  uint64            `json:"bytes_done"`
	BytesTotal uint64            `json:"bytes_total"`
	// ResumedBytes is the part of BytesDone that an earlier run of the unit
	// finished, so a scan rate can leave it out.
	ResumedBytes uint64 `json:"resumed_bytes"`
}

// Percent returns the share of bytes done once the totals are final, and 0
// before. It is capped at 99, because only the unit finishing makes it complete.
func (u UnitProgress) Percent() int64 {
	if u.State != UnitProgressReady || u.BytesTotal == 0 {
		return 0
	}
	return int64(min(u.BytesDone*100/u.BytesTotal, 99))
}

// SetUnitProgressFor records the progress of the unit with the provided ID.
func (p *Progress) SetUnitProgressFor(id string, progress UnitProgress) {
	p.mut.Lock()
	defer p.mut.Unlock()
	if p.unitProgressByID == nil {
		p.unitProgressByID = make(map[string]UnitProgress)
	}
	p.unitProgressByID[id] = progress
}

// GetUnitProgressFor returns the progress last recorded for the unit with the
// provided ID, and whether any was recorded.
func (p *Progress) GetUnitProgressFor(id string) (UnitProgress, bool) {
	p.mut.Lock()
	defer p.mut.Unlock()
	progress, ok := p.unitProgressByID[id]
	return progress, ok
}

// -/unit-progress-----------------------------------------------------------------