Introduce relationships as first-class objects (#607)

* migrate pkg.ID and pkg.Relationship to artifact package

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

* return relationships from tasks

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

* fix more tests

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

* add artifact.Identifiable by Identity() method

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

* fix linting

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

* remove catalog ID assignment

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

* adjust spdx helpers to use copy of packages

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

* stabilize package ID relative to encode-decode format cycles

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

* rename Identity() to ID()

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

* use zero value for nils in ID generation

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

* enable source.Location to be identifiable

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

* hoist up package relationship discovery to analysis stage

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

* update ownership-by-file-overlap relationship description

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

* add test reminders to put new relationships under test

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

* adjust PHP composer.lock parser function to return relationships

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

cmd/packages.go

@@ -253,7 +253,7 @@ func packagesExecWorker(userInput string) <-chan error {
 		}
 		defer cleanup()
 
-		catalog, d, err := syft.CatalogPackages(src, appConfig.Package.Cataloger.ScopeOpt)
+		catalog, relationships, d, err := syft.CatalogPackages(src, appConfig.Package.Cataloger.ScopeOpt)
 		if err != nil {
 			errs <- fmt.Errorf("failed to catalog input: %w", err)
 			return
@@ -264,7 +264,8 @@ func packagesExecWorker(userInput string) <-chan error {
 				PackageCatalog: catalog,
 				Distro:         d,
 			},
-			Source: src.Metadata,
+			Relationships: relationships,
+			Source:        src.Metadata,
 		}
 
 		if appConfig.Anchore.Host != "" {

cmd/power_user.go

@@ -4,6 +4,8 @@ import (
 	"fmt"
 	"sync"
 
+	"github.com/anchore/syft/syft/artifact"
+
 	"github.com/anchore/syft/syft/sbom"
 
 	"github.com/anchore/stereoscope"
@@ -88,7 +90,6 @@ func powerUserExec(_ *cobra.Command, args []string) error {
 		ui.Select(isVerbose(), appConfig.Quiet, reporter)...,
 	)
 }
-
 func powerUserExecWorker(userInput string) <-chan error {
 	errs := make(chan error)
 	go func() {
@@ -109,28 +110,61 @@ func powerUserExecWorker(userInput string) <-chan error {
 		}
 		defer cleanup()
 
-		analysisResults := sbom.SBOM{
+		s := sbom.SBOM{
 			Source: src.Metadata,
 		}
 
-		wg := &sync.WaitGroup{}
+		var results []<-chan artifact.Relationship
 		for _, task := range tasks {
-			wg.Add(1)
+			c := make(chan artifact.Relationship)
-			go func(task powerUserTask) {
+			results = append(results, c)
-				defer wg.Done()
+
-				if err = task(&analysisResults.Artifacts, src); err != nil {
+			go runTask(task, &s.Artifacts, src, c, errs)
-					errs <- err
-					return
-				}
-			}(task)
 		}
 
-		wg.Wait()
+		for relationship := range mergeResults(results...) {
+			s.Relationships = append(s.Relationships, relationship)
+		}
 
 		bus.Publish(partybus.Event{
 			Type:  event.PresenterReady,
-			Value: poweruser.NewJSONPresenter(analysisResults, *appConfig),
+			Value: poweruser.NewJSONPresenter(s, *appConfig),
 		})
 	}()
 	return errs
 }
+
+func runTask(t powerUserTask, a *sbom.Artifacts, src *source.Source, c chan<- artifact.Relationship, errs chan<- error) {
+	defer close(c)
+
+	relationships, err := t(a, src)
+	if err != nil {
+		errs <- err
+		return
+	}
+
+	for _, relationship := range relationships {
+		c <- relationship
+	}
+}
+
+func mergeResults(cs ...<-chan artifact.Relationship) <-chan artifact.Relationship {
+	var wg sync.WaitGroup
+	var results = make(chan artifact.Relationship)
+
+	wg.Add(len(cs))
+	for _, c := range cs {
+		go func(c <-chan artifact.Relationship) {
+			for n := range c {
+				results <- n
+			}
+			wg.Done()
+		}(c)
+	}
+
+	go func() {
+		wg.Wait()
+		close(results)
+	}()
+	return results
+}

cmd/power_user_tasks.go

@@ -4,6 +4,8 @@ import (
 	"crypto"
 	"fmt"
 
+	"github.com/anchore/syft/syft/artifact"
+
 	"github.com/anchore/syft/syft/sbom"
 
 	"github.com/anchore/syft/syft"
@@ -11,7 +13,7 @@ import (
 	"github.com/anchore/syft/syft/source"
 )
 
-type powerUserTask func(*sbom.Artifacts, *source.Source) error
+type powerUserTask func(*sbom.Artifacts, *source.Source) ([]artifact.Relationship, error)
 
 func powerUserTasks() ([]powerUserTask, error) {
 	var tasks []powerUserTask
@@ -43,16 +45,16 @@ func catalogPackagesTask() (powerUserTask, error) {
 		return nil, nil
 	}
 
-	task := func(results *sbom.Artifacts, src *source.Source) error {
+	task := func(results *sbom.Artifacts, src *source.Source) ([]artifact.Relationship, error) {
-		packageCatalog, theDistro, err := syft.CatalogPackages(src, appConfig.Package.Cataloger.ScopeOpt)
+		packageCatalog, relationships, theDistro, err := syft.CatalogPackages(src, appConfig.Package.Cataloger.ScopeOpt)
 		if err != nil {
-			return err
+			return nil, err
 		}
 
 		results.PackageCatalog = packageCatalog
 		results.Distro = theDistro
 
-		return nil
+		return relationships, nil
 	}
 
 	return task, nil
@@ -65,18 +67,18 @@ func catalogFileMetadataTask() (powerUserTask, error) {
 
 	metadataCataloger := file.NewMetadataCataloger()
 
-	task := func(results *sbom.Artifacts, src *source.Source) error {
+	task := func(results *sbom.Artifacts, src *source.Source) ([]artifact.Relationship, error) {
 		resolver, err := src.FileResolver(appConfig.FileMetadata.Cataloger.ScopeOpt)
 		if err != nil {
-			return err
+			return nil, err
 		}
 
 		result, err := metadataCataloger.Catalog(resolver)
 		if err != nil {
-			return err
+			return nil, err
 		}
 		results.FileMetadata = result
-		return nil
+		return nil, nil
 	}
 
 	return task, nil
@@ -111,18 +113,18 @@ func catalogFileDigestsTask() (powerUserTask, error) {
 		return nil, err
 	}
 
-	task := func(results *sbom.Artifacts, src *source.Source) error {
+	task := func(results *sbom.Artifacts, src *source.Source) ([]artifact.Relationship, error) {
 		resolver, err := src.FileResolver(appConfig.FileMetadata.Cataloger.ScopeOpt)
 		if err != nil {
-			return err
+			return nil, err
 		}
 
 		result, err := digestsCataloger.Catalog(resolver)
 		if err != nil {
-			return err
+			return nil, err
 		}
 		results.FileDigests = result
-		return nil
+		return nil, nil
 	}
 
 	return task, nil
@@ -143,18 +145,18 @@ func catalogSecretsTask() (powerUserTask, error) {
 		return nil, err
 	}
 
-	task := func(results *sbom.Artifacts, src *source.Source) error {
+	task := func(results *sbom.Artifacts, src *source.Source) ([]artifact.Relationship, error) {
 		resolver, err := src.FileResolver(appConfig.Secrets.Cataloger.ScopeOpt)
 		if err != nil {
-			return err
+			return nil, err
 		}
 
 		result, err := secretsCataloger.Catalog(resolver)
 		if err != nil {
-			return err
+			return nil, err
 		}
 		results.Secrets = result
-		return nil
+		return nil, nil
 	}
 
 	return task, nil
@@ -171,18 +173,18 @@ func catalogFileClassificationsTask() (powerUserTask, error) {
 		return nil, err
 	}
 
-	task := func(results *sbom.Artifacts, src *source.Source) error {
+	task := func(results *sbom.Artifacts, src *source.Source) ([]artifact.Relationship, error) {
 		resolver, err := src.FileResolver(appConfig.FileClassification.Cataloger.ScopeOpt)
 		if err != nil {
-			return err
+			return nil, err
 		}
 
 		result, err := classifierCataloger.Catalog(resolver)
 		if err != nil {
-			return err
+			return nil, err
 		}
 		results.FileClassifications = result
-		return nil
+		return nil, nil
 	}
 
 	return task, nil
@@ -198,18 +200,18 @@ func catalogContentsTask() (powerUserTask, error) {
 		return nil, err
 	}
 
-	task := func(results *sbom.Artifacts, src *source.Source) error {
+	task := func(results *sbom.Artifacts, src *source.Source) ([]artifact.Relationship, error) {
 		resolver, err := src.FileResolver(appConfig.FileContents.Cataloger.ScopeOpt)
 		if err != nil {
-			return err
+			return nil, err
 		}
 
 		result, err := contentsCataloger.Catalog(resolver)
 		if err != nil {
-			return err
+			return nil, err
 		}
 		results.FileContents = result
-		return nil
+		return nil, nil
 	}
 
 	return task, nil

go.mod

@@ -24,8 +24,9 @@ require (
 	github.com/gookit/color v1.2.7
 	github.com/hashicorp/go-multierror v1.1.0
 	github.com/hashicorp/go-version v1.2.0
+	github.com/jinzhu/copier v0.3.2
 	github.com/mitchellh/go-homedir v1.1.0
-	github.com/mitchellh/hashstructure v1.1.0
+	github.com/mitchellh/hashstructure/v2 v2.0.2
 	github.com/mitchellh/mapstructure v1.3.1
 	github.com/olekukonko/tablewriter v0.0.4
 	github.com/pelletier/go-toml v1.8.1

go.sum

@@ -454,6 +454,8 @@ github.com/inconshreveable/mousetrap v1.0.0 h1:Z8tu5sraLXCXIcARxBp/8cbvlwVa7Z1NH
 github.com/inconshreveable/mousetrap v1.0.0/go.mod h1:PxqpIevigyE2G7u3NXJIT2ANytuPF1OarO4DADm73n8=
 github.com/jarcoal/httpmock v1.0.5/go.mod h1:ATjnClrvW/3tijVmpL/va5Z3aAyGvqU3gCT8nX0Txik=
 github.com/jingyugao/rowserrcheck v0.0.0-20191204022205-72ab7603b68a/go.mod h1:xRskid8CManxVta/ALEhJha/pweKBaVG6fWgc0yH25s=
+github.com/jinzhu/copier v0.3.2 h1:QdBOCbaouLDYaIPFfi1bKv5F5tPpeTwXe4sD0jqtz5w=
+github.com/jinzhu/copier v0.3.2/go.mod h1:24xnZezI2Yqac9J61UC6/dG/k76ttpq0DdJI3QmUvro=
 github.com/jirfag/go-printf-func-name v0.0.0-20191110105641-45db9963cdd3/go.mod h1:HEWGJkRDzjJY2sqdDwxccsGicWEf9BQOZsq2tV+xzM0=
 github.com/jirfag/go-printf-func-name v0.0.0-20200119135958-7558a9eaa5af/go.mod h1:HEWGJkRDzjJY2sqdDwxccsGicWEf9BQOZsq2tV+xzM0=
 github.com/jmespath/go-jmespath v0.0.0-20160202185014-0b12d6b521d8/go.mod h1:Nht3zPeWKUH0NzdCt2Blrr5ys8VGpn0CEB0cQHVjt7k=
@@ -544,8 +546,8 @@ github.com/mitchellh/go-homedir v1.1.0/go.mod h1:SfyaCUpYCn1Vlf4IUYiD9fPX4A5wJrk
 github.com/mitchellh/go-ps v0.0.0-20190716172923-621e5597135b/go.mod h1:r1VsdOzOPt1ZSrGZWFoNhsAedKnEd6r9Np1+5blZCWk=
 github.com/mitchellh/go-testing-interface v1.0.0/go.mod h1:kRemZodwjscx+RGhAo8eIhFbs2+BFgRtFPeD/KE+zxI=
 github.com/mitchellh/gox v0.4.0/go.mod h1:Sd9lOJ0+aimLBi73mGofS1ycjY8lL3uZM3JPS42BGNg=
-github.com/mitchellh/hashstructure v1.1.0 h1:P6P1hdjqAAknpY/M1CGipelZgp+4y9ja9kmUZPXP+H0=
+github.com/mitchellh/hashstructure/v2 v2.0.2 h1:vGKWl0YJqUNxE8d+h8f6NJLcCJrgbhC4NcD46KavDd4=
-github.com/mitchellh/hashstructure v1.1.0/go.mod h1:xUDAozZz0Wmdiufv0uyhnHkUTN6/6d8ulp4AwfLKrmA=
+github.com/mitchellh/hashstructure/v2 v2.0.2/go.mod h1:MG3aRVU/N29oo/V/IhBX8GR/zz4kQkprJgF2EVszyDE=
 github.com/mitchellh/iochan v1.0.0/go.mod h1:JwYml1nuB7xOzsp52dPpHFffvOCDupsG0QubkSMEySY=
 github.com/mitchellh/mapstructure v0.0.0-20160808181253-ca63d7c062ee/go.mod h1:FVVH3fgwuzCH5S8UJGiWEs2h04kUh9fWfEaFds41c1Y=
 github.com/mitchellh/mapstructure v1.1.2/go.mod h1:FVVH3fgwuzCH5S8UJGiWEs2h04kUh9fWfEaFds41c1Y=

internal/formats/common/spdxhelpers/description.go

@@ -2,7 +2,7 @@ package spdxhelpers
 
 import "github.com/anchore/syft/syft/pkg"
 
-func Description(p *pkg.Package) string {
+func Description(p pkg.Package) string {
 	if hasMetadata(p) {
 		switch metadata := p.Metadata.(type) {
 		case pkg.ApkMetadata:
@@ -14,10 +14,6 @@ func Description(p *pkg.Package) string {
 	return ""
 }
 
-func packageExists(p *pkg.Package) bool {
+func hasMetadata(p pkg.Package) bool {
-	return p != nil
+	return p.Metadata != nil
-}
-
-func hasMetadata(p *pkg.Package) bool {
-	return packageExists(p) && p.Metadata != nil
 }

internal/formats/common/spdxhelpers/description_test.go

@@ -50,7 +50,7 @@ func Test_Description(t *testing.T) {
 	}
 	for _, test := range tests {
 		t.Run(test.name, func(t *testing.T) {
-			assert.Equal(t, test.expected, Description(&test.input))
+			assert.Equal(t, test.expected, Description(test.input))
 		})
 	}
 }

internal/formats/common/spdxhelpers/download_location.go

@@ -2,7 +2,7 @@ package spdxhelpers
 
 import "github.com/anchore/syft/syft/pkg"
 
-func DownloadLocation(p *pkg.Package) string {
+func DownloadLocation(p pkg.Package) string {
 	// 3.7: Package Download Location
 	// Cardinality: mandatory, one
 	// NONE if there is no download location whatsoever.

internal/formats/common/spdxhelpers/download_location_test.go

@@ -48,7 +48,7 @@ func Test_DownloadLocation(t *testing.T) {
 	}
 	for _, test := range tests {
 		t.Run(test.name, func(t *testing.T) {
-			assert.Equal(t, test.expected, DownloadLocation(&test.input))
+			assert.Equal(t, test.expected, DownloadLocation(test.input))
 		})
 	}
 }

internal/formats/common/spdxhelpers/external_refs.go

@@ -6,13 +6,9 @@ import (
 	"github.com/anchore/syft/syft/pkg"
 )
 
-func ExternalRefs(p *pkg.Package) (externalRefs []model.ExternalRef) {
+func ExternalRefs(p pkg.Package) (externalRefs []model.ExternalRef) {
 	externalRefs = make([]model.ExternalRef, 0)
 
-	if !packageExists(p) {
-		return externalRefs
-	}
-
 	for _, c := range p.CPEs {
 		externalRefs = append(externalRefs, model.ExternalRef{
 			ReferenceCategory: model.SecurityReferenceCategory,

internal/formats/common/spdxhelpers/external_refs_test.go

@@ -39,7 +39,7 @@ func Test_ExternalRefs(t *testing.T) {
 	}
 	for _, test := range tests {
 		t.Run(test.name, func(t *testing.T) {
-			assert.ElementsMatch(t, test.expected, ExternalRefs(&test.input))
+			assert.ElementsMatch(t, test.expected, ExternalRefs(test.input))
 		})
 	}
 }

internal/formats/common/spdxhelpers/files.go

@@ -9,7 +9,7 @@ import (
 	"github.com/anchore/syft/syft/pkg"
 )
 
-func Files(packageSpdxID string, p *pkg.Package) (files []model.File, fileIDs []string, relationships []model.Relationship) {
+func Files(packageSpdxID string, p pkg.Package) (files []model.File, fileIDs []string, relationships []model.Relationship) {
 	files = make([]model.File, 0)
 	fileIDs = make([]string, 0)
 	relationships = make([]model.Relationship, 0)

internal/formats/common/spdxhelpers/homepage.go

@@ -2,7 +2,7 @@ package spdxhelpers
 
 import "github.com/anchore/syft/syft/pkg"
 
-func Homepage(p *pkg.Package) string {
+func Homepage(p pkg.Package) string {
 	if hasMetadata(p) {
 		switch metadata := p.Metadata.(type) {
 		case pkg.GemMetadata:

internal/formats/common/spdxhelpers/homepage_test.go

@@ -50,7 +50,7 @@ func Test_Homepage(t *testing.T) {
 	}
 	for _, test := range tests {
 		t.Run(test.name, func(t *testing.T) {
-			assert.Equal(t, test.expected, Homepage(&test.input))
+			assert.Equal(t, test.expected, Homepage(test.input))
 		})
 	}
 }

internal/formats/common/spdxhelpers/license.go

@@ -7,7 +7,7 @@ import (
 	"github.com/anchore/syft/syft/pkg"
 )
 
-func License(p *pkg.Package) string {
+func License(p pkg.Package) string {
 	// source: https://spdx.github.io/spdx-spec/3-package-information/#313-concluded-license
 	// The options to populate this field are limited to:
 	// A valid SPDX License Expression as defined in Appendix IV;
@@ -17,7 +17,7 @@ func License(p *pkg.Package) string {
 	//   (ii) the SPDX file creator has made no attempt to determine this field; or
 	//   (iii) the SPDX file creator has intentionally provided no information (no meaning should be implied by doing so).
 
-	if !packageExists(p) || len(p.Licenses) == 0 {
+	if len(p.Licenses) == 0 {
 		return "NONE"
 	}
 

internal/formats/common/spdxhelpers/license_test.go

@@ -67,7 +67,7 @@ func Test_License(t *testing.T) {
 	}
 	for _, test := range tests {
 		t.Run(test.name, func(t *testing.T) {
-			assert.Equal(t, test.expected, License(&test.input))
+			assert.Equal(t, test.expected, License(test.input))
 		})
 	}
 }

internal/formats/common/spdxhelpers/originator_test.go

@@ -108,7 +108,7 @@ func Test_Originator(t *testing.T) {
 	}
 	for _, test := range tests {
 		t.Run(test.name, func(t *testing.T) {
-			assert.Equal(t, test.expected, Originator(&test.input))
+			assert.Equal(t, test.expected, Originator(test.input))
 		})
 	}
 }

internal/formats/common/spdxhelpers/origintor.go

@@ -6,7 +6,7 @@ import (
 	"github.com/anchore/syft/syft/pkg"
 )
 
-func Originator(p *pkg.Package) string {
+func Originator(p pkg.Package) string {
 	if hasMetadata(p) {
 		switch metadata := p.Metadata.(type) {
 		case pkg.ApkMetadata:

internal/formats/common/spdxhelpers/source_info.go

@@ -6,11 +6,7 @@ import (
 	"github.com/anchore/syft/syft/pkg"
 )
 
-func SourceInfo(p *pkg.Package) string {
+func SourceInfo(p pkg.Package) string {
-	if !packageExists(p) {
-		return ""
-	}
-
 	answer := ""
 	switch p.Type {
 	case pkg.RpmPkg:

internal/formats/common/spdxhelpers/source_info_test.go

@@ -139,7 +139,7 @@ func Test_SourceInfo(t *testing.T) {
 			if test.input.Type != "" {
 				pkgTypes = append(pkgTypes, test.input.Type)
 			}
-			actual := SourceInfo(&test.input)
+			actual := SourceInfo(test.input)
 			for _, expected := range test.expected {
 				assert.Contains(t, actual, expected)
 			}

internal/formats/common/testutils/utils.go

@@ -133,7 +133,6 @@ func populateImageCatalog(catalog *pkg.Catalog, img *image.Image) {
 
 	// populate catalog with test data
 	catalog.Add(pkg.Package{
-		ID:      "package-1-id",
 		Name:    "package-1",
 		Version: "1.0.1",
 		Locations: []source.Location{
@@ -154,7 +153,6 @@ func populateImageCatalog(catalog *pkg.Catalog, img *image.Image) {
 		},
 	})
 	catalog.Add(pkg.Package{
-		ID:      "package-2-id",
 		Name:    "package-2",
 		Version: "2.0.1",
 		Locations: []source.Location{
@@ -197,7 +195,6 @@ func newDirectoryCatalog() *pkg.Catalog {
 
 	// populate catalog with test data
 	catalog.Add(pkg.Package{
-		ID:      "package-1-id",
 		Name:    "package-1",
 		Version: "1.0.1",
 		Type:    pkg.PythonPkg,
@@ -223,7 +220,6 @@ func newDirectoryCatalog() *pkg.Catalog {
 		},
 	})
 	catalog.Add(pkg.Package{
-		ID:      "package-2-id",
 		Name:    "package-2",
 		Version: "2.0.1",
 		Type:    pkg.DebPkg,

internal/formats/cyclonedx12xml/to_format_model.go

@@ -33,7 +33,7 @@ func toFormatModel(s sbom.SBOM) model.Document {
 	return doc
 }
 
-func toComponent(p *pkg.Package) model.Component {
+func toComponent(p pkg.Package) model.Component {
 	return model.Component{
 		Type:       "library", // TODO: this is not accurate
 		Name:       p.Name,

internal/formats/spdx22tagvalue/to_format_model.go

@@ -256,7 +256,7 @@ func toFormatPackages(catalog *pkg.Catalog) map[spdx.ElementID]*spdx.Package2_2
 	return results
 }
 
-func formatSPDXExternalRefs(p *pkg.Package) (refs []*spdx.PackageExternalReference2_2) {
+func formatSPDXExternalRefs(p pkg.Package) (refs []*spdx.PackageExternalReference2_2) {
 	for _, ref := range spdxhelpers.ExternalRefs(p) {
 		refs = append(refs, &spdx.PackageExternalReference2_2{
 			Category:           string(ref.ReferenceCategory),

internal/formats/syftjson/decoder_test.go

@@ -31,11 +31,7 @@ func TestEncodeDecodeCycle(t *testing.T) {
 			continue
 		}
 
-		// ids will never be equal
+		for _, d := range deep.Equal(p, actualPackages[idx]) {
-		p.ID = ""
-		actualPackages[idx].ID = ""
-
-		for _, d := range deep.Equal(*p, *actualPackages[idx]) {
 			if strings.Contains(d, ".VirtualPath: ") {
 				// location.Virtual path is not exposed in the json output
 				continue

internal/formats/syftjson/test-fixtures/snapshot/TestDirectoryPresenter.golden

@@ -1,7 +1,7 @@
 {
  "artifacts": [
   {
-   "id": "package-1-id",
+   "id": "cbf4f3077fc7deee",
    "name": "package-1",
    "version": "1.0.1",
    "type": "python",
@@ -36,7 +36,7 @@
    }
   },
   {
-   "id": "package-2-id",
+   "id": "1a39aadd9705c2b9",
    "name": "package-2",
    "version": "2.0.1",
    "type": "deb",

internal/formats/syftjson/test-fixtures/snapshot/TestImagePresenter.golden

@@ -1,7 +1,7 @@
 {
  "artifacts": [
   {
-   "id": "package-1-id",
+   "id": "d1d433485a31ed07",
    "name": "package-1",
    "version": "1.0.1",
    "type": "python",
@@ -32,7 +32,7 @@
    }
   },
   {
-   "id": "package-2-id",
+   "id": "2db629ca48fa6786",
    "name": "package-2",
    "version": "2.0.1",
    "type": "deb",

internal/formats/syftjson/to_format_model.go

@@ -3,6 +3,8 @@ package syftjson
 import (
 	"fmt"
 
+	"github.com/anchore/syft/syft/artifact"
+
 	"github.com/anchore/syft/syft/sbom"
 
 	"github.com/anchore/syft/internal"
@@ -23,7 +25,7 @@ func ToFormatModel(s sbom.SBOM, applicationConfig interface{}) model.Document {
 
 	return model.Document{
 		Artifacts:             toPackageModels(s.Artifacts.PackageCatalog),
-		ArtifactRelationships: toRelationshipModel(pkg.NewRelationships(s.Artifacts.PackageCatalog)),
+		ArtifactRelationships: toRelationshipModel(s.Relationships),
 		Source:                src,
 		Distro:                toDistroModel(s.Artifacts.Distro),
 		Descriptor: model.Descriptor{
@@ -50,7 +52,7 @@ func toPackageModels(catalog *pkg.Catalog) []model.Package {
 }
 
 // toPackageModel crates a new Package from the given pkg.Package.
-func toPackageModel(p *pkg.Package) model.Package {
+func toPackageModel(p pkg.Package) model.Package {
 	var cpes = make([]string, len(p.CPEs))
 	for i, c := range p.CPEs {
 		cpes[i] = c.BindToFmtString()
@@ -69,7 +71,7 @@ func toPackageModel(p *pkg.Package) model.Package {
 
 	return model.Package{
 		PackageBasicData: model.PackageBasicData{
-			ID:        string(p.ID),
+			ID:        string(p.ID()),
 			Name:      p.Name,
 			Version:   p.Version,
 			Type:      p.Type,
@@ -87,14 +89,14 @@ func toPackageModel(p *pkg.Package) model.Package {
 	}
 }
 
-func toRelationshipModel(relationships []pkg.Relationship) []model.Relationship {
+func toRelationshipModel(relationships []artifact.Relationship) []model.Relationship {
 	result := make([]model.Relationship, len(relationships))
 	for i, r := range relationships {
 		result[i] = model.Relationship{
-			Parent:   string(r.Parent),
+			Parent:   string(r.From.ID()),
-			Child:    string(r.Child),
+			Child:    string(r.To.ID()),
 			Type:     string(r.Type),
-			Metadata: r.Metadata,
+			Metadata: r.Data,
 		}
 	}
 	return result

internal/formats/syftjson/to_syft_model.go

@@ -61,7 +61,6 @@ func toSyftPackage(p model.Package) pkg.Package {
 	}
 
 	return pkg.Package{
-		ID:           pkg.ID(p.ID),
 		Name:         p.Name,
 		Version:      p.Version,
 		FoundBy:      p.FoundBy,

internal/presenter/poweruser/json_presenter_test.go

@@ -33,7 +33,6 @@ func TestJSONPresenter(t *testing.T) {
 	catalog := pkg.NewCatalog()
 
 	catalog.Add(pkg.Package{
-		ID:      "package-1-id",
 		Name:    "package-1",
 		Version: "1.0.1",
 		Locations: []source.Location{
@@ -57,7 +56,6 @@ func TestJSONPresenter(t *testing.T) {
 		},
 	})
 	catalog.Add(pkg.Package{
-		ID:      "package-2-id",
 		Name:    "package-2",
 		Version: "2.0.1",
 		Locations: []source.Location{

internal/presenter/poweruser/test-fixtures/snapshot/TestJSONPresenter.golden

@@ -72,7 +72,7 @@
  ],
  "artifacts": [
   {
-   "id": "package-1-id",
+   "id": "b84dfe0eb2c5670f",
    "name": "package-1",
    "version": "1.0.1",
    "type": "python",
@@ -102,7 +102,7 @@
    }
   },
   {
-   "id": "package-2-id",
+   "id": "6619226d6979963f",
    "name": "package-2",
    "version": "2.0.1",
    "type": "deb",

syft/artifact/id.go

@@ -0,0 +1,26 @@
+package artifact
+
+import (
+	"fmt"
+
+	"github.com/mitchellh/hashstructure/v2"
+)
+
+// ID represents a unique value for each package added to a package catalog.
+type ID string
+
+type Identifiable interface {
+	ID() ID
+}
+
+func IDFromHash(obj interface{}) (ID, error) {
+	f, err := hashstructure.Hash(obj, hashstructure.FormatV2, &hashstructure.HashOptions{
+		ZeroNil:      true,
+		SlicesAsSets: true,
+	})
+	if err != nil {
+		return "", fmt.Errorf("could not build ID for object=%+v: %+v", obj, err)
+	}
+
+	return ID(fmt.Sprintf("%x", f)), nil
+}

syft/artifact/relationship.go

@@ -0,0 +1,17 @@
+package artifact
+
+const (
+	// OwnershipByFileOverlapRelationship indicates that the parent package claims ownership of a child package since
+	// the parent metadata indicates overlap with a location that a cataloger found the child package by. This is
+	// by definition a package-to-package relationship and is created only after all package cataloging has been completed.
+	OwnershipByFileOverlapRelationship RelationshipType = "ownership-by-file-overlap"
+)
+
+type RelationshipType string
+
+type Relationship struct {
+	From Identifiable     `json:"from"`
+	To   Identifiable     `json:"to"`
+	Type RelationshipType `json:"type"`
+	Data interface{}      `json:"data,omitempty"`
+}

syft/lib.go

@@ -19,6 +19,8 @@ package syft
 import (
 	"fmt"
 
+	"github.com/anchore/syft/syft/artifact"
+
 	"github.com/anchore/syft/internal/bus"
 	"github.com/anchore/syft/internal/log"
 	"github.com/anchore/syft/syft/distro"
@@ -32,10 +34,10 @@ import (
 // CatalogPackages takes an inventory of packages from the given image from a particular perspective
 // (e.g. squashed source, all-layers source). Returns the discovered  set of packages, the identified Linux
 // distribution, and the source object used to wrap the data source.
-func CatalogPackages(src *source.Source, scope source.Scope) (*pkg.Catalog, *distro.Distro, error) {
+func CatalogPackages(src *source.Source, scope source.Scope) (*pkg.Catalog, []artifact.Relationship, *distro.Distro, error) {
 	resolver, err := src.FileResolver(scope)
 	if err != nil {
-		return nil, nil, fmt.Errorf("unable to determine resolver while cataloging packages: %w", err)
+		return nil, nil, nil, fmt.Errorf("unable to determine resolver while cataloging packages: %w", err)
 	}
 
 	// find the distro
@@ -59,15 +61,15 @@ func CatalogPackages(src *source.Source, scope source.Scope) (*pkg.Catalog, *dis
 		log.Info("cataloging directory")
 		catalogers = cataloger.DirectoryCatalogers()
 	default:
-		return nil, nil, fmt.Errorf("unable to determine cataloger set from scheme=%+v", src.Metadata.Scheme)
+		return nil, nil, nil, fmt.Errorf("unable to determine cataloger set from scheme=%+v", src.Metadata.Scheme)
 	}
 
-	catalog, err := cataloger.Catalog(resolver, theDistro, catalogers...)
+	catalog, relationships, err := cataloger.Catalog(resolver, theDistro, catalogers...)
 	if err != nil {
-		return nil, nil, err
+		return nil, nil, nil, err
 	}
 
-	return catalog, theDistro, nil
+	return catalog, relationships, theDistro, nil
 }
 
 // SetLogger sets the logger object used for all syft logging calls.

syft/pkg/catalog.go

@@ -4,6 +4,9 @@ import (
 	"sort"
 	"sync"
 
+	"github.com/anchore/syft/syft/artifact"
+	"github.com/jinzhu/copier"
+
 	"github.com/anchore/syft/internal"
 
 	"github.com/anchore/syft/internal/log"
@@ -11,18 +14,18 @@ import (
 
 // Catalog represents a collection of Packages.
 type Catalog struct {
-	byID      map[ID]*Package
+	byID      map[artifact.ID]Package
-	idsByType map[Type][]ID
+	idsByType map[Type][]artifact.ID
-	idsByPath map[string][]ID // note: this is real path or virtual path
+	idsByPath map[string][]artifact.ID // note: this is real path or virtual path
 	lock      sync.RWMutex
 }
 
 // NewCatalog returns a new empty Catalog
 func NewCatalog(pkgs ...Package) *Catalog {
 	catalog := Catalog{
-		byID:      make(map[ID]*Package),
+		byID:      make(map[artifact.ID]Package),
-		idsByType: make(map[Type][]ID),
+		idsByType: make(map[Type][]artifact.ID),
-		idsByPath: make(map[string][]ID),
+		idsByPath: make(map[string][]artifact.ID),
 	}
 
 	for _, p := range pkgs {
@@ -38,21 +41,26 @@ func (c *Catalog) PackageCount() int {
 }
 
 // Package returns the package with the given ID.
-func (c *Catalog) Package(id ID) *Package {
+func (c *Catalog) Package(id artifact.ID) *Package {
 	v, exists := c.byID[id]
 	if !exists {
 		return nil
 	}
-	return v
+	var p Package
+	if err := copier.Copy(&p, &v); err != nil {
+		log.Warnf("unable to copy package id=%q name=%q: %+v", id, v.Name, err)
+		return nil
+	}
+	return &p
 }
 
 // PackagesByPath returns all packages that were discovered from the given path.
-func (c *Catalog) PackagesByPath(path string) []*Package {
+func (c *Catalog) PackagesByPath(path string) []Package {
 	return c.Packages(c.idsByPath[path])
 }
 
 // Packages returns all packages for the given ID.
-func (c *Catalog) Packages(ids []ID) (result []*Package) {
+func (c *Catalog) Packages(ids []artifact.ID) (result []Package) {
 	for _, i := range ids {
 		p, exists := c.byID[i]
 		if exists {
@@ -67,68 +75,32 @@ func (c *Catalog) Add(p Package) {
 	c.lock.Lock()
 	defer c.lock.Unlock()
 
-	if p.ID == "" {
+	// note: since we are capturing the ID, we cannot modify the package being added from this point forward
-		fingerprint, err := p.Fingerprint()
+	id := p.ID()
-		if err != nil {
-			log.Warnf("failed to add package to catalog: %w", err)
-			return
-		}
-
-		p.ID = ID(fingerprint)
-	}
 
 	// store by package ID
-	c.byID[p.ID] = &p
+	c.byID[id] = p
 
 	// store by package type
-	c.idsByType[p.Type] = append(c.idsByType[p.Type], p.ID)
+	c.idsByType[p.Type] = append(c.idsByType[p.Type], id)
 
 	// store by file location paths
 	observedPaths := internal.NewStringSet()
 	for _, l := range p.Locations {
 		if l.RealPath != "" && !observedPaths.Contains(l.RealPath) {
-			c.idsByPath[l.RealPath] = append(c.idsByPath[l.RealPath], p.ID)
+			c.idsByPath[l.RealPath] = append(c.idsByPath[l.RealPath], id)
 			observedPaths.Add(l.RealPath)
 		}
 		if l.VirtualPath != "" && l.RealPath != l.VirtualPath && !observedPaths.Contains(l.VirtualPath) {
-			c.idsByPath[l.VirtualPath] = append(c.idsByPath[l.VirtualPath], p.ID)
+			c.idsByPath[l.VirtualPath] = append(c.idsByPath[l.VirtualPath], id)
 			observedPaths.Add(l.VirtualPath)
 		}
 	}
 }
 
-func (c *Catalog) Remove(id ID) {
-	c.lock.Lock()
-	defer c.lock.Unlock()
-
-	_, exists := c.byID[id]
-	if !exists {
-		log.Errorf("package ID does not exist in the catalog : id=%+v", id)
-		return
-	}
-
-	// Remove all index references to this package ID
-	for t, ids := range c.idsByType {
-		c.idsByType[t] = removeID(id, ids)
-		if len(c.idsByType[t]) == 0 {
-			delete(c.idsByType, t)
-		}
-	}
-
-	for p, ids := range c.idsByPath {
-		c.idsByPath[p] = removeID(id, ids)
-		if len(c.idsByPath[p]) == 0 {
-			delete(c.idsByPath, p)
-		}
-	}
-
-	// Remove package
-	delete(c.byID, id)
-}
-
 // Enumerate all packages for the given type(s), enumerating all packages if no type is specified.
-func (c *Catalog) Enumerate(types ...Type) <-chan *Package {
+func (c *Catalog) Enumerate(types ...Type) <-chan Package {
-	channel := make(chan *Package)
+	channel := make(chan Package)
 	go func() {
 		defer close(channel)
 		for ty, ids := range c.idsByType {
@@ -146,7 +118,10 @@ func (c *Catalog) Enumerate(types ...Type) <-chan *Package {
 				}
 			}
 			for _, id := range ids {
-				channel <- c.Package(id)
+				p := c.Package(id)
+				if p != nil {
+					channel <- *p
+				}
 			}
 		}
 	}()
@@ -155,8 +130,7 @@ func (c *Catalog) Enumerate(types ...Type) <-chan *Package {
 
 // Sorted enumerates all packages for the given types sorted by package name. Enumerates all packages if no type
 // is specified.
-func (c *Catalog) Sorted(types ...Type) []*Package {
+func (c *Catalog) Sorted(types ...Type) (pkgs []Package) {
-	pkgs := make([]*Package, 0)
 	for p := range c.Enumerate(types...) {
 		pkgs = append(pkgs, p)
 	}
@@ -176,12 +150,3 @@ func (c *Catalog) Sorted(types ...Type) []*Package {
 
 	return pkgs
 }
-
-func removeID(id ID, target []ID) (result []ID) {
-	for _, value := range target {
-		if value != id {
-			result = append(result, value)
-		}
-	}
-	return result
-}

syft/pkg/catalog_test.go

@@ -10,7 +10,6 @@ import (
 
 var catalogAddAndRemoveTestPkgs = []Package{
 	{
-		ID: "my-id",
 		Locations: []source.Location{
 			{
 				RealPath:    "/a/path",
@@ -24,7 +23,6 @@ var catalogAddAndRemoveTestPkgs = []Package{
 		Type: RpmPkg,
 	},
 	{
-		ID: "my-other-id",
 		Locations: []source.Location{
 			{
 				RealPath:    "/c/path",
@@ -45,6 +43,11 @@ type expectedIndexes struct {
 }
 
 func TestCatalogAddPopulatesIndex(t *testing.T) {
+
+	fixtureID := func(i int) string {
+		return string(catalogAddAndRemoveTestPkgs[i].ID())
+	}
+
 	tests := []struct {
 		name            string
 		pkgs            []Package
@@ -55,16 +58,16 @@ func TestCatalogAddPopulatesIndex(t *testing.T) {
 			pkgs: catalogAddAndRemoveTestPkgs,
 			expectedIndexes: expectedIndexes{
 				byType: map[Type]*strset.Set{
-					RpmPkg: strset.New("my-id"),
+					RpmPkg: strset.New(fixtureID(0)),
-					NpmPkg: strset.New("my-other-id"),
+					NpmPkg: strset.New(fixtureID(1)),
 				},
 				byPath: map[string]*strset.Set{
-					"/another/path": strset.New("my-id", "my-other-id"),
+					"/another/path": strset.New(fixtureID(0), fixtureID(1)),
-					"/a/path":       strset.New("my-id"),
+					"/a/path":       strset.New(fixtureID(0)),
-					"/b/path":       strset.New("my-id"),
+					"/b/path":       strset.New(fixtureID(0)),
-					"/bee/path":     strset.New("my-id"),
+					"/bee/path":     strset.New(fixtureID(0)),
-					"/c/path":       strset.New("my-other-id"),
+					"/c/path":       strset.New(fixtureID(1)),
-					"/d/path":       strset.New("my-other-id"),
+					"/d/path":       strset.New(fixtureID(1)),
 				},
 			},
 		},
@@ -80,50 +83,6 @@ func TestCatalogAddPopulatesIndex(t *testing.T) {
 	}
 }
 
-func TestCatalogRemove(t *testing.T) {
-	tests := []struct {
-		name            string
-		pkgs            []Package
-		removeId        ID
-		expectedIndexes expectedIndexes
-	}{
-		{
-			name:     "vanilla-add",
-			removeId: "my-other-id",
-			pkgs:     catalogAddAndRemoveTestPkgs,
-			expectedIndexes: expectedIndexes{
-				byType: map[Type]*strset.Set{
-					RpmPkg: strset.New("my-id"),
-				},
-				byPath: map[string]*strset.Set{
-					"/another/path": strset.New("my-id"),
-					"/a/path":       strset.New("my-id"),
-					"/b/path":       strset.New("my-id"),
-					"/bee/path":     strset.New("my-id"),
-				},
-			},
-		},
-	}
-
-	for _, test := range tests {
-		t.Run(test.name, func(t *testing.T) {
-			c := NewCatalog(test.pkgs...)
-			c.Remove(test.removeId)
-
-			assertIndexes(t, c, test.expectedIndexes)
-
-			if c.Package(test.removeId) != nil {
-				t.Errorf("expected package to be removed, but was found!")
-			}
-
-			if c.PackageCount() != len(test.pkgs)-1 {
-				t.Errorf("expected count to be affected but was not")
-			}
-
-		})
-	}
-}
-
 func assertIndexes(t *testing.T, c *Catalog, expectedIndexes expectedIndexes) {
 	// assert path index
 	if len(c.idsByPath) != len(expectedIndexes.byPath) {
@@ -132,7 +91,7 @@ func assertIndexes(t *testing.T, c *Catalog, expectedIndexes expectedIndexes) {
 	for path, expectedIds := range expectedIndexes.byPath {
 		actualIds := strset.New()
 		for _, p := range c.PackagesByPath(path) {
-			actualIds.Add(string(p.ID))
+			actualIds.Add(string(p.ID()))
 		}
 
 		if !expectedIds.IsEqual(actualIds) {
@@ -147,7 +106,7 @@ func assertIndexes(t *testing.T, c *Catalog, expectedIndexes expectedIndexes) {
 	for ty, expectedIds := range expectedIndexes.byType {
 		actualIds := strset.New()
 		for p := range c.Enumerate(ty) {
-			actualIds.Add(string(p.ID))
+			actualIds.Add(string(p.ID()))
 		}
 
 		if !expectedIds.IsEqual(actualIds) {
@@ -157,39 +116,42 @@ func assertIndexes(t *testing.T, c *Catalog, expectedIndexes expectedIndexes) {
 }
 
 func TestCatalog_PathIndexDeduplicatesRealVsVirtualPaths(t *testing.T) {
+	p1 := Package{
+		Locations: []source.Location{
+			{
+				RealPath:    "/b/path",
+				VirtualPath: "/another/path",
+			},
+			{
+				RealPath:    "/b/path",
+				VirtualPath: "/b/path",
+			},
+		},
+		Type: RpmPkg,
+		Name: "Package-1",
+	}
+
+	p2 := Package{
+		Locations: []source.Location{
+			{
+				RealPath:    "/b/path",
+				VirtualPath: "/b/path",
+			},
+		},
+		Type: RpmPkg,
+		Name: "Package-2",
+	}
 	tests := []struct {
 		name string
 		pkg  Package
 	}{
 		{
 			name: "multiple locations with shared path",
-			pkg: Package{
+			pkg:  p1,
-				ID: "my-id",
-				Locations: []source.Location{
-					{
-						RealPath:    "/b/path",
-						VirtualPath: "/another/path",
-					},
-					{
-						RealPath:    "/b/path",
-						VirtualPath: "/b/path",
-					},
-				},
-				Type: RpmPkg,
-			},
 		},
 		{
 			name: "one location with shared path",
-			pkg: Package{
+			pkg:  p2,
-				ID: "my-id",
-				Locations: []source.Location{
-					{
-						RealPath:    "/b/path",
-						VirtualPath: "/b/path",
-					},
-				},
-				Type: RpmPkg,
-			},
 		},
 	}
 

syft/pkg/cataloger/apkdb/parse_apk_db.go

@@ -8,6 +8,8 @@ import (
 	"strconv"
 	"strings"
 
+	"github.com/anchore/syft/syft/artifact"
+
 	"github.com/anchore/syft/syft/file"
 
 	"github.com/anchore/syft/internal/log"
@@ -21,7 +23,7 @@ var _ common.ParserFn = parseApkDB
 
 // parseApkDb parses individual packages from a given Alpine DB file. For more information on specific fields
 // see https://wiki.alpinelinux.org/wiki/Apk_spec .
-func parseApkDB(_ string, reader io.Reader) ([]pkg.Package, error) {
+func parseApkDB(_ string, reader io.Reader) ([]pkg.Package, []artifact.Relationship, error) {
 	// larger capacity for the scanner.
 	const maxScannerCapacity = 1024 * 1024
 	// a new larger buffer for the scanner
@@ -47,7 +49,7 @@ func parseApkDB(_ string, reader io.Reader) ([]pkg.Package, error) {
 	for scanner.Scan() {
 		metadata, err := parseApkDBEntry(strings.NewReader(scanner.Text()))
 		if err != nil {
-			return nil, err
+			return nil, nil, err
 		}
 		if metadata != nil {
 			packages = append(packages, pkg.Package{
@@ -62,10 +64,10 @@ func parseApkDB(_ string, reader io.Reader) ([]pkg.Package, error) {
 	}
 
 	if err := scanner.Err(); err != nil {
-		return nil, fmt.Errorf("failed to parse APK DB file: %w", err)
+		return nil, nil, fmt.Errorf("failed to parse APK DB file: %w", err)
 	}
 
-	return packages, nil
+	return packages, nil, nil
 }
 
 // nolint:funlen

syft/pkg/cataloger/apkdb/parse_apk_db_test.go

@@ -775,7 +775,8 @@ func TestMultiplePackages(t *testing.T) {
 				}
 			}()
 
-			pkgs, err := parseApkDB(file.Name(), file)
+			// TODO: no relationships are under test yet
+			pkgs, _, err := parseApkDB(file.Name(), file)
 			if err != nil {
 				t.Fatal("Unable to read file contents: ", err)
 			}

syft/pkg/cataloger/catalog.go

@@ -3,6 +3,7 @@ package cataloger
 import (
 	"github.com/anchore/syft/internal/bus"
 	"github.com/anchore/syft/internal/log"
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/distro"
 	"github.com/anchore/syft/syft/event"
 	"github.com/anchore/syft/syft/pkg"
@@ -38,8 +39,9 @@ func newMonitor() (*progress.Manual, *progress.Manual) {
 // In order to efficiently retrieve contents from a underlying container image the content fetch requests are
 // done in bulk. Specifically, all files of interest are collected from each catalogers and accumulated into a single
 // request.
-func Catalog(resolver source.FileResolver, theDistro *distro.Distro, catalogers ...Cataloger) (*pkg.Catalog, error) {
+func Catalog(resolver source.FileResolver, theDistro *distro.Distro, catalogers ...Cataloger) (*pkg.Catalog, []artifact.Relationship, error) {
 	catalog := pkg.NewCatalog()
+	var allRelationships []artifact.Relationship
 
 	filesProcessed, packagesDiscovered := newMonitor()
 
@@ -47,7 +49,7 @@ func Catalog(resolver source.FileResolver, theDistro *distro.Distro, catalogers
 	var errs error
 	for _, theCataloger := range catalogers {
 		// find packages from the underlying raw data
-		packages, err := theCataloger.Catalog(resolver)
+		packages, relationships, err := theCataloger.Catalog(resolver)
 		if err != nil {
 			errs = multierror.Append(errs, err)
 			continue
@@ -68,14 +70,18 @@ func Catalog(resolver source.FileResolver, theDistro *distro.Distro, catalogers
 			// add to catalog
 			catalog.Add(p)
 		}
+
+		allRelationships = append(allRelationships, relationships...)
 	}
 
+	allRelationships = append(allRelationships, pkg.NewRelationships(catalog)...)
+
 	if errs != nil {
-		return nil, errs
+		return nil, nil, errs
 	}
 
 	filesProcessed.SetCompleted()
 	packagesDiscovered.SetCompleted()
 
-	return catalog, nil
+	return catalog, allRelationships, nil
 }

syft/pkg/cataloger/cataloger.go

@@ -6,6 +6,7 @@ catalogers defined in child packages as well as the interface definition to impl
 package cataloger
 
 import (
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/pkg/cataloger/apkdb"
 	"github.com/anchore/syft/syft/pkg/cataloger/deb"
@@ -27,7 +28,7 @@ type Cataloger interface {
 	// Name returns a string that uniquely describes a cataloger
 	Name() string
 	// Catalog is given an object to resolve file references and content, this function returns any discovered Packages after analyzing the catalog source.
-	Catalog(resolver source.FileResolver) ([]pkg.Package, error)
+	Catalog(resolver source.FileResolver) ([]pkg.Package, []artifact.Relationship, error)
 }
 
 // ImageCatalogers returns a slice of locally implemented catalogers that are fit for detecting installations of packages.

syft/pkg/cataloger/common/cpe/generate.go

@@ -83,7 +83,7 @@ func candidateVendors(p pkg.Package) []string {
 	// allow * as a candidate. Note: do NOT allow Java packages to have * vendors.
 	switch p.Language {
 	case pkg.Ruby, pkg.JavaScript:
-		vendors.addValue("*")
+		vendors.addValue(wfn.Any)
 	}
 
 	switch p.MetadataType {

syft/pkg/cataloger/common/cpe/generate_test.go

@@ -637,7 +637,7 @@ func TestCandidateProducts(t *testing.T) {
 	}
 
 	for _, test := range tests {
-		t.Run(fmt.Sprintf("%+v %+v", test.p, test.expected), func(t *testing.T) {
+		t.Run(test.name, func(t *testing.T) {
 			assert.ElementsMatch(t, test.expected, candidateProducts(test.p))
 		})
 	}

syft/pkg/cataloger/common/generic_cataloger.go

@@ -6,6 +6,8 @@ package common
 import (
 	"fmt"
 
+	"github.com/anchore/syft/syft/artifact"
+
 	"github.com/anchore/syft/internal"
 	"github.com/anchore/syft/internal/log"
 	"github.com/anchore/syft/syft/pkg"
@@ -35,18 +37,18 @@ func (c *GenericCataloger) Name() string {
 }
 
 // Catalog is given an object to resolve file references and content, this function returns any discovered Packages after analyzing the catalog source.
-func (c *GenericCataloger) Catalog(resolver source.FileResolver) ([]pkg.Package, error) {
+func (c *GenericCataloger) Catalog(resolver source.FileResolver) ([]pkg.Package, []artifact.Relationship, error) {
 	var packages []pkg.Package
-	parserByLocation := c.selectFiles(resolver)
+	var relationships []artifact.Relationship
 
-	for location, parser := range parserByLocation {
+	for location, parser := range c.selectFiles(resolver) {
 		contentReader, err := resolver.FileContentsByLocation(location)
 		if err != nil {
 			// TODO: fail or log?
-			return nil, fmt.Errorf("unable to fetch contents for location=%v : %w", location, err)
+			return nil, nil, fmt.Errorf("unable to fetch contents for location=%v : %w", location, err)
 		}
 
-		entries, err := parser(location.RealPath, contentReader)
+		discoveredPackages, discoveredRelationships, err := parser(location.RealPath, contentReader)
 		internal.CloseAndLogError(contentReader, location.VirtualPath)
 		if err != nil {
 			// TODO: should we fail? or only log?
@@ -54,14 +56,16 @@ func (c *GenericCataloger) Catalog(resolver source.FileResolver) ([]pkg.Package,
 			continue
 		}
 
-		for _, entry := range entries {
+		for _, p := range discoveredPackages {
-			entry.FoundBy = c.upstreamCataloger
+			p.FoundBy = c.upstreamCataloger
-			entry.Locations = []source.Location{location}
+			p.Locations = append(p.Locations, location)
 
-			packages = append(packages, entry)
+			packages = append(packages, p)
 		}
+
+		relationships = append(relationships, discoveredRelationships...)
 	}
-	return packages, nil
+	return packages, relationships, nil
 }
 
 // SelectFiles takes a set of file trees and resolves and file references of interest for future cataloging

syft/pkg/cataloger/common/generic_cataloger_test.go

@@ -8,11 +8,12 @@ import (
 
 	"github.com/stretchr/testify/assert"
 
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/source"
 )
 
-func parser(_ string, reader io.Reader) ([]pkg.Package, error) {
+func parser(_ string, reader io.Reader) ([]pkg.Package, []artifact.Relationship, error) {
 	contents, err := ioutil.ReadAll(reader)
 	if err != nil {
 		panic(err)
@@ -21,7 +22,7 @@ func parser(_ string, reader io.Reader) ([]pkg.Package, error) {
 		{
 			Name: string(contents),
 		},
-	}, nil
+	}, nil, nil
 }
 
 func TestGenericCataloger(t *testing.T) {
@@ -47,7 +48,7 @@ func TestGenericCataloger(t *testing.T) {
 		}
 	}
 
-	actualPkgs, err := cataloger.Catalog(resolver)
+	actualPkgs, _, err := cataloger.Catalog(resolver)
 	assert.NoError(t, err)
 	assert.Len(t, actualPkgs, len(expectedPkgs))
 

syft/pkg/cataloger/common/parser.go

@@ -3,8 +3,9 @@ package common
 import (
 	"io"
 
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 )
 
 // ParserFn standardizes a function signature for parser functions that accept the virtual file path (not usable for file reads) and contents and return any discovered packages from that file
-type ParserFn func(string, io.Reader) ([]pkg.Package, error)
+type ParserFn func(string, io.Reader) ([]pkg.Package, []artifact.Relationship, error)

syft/pkg/cataloger/deb/cataloger.go

@@ -13,6 +13,7 @@ import (
 	"github.com/anchore/syft/internal"
 
 	"github.com/anchore/syft/internal/log"
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/source"
 )
@@ -36,24 +37,23 @@ func (c *Cataloger) Name() string {
 }
 
 // Catalog is given an object to resolve file references and content, this function returns any discovered Packages after analyzing dpkg support files.
-func (c *Cataloger) Catalog(resolver source.FileResolver) ([]pkg.Package, error) {
+func (c *Cataloger) Catalog(resolver source.FileResolver) ([]pkg.Package, []artifact.Relationship, error) {
 	dbFileMatches, err := resolver.FilesByGlob(pkg.DpkgDBGlob)
 	if err != nil {
-		return nil, fmt.Errorf("failed to find dpkg status files's by glob: %w", err)
+		return nil, nil, fmt.Errorf("failed to find dpkg status files's by glob: %w", err)
 	}
 
-	var results []pkg.Package
+	var allPackages []pkg.Package
-	var pkgs []pkg.Package
 	for _, dbLocation := range dbFileMatches {
 		dbContents, err := resolver.FileContentsByLocation(dbLocation)
 		if err != nil {
-			return nil, err
+			return nil, nil, err
 		}
 
-		pkgs, err = parseDpkgStatus(dbContents)
+		pkgs, err := parseDpkgStatus(dbContents)
 		internal.CloseAndLogError(dbContents, dbLocation.VirtualPath)
 		if err != nil {
-			return nil, fmt.Errorf("unable to catalog dpkg package=%+v: %w", dbLocation.RealPath, err)
+			return nil, nil, fmt.Errorf("unable to catalog dpkg package=%+v: %w", dbLocation.RealPath, err)
 		}
 
 		for i := range pkgs {
@@ -70,9 +70,9 @@ func (c *Cataloger) Catalog(resolver source.FileResolver) ([]pkg.Package, error)
 			addLicenses(resolver, dbLocation, p)
 		}
 
-		results = append(results, pkgs...)
+		allPackages = append(allPackages, pkgs...)
 	}
-	return results, nil
+	return allPackages, nil, nil
 }
 
 func addLicenses(resolver source.FileResolver, dbLocation source.Location, p *pkg.Package) {

syft/pkg/cataloger/deb/cataloger_test.go

@@ -100,7 +100,7 @@ func TestDpkgCataloger(t *testing.T) {
 				t.Errorf("could not get resolver error: %+v", err)
 			}
 
-			actual, err := c.Catalog(resolver)
+			actual, _, err := c.Catalog(resolver)
 			if err != nil {
 				t.Fatalf("failed to catalog: %+v", err)
 			}

syft/pkg/cataloger/deb/parse_dpkg_status.go

@@ -23,7 +23,7 @@ var (
 // parseDpkgStatus is a parser function for Debian DB status contents, returning all Debian packages listed.
 func parseDpkgStatus(reader io.Reader) ([]pkg.Package, error) {
 	buffedReader := bufio.NewReader(reader)
-	var packages = make([]pkg.Package, 0)
+	var packages []pkg.Package
 
 	continueProcessing := true
 	for continueProcessing {

syft/pkg/cataloger/golang/binary_cataloger.go

@@ -6,7 +6,10 @@ package golang
 import (
 	"fmt"
 
+	"github.com/anchore/syft/internal"
+
 	"github.com/anchore/syft/internal/log"
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/source"
 )
@@ -35,17 +38,18 @@ func (c *Cataloger) Name() string {
 }
 
 // Catalog is given an object to resolve file references and content, this function returns any discovered Packages after analyzing rpm db installation.
-func (c *Cataloger) Catalog(resolver source.FileResolver) ([]pkg.Package, error) {
+func (c *Cataloger) Catalog(resolver source.FileResolver) ([]pkg.Package, []artifact.Relationship, error) {
-	pkgs := make([]pkg.Package, 0)
+	var pkgs []pkg.Package
+
 	fileMatches, err := resolver.FilesByMIMEType(mimeTypes...)
 	if err != nil {
-		return pkgs, fmt.Errorf("failed to find bin by mime types: %w", err)
+		return pkgs, nil, fmt.Errorf("failed to find bin by mime types: %w", err)
 	}
 
 	for _, location := range fileMatches {
 		r, err := resolver.FileContentsByLocation(location)
 		if err != nil {
-			return pkgs, fmt.Errorf("failed to resolve file contents by location: %w", err)
+			return pkgs, nil, fmt.Errorf("failed to resolve file contents by location: %w", err)
 		}
 
 		goPkgs, err := parseGoBin(location, r)
@@ -53,9 +57,9 @@ func (c *Cataloger) Catalog(resolver source.FileResolver) ([]pkg.Package, error)
 			log.Warnf("could not parse possible go binary: %+v", err)
 		}
 
-		r.Close()
+		internal.CloseAndLogError(r, location.RealPath)
 		pkgs = append(pkgs, goPkgs...)
 	}
 
-	return pkgs, nil
+	return pkgs, nil, nil
 }

syft/pkg/cataloger/golang/parse_go_bin.go

@@ -23,9 +23,7 @@ func parseGoBin(location source.Location, reader io.ReadCloser) ([]pkg.Package,
 
 	goVersion, mod := findVers(x)
 
-	pkgs := buildGoPkgInfo(location, mod, goVersion)
+	return buildGoPkgInfo(location, mod, goVersion), nil
-
-	return pkgs, nil
 }
 
 func buildGoPkgInfo(location source.Location, mod, goVersion string) []pkg.Package {

syft/pkg/cataloger/golang/parse_go_mod.go

@@ -6,22 +6,23 @@ import (
 	"io/ioutil"
 	"sort"
 
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"golang.org/x/mod/modfile"
 )
 
 // parseGoMod takes a go.mod and lists all packages discovered.
-func parseGoMod(path string, reader io.Reader) ([]pkg.Package, error) {
+func parseGoMod(path string, reader io.Reader) ([]pkg.Package, []artifact.Relationship, error) {
 	packages := make(map[string]pkg.Package)
 
 	contents, err := ioutil.ReadAll(reader)
 	if err != nil {
-		return nil, fmt.Errorf("failed to read go module: %w", err)
+		return nil, nil, fmt.Errorf("failed to read go module: %w", err)
 	}
 
 	file, err := modfile.Parse(path, contents, nil)
 	if err != nil {
-		return nil, fmt.Errorf("failed to parse go module: %w", err)
+		return nil, nil, fmt.Errorf("failed to parse go module: %w", err)
 	}
 
 	for _, m := range file.Require {
@@ -59,5 +60,5 @@ func parseGoMod(path string, reader io.Reader) ([]pkg.Package, error) {
 		return pkgsSlice[i].Name < pkgsSlice[j].Name
 	})
 
-	return pkgsSlice, nil
+	return pkgsSlice, nil, nil
 }

syft/pkg/cataloger/golang/parse_go_mod_test.go

@@ -70,7 +70,8 @@ func TestParseGoMod(t *testing.T) {
 				t.Fatalf(err.Error())
 			}
 
-			actual, err := parseGoMod(test.fixture, f)
+			// TODO: no relationships are under test yet
+			actual, _, err := parseGoMod(test.fixture, f)
 			if err != nil {
 				t.Fatalf(err.Error())
 			}

syft/pkg/cataloger/java/archive_parser.go

@@ -9,6 +9,7 @@ import (
 	"github.com/anchore/syft/internal/log"
 
 	"github.com/anchore/syft/internal/file"
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/pkg/cataloger/common"
 )
@@ -34,12 +35,12 @@ type archiveParser struct {
 }
 
 // parseJavaArchive is a parser function for java archive contents, returning all Java libraries and nested archives.
-func parseJavaArchive(virtualPath string, reader io.Reader) ([]pkg.Package, error) {
+func parseJavaArchive(virtualPath string, reader io.Reader) ([]pkg.Package, []artifact.Relationship, error) {
 	parser, cleanupFn, err := newJavaArchiveParser(virtualPath, reader, true)
 	// note: even on error, we should always run cleanup functions
 	defer cleanupFn()
 	if err != nil {
-		return nil, err
+		return nil, nil, err
 	}
 	return parser.parse()
 }
@@ -80,29 +81,31 @@ func newJavaArchiveParser(virtualPath string, reader io.Reader, detectNested boo
 }
 
 // parse the loaded archive and return all packages found.
-func (j *archiveParser) parse() ([]pkg.Package, error) {
+func (j *archiveParser) parse() ([]pkg.Package, []artifact.Relationship, error) {
-	var pkgs = make([]pkg.Package, 0)
+	var pkgs []pkg.Package
+	var relationships []artifact.Relationship
 
 	// find the parent package from the java manifest
 	parentPkg, err := j.discoverMainPackage()
 	if err != nil {
-		return nil, fmt.Errorf("could not generate package from %s: %w", j.virtualPath, err)
+		return nil, nil, fmt.Errorf("could not generate package from %s: %w", j.virtualPath, err)
 	}
 
 	// find aux packages from pom.properties/pom.xml and potentially modify the existing parentPkg
 	auxPkgs, err := j.discoverPkgsFromAllMavenFiles(parentPkg)
 	if err != nil {
-		return nil, err
+		return nil, nil, err
 	}
 	pkgs = append(pkgs, auxPkgs...)
 
 	if j.detectNested {
 		// find nested java archive packages
-		nestedPkgs, err := j.discoverPkgsFromNestedArchives(parentPkg)
+		nestedPkgs, nestedRelationships, err := j.discoverPkgsFromNestedArchives(parentPkg)
 		if err != nil {
-			return nil, err
+			return nil, nil, err
 		}
 		pkgs = append(pkgs, nestedPkgs...)
+		relationships = append(relationships, nestedRelationships...)
 	}
 
 	// lastly, add the parent package to the list (assuming the parent exists)
@@ -110,7 +113,7 @@ func (j *archiveParser) parse() ([]pkg.Package, error) {
 		pkgs = append([]pkg.Package{*parentPkg}, pkgs...)
 	}
 
-	return pkgs, nil
+	return pkgs, relationships, nil
 }
 
 // discoverMainPackage parses the root Java manifest used as the parent package to all discovered nested packages.
@@ -189,31 +192,32 @@ func (j *archiveParser) discoverPkgsFromAllMavenFiles(parentPkg *pkg.Package) ([
 
 // discoverPkgsFromNestedArchives finds Java archives within Java archives, returning all listed Java packages found and
 // associating each discovered package to the given parent package.
-func (j *archiveParser) discoverPkgsFromNestedArchives(parentPkg *pkg.Package) ([]pkg.Package, error) {
+func (j *archiveParser) discoverPkgsFromNestedArchives(parentPkg *pkg.Package) ([]pkg.Package, []artifact.Relationship, error) {
-	var pkgs = make([]pkg.Package, 0)
+	var pkgs []pkg.Package
+	var relationships []artifact.Relationship
 
 	// search and parse pom.properties files & fetch the contents
 	openers, err := file.ExtractFromZipToUniqueTempFile(j.archivePath, j.contentPath, j.fileManifest.GlobMatch(archiveFormatGlobs...)...)
 	if err != nil {
-		return nil, fmt.Errorf("unable to extract files from zip: %w", err)
+		return nil, nil, fmt.Errorf("unable to extract files from zip: %w", err)
 	}
 
 	// discover nested artifacts
 	for archivePath, archiveOpener := range openers {
 		archiveReadCloser, err := archiveOpener.Open()
 		if err != nil {
-			return nil, fmt.Errorf("unable to open archived file from tempdir: %w", err)
+			return nil, nil, fmt.Errorf("unable to open archived file from tempdir: %w", err)
 		}
 		nestedPath := fmt.Sprintf("%s:%s", j.virtualPath, archivePath)
-		nestedPkgs, err := parseJavaArchive(nestedPath, archiveReadCloser)
+		nestedPkgs, nestedRelationships, err := parseJavaArchive(nestedPath, archiveReadCloser)
 		if err != nil {
 			if closeErr := archiveReadCloser.Close(); closeErr != nil {
 				log.Warnf("unable to close archived file from tempdir: %+v", closeErr)
 			}
-			return nil, fmt.Errorf("unable to process nested java archive (%s): %w", archivePath, err)
+			return nil, nil, fmt.Errorf("unable to process nested java archive (%s): %w", archivePath, err)
 		}
 		if err = archiveReadCloser.Close(); err != nil {
-			return nil, fmt.Errorf("unable to close archived file from tempdir: %w", err)
+			return nil, nil, fmt.Errorf("unable to close archived file from tempdir: %w", err)
 		}
 
 		// attach the parent package to all discovered packages that are not already associated with a java archive
@@ -226,9 +230,11 @@ func (j *archiveParser) discoverPkgsFromNestedArchives(parentPkg *pkg.Package) (
 			}
 			pkgs = append(pkgs, p)
 		}
+
+		relationships = append(relationships, nestedRelationships...)
 	}
 
-	return pkgs, nil
+	return pkgs, relationships, nil
 }
 
 func pomPropertiesByParentPath(archivePath string, extractPaths []string, virtualPath string) (map[string]pkg.PomProperties, error) {

syft/pkg/cataloger/java/archive_parser_test.go

@@ -242,7 +242,7 @@ func TestParseJar(t *testing.T) {
 				t.Fatalf("should not have filed... %+v", err)
 			}
 
-			actual, err := parser.parse()
+			actual, _, err := parser.parse()
 			if err != nil {
 				t.Fatalf("failed to parse java archive: %+v", err)
 			}
@@ -507,7 +507,7 @@ func TestParseNestedJar(t *testing.T) {
 				t.Fatalf("failed to open fixture: %+v", err)
 			}
 
-			actual, err := parseJavaArchive(fixture.Name(), fixture)
+			actual, _, err := parseJavaArchive(fixture.Name(), fixture)
 			if err != nil {
 				t.Fatalf("failed to parse java archive: %+v", err)
 			}

syft/pkg/cataloger/javascript/parse_package_json.go

@@ -13,6 +13,7 @@ import (
 
 	"github.com/mitchellh/mapstructure"
 
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/pkg/cataloger/common"
 )
@@ -162,8 +163,8 @@ func licensesFromJSON(p PackageJSON) ([]string, error) {
 }
 
 // parsePackageJSON parses a package.json and returns the discovered JavaScript packages.
-func parsePackageJSON(_ string, reader io.Reader) ([]pkg.Package, error) {
+func parsePackageJSON(_ string, reader io.Reader) ([]pkg.Package, []artifact.Relationship, error) {
-	packages := make([]pkg.Package, 0)
+	var packages []pkg.Package
 	dec := json.NewDecoder(reader)
 
 	for {
@@ -171,17 +172,17 @@ func parsePackageJSON(_ string, reader io.Reader) ([]pkg.Package, error) {
 		if err := dec.Decode(&p); err == io.EOF {
 			break
 		} else if err != nil {
-			return nil, fmt.Errorf("failed to parse package.json file: %w", err)
+			return nil, nil, fmt.Errorf("failed to parse package.json file: %w", err)
 		}
 
 		if !p.hasNameAndVersionValues() {
 			log.Debug("encountered package.json file without a name and/or version field, ignoring this file")
-			return nil, nil
+			return nil, nil, nil
 		}
 
 		licenses, err := licensesFromJSON(p)
 		if err != nil {
-			return nil, fmt.Errorf("failed to parse package.json file: %w", err)
+			return nil, nil, fmt.Errorf("failed to parse package.json file: %w", err)
 		}
 
 		packages = append(packages, pkg.Package{
@@ -200,7 +201,7 @@ func parsePackageJSON(_ string, reader io.Reader) ([]pkg.Package, error) {
 		})
 	}
 
-	return packages, nil
+	return packages, nil, nil
 }
 
 func (p PackageJSON) hasNameAndVersionValues() bool {

syft/pkg/cataloger/javascript/parse_package_json_test.go

@@ -124,7 +124,7 @@ func TestParsePackageJSON(t *testing.T) {
 				t.Fatalf("failed to open fixture: %+v", err)
 			}
 
-			actual, err := parsePackageJSON("", fixture)
+			actual, _, err := parsePackageJSON("", fixture)
 			if err != nil {
 				t.Fatalf("failed to parse package-lock.json: %+v", err)
 			}
@@ -150,7 +150,8 @@ func TestParsePackageJSON_Partial(t *testing.T) { // see https://github.com/anch
 		t.Fatalf("failed to open fixture: %+v", err)
 	}
 
-	actual, err := parsePackageJSON("", fixture)
+	// TODO: no relationships are under test yet
+	actual, _, err := parsePackageJSON("", fixture)
 	if err != nil {
 		t.Fatalf("failed to parse package-lock.json: %+v", err)
 	}

syft/pkg/cataloger/javascript/parse_package_lock.go

@@ -5,6 +5,7 @@ import (
 	"fmt"
 	"io"
 
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/pkg/cataloger/common"
 )
@@ -28,11 +29,11 @@ type Dependency struct {
 }
 
 // parsePackageLock parses a package-lock.json and returns the discovered JavaScript packages.
-func parsePackageLock(path string, reader io.Reader) ([]pkg.Package, error) {
+func parsePackageLock(path string, reader io.Reader) ([]pkg.Package, []artifact.Relationship, error) {
 	// in the case we find package-lock.json files in the node_modules directories, skip those
 	// as the whole purpose of the lock file is for the specific dependencies of the root project
 	if pathContainsNodeModulesDirectory(path) {
-		return nil, nil
+		return nil, nil, nil
 	}
 
 	var packages []pkg.Package
@@ -43,7 +44,7 @@ func parsePackageLock(path string, reader io.Reader) ([]pkg.Package, error) {
 		if err := dec.Decode(&lock); err == io.EOF {
 			break
 		} else if err != nil {
-			return nil, fmt.Errorf("failed to parse package-lock.json file: %w", err)
+			return nil, nil, fmt.Errorf("failed to parse package-lock.json file: %w", err)
 		}
 		for name, pkgMeta := range lock.Dependencies {
 			packages = append(packages, pkg.Package{
@@ -55,5 +56,5 @@ func parsePackageLock(path string, reader io.Reader) ([]pkg.Package, error) {
 		}
 	}
 
-	return packages, nil
+	return packages, nil, nil
 }

syft/pkg/cataloger/javascript/parse_package_lock_test.go

@@ -109,7 +109,8 @@ func TestParsePackageLock(t *testing.T) {
 		t.Fatalf("failed to open fixture: %+v", err)
 	}
 
-	actual, err := parsePackageLock(fixture.Name(), fixture)
+	// TODO: no relationships are under test yet
+	actual, _, err := parsePackageLock(fixture.Name(), fixture)
 	if err != nil {
 		t.Fatalf("failed to parse package-lock.json: %+v", err)
 	}

syft/pkg/cataloger/javascript/parse_yarn_lock.go

@@ -7,6 +7,7 @@ import (
 	"regexp"
 
 	"github.com/anchore/syft/internal"
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/pkg/cataloger/common"
 )
@@ -34,11 +35,11 @@ const (
 	noVersion = ""
 )
 
-func parseYarnLock(path string, reader io.Reader) ([]pkg.Package, error) {
+func parseYarnLock(path string, reader io.Reader) ([]pkg.Package, []artifact.Relationship, error) {
 	// in the case we find yarn.lock files in the node_modules directories, skip those
 	// as the whole purpose of the lock file is for the specific dependencies of the project
 	if pathContainsNodeModulesDirectory(path) {
-		return nil, nil
+		return nil, nil, nil
 	}
 
 	var packages []pkg.Package
@@ -79,10 +80,10 @@ func parseYarnLock(path string, reader io.Reader) ([]pkg.Package, error) {
 	}
 
 	if err := scanner.Err(); err != nil {
-		return nil, fmt.Errorf("failed to parse yarn.lock file: %w", err)
+		return nil, nil, fmt.Errorf("failed to parse yarn.lock file: %w", err)
 	}
 
-	return packages, nil
+	return packages, nil, nil
 }
 
 func findPackageName(line string) string {

syft/pkg/cataloger/javascript/parse_yarn_lock_test.go

@@ -70,7 +70,8 @@ func TestParseYarnLock(t *testing.T) {
 		t.Fatalf("failed to open fixture: %+v", err)
 	}
 
-	actual, err := parseYarnLock(fixture.Name(), fixture)
+	// TODO: no relationships are under test yet
+	actual, _, err := parseYarnLock(fixture.Name(), fixture)
 	if err != nil {
 		t.Fatalf("failed to parse yarn.lock: %+v", err)
 	}

syft/pkg/cataloger/php/parse_composer_lock.go

@@ -5,6 +5,8 @@ import (
 	"fmt"
 	"io"
 
+	"github.com/anchore/syft/syft/artifact"
+
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/pkg/cataloger/common"
 )
@@ -23,7 +25,7 @@ type Dependency struct {
 var _ common.ParserFn = parseComposerLock
 
 // parseComposerLock is a parser function for Composer.lock contents, returning "Default" php packages discovered.
-func parseComposerLock(_ string, reader io.Reader) ([]pkg.Package, error) {
+func parseComposerLock(_ string, reader io.Reader) ([]pkg.Package, []artifact.Relationship, error) {
 	packages := make([]pkg.Package, 0)
 	dec := json.NewDecoder(reader)
 
@@ -32,7 +34,7 @@ func parseComposerLock(_ string, reader io.Reader) ([]pkg.Package, error) {
 		if err := dec.Decode(&lock); err == io.EOF {
 			break
 		} else if err != nil {
-			return nil, fmt.Errorf("failed to parse composer.lock file: %w", err)
+			return nil, nil, fmt.Errorf("failed to parse composer.lock file: %w", err)
 		}
 		for _, pkgMeta := range lock.Packages {
 			version := pkgMeta.Version
@@ -46,5 +48,5 @@ func parseComposerLock(_ string, reader io.Reader) ([]pkg.Package, error) {
 		}
 	}
 
-	return packages, nil
+	return packages, nil, nil
 }

syft/pkg/cataloger/php/parse_composer_lock_test.go

@@ -28,7 +28,8 @@ func TestParseComposerFileLock(t *testing.T) {
 		t.Fatalf("failed to open fixture: %+v", err)
 	}
 
-	actual, err := parseComposerLock(fixture.Name(), fixture)
+	// TODO: no relationships are under test yet
+	actual, _, err := parseComposerLock(fixture.Name(), fixture)
 	if err != nil {
 		t.Fatalf("failed to parse requirements: %+v", err)
 	}

syft/pkg/cataloger/python/package_cataloger.go

@@ -7,6 +7,7 @@ import (
 
 	"github.com/anchore/syft/internal"
 
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 
 	"github.com/anchore/syft/syft/source"
@@ -31,13 +32,13 @@ func (c *PackageCataloger) Name() string {
 }
 
 // Catalog is given an object to resolve file references and content, this function returns any discovered Packages after analyzing python egg and wheel installations.
-func (c *PackageCataloger) Catalog(resolver source.FileResolver) ([]pkg.Package, error) {
+func (c *PackageCataloger) Catalog(resolver source.FileResolver) ([]pkg.Package, []artifact.Relationship, error) {
 	var fileMatches []source.Location
 
 	for _, glob := range []string{eggMetadataGlob, wheelMetadataGlob, eggFileMetadataGlob} {
 		matches, err := resolver.FilesByGlob(glob)
 		if err != nil {
-			return nil, fmt.Errorf("failed to find files by glob: %s", glob)
+			return nil, nil, fmt.Errorf("failed to find files by glob: %s", glob)
 		}
 		fileMatches = append(fileMatches, matches...)
 	}
@@ -46,13 +47,13 @@ func (c *PackageCataloger) Catalog(resolver source.FileResolver) ([]pkg.Package,
 	for _, location := range fileMatches {
 		p, err := c.catalogEggOrWheel(resolver, location)
 		if err != nil {
-			return nil, fmt.Errorf("unable to catalog python package=%+v: %w", location.RealPath, err)
+			return nil, nil, fmt.Errorf("unable to catalog python package=%+v: %w", location.RealPath, err)
 		}
 		if p != nil {
 			pkgs = append(pkgs, *p)
 		}
 	}
-	return pkgs, nil
+	return pkgs, nil, nil
 }
 
 // catalogEggOrWheel takes the primary metadata file reference and returns the python package it represents.

syft/pkg/cataloger/python/package_cataloger_test.go

@@ -144,7 +144,7 @@ func TestPythonPackageWheelCataloger(t *testing.T) {
 
 			test.expectedPackage.Locations = locations
 
-			actual, err := NewPythonPackageCataloger().Catalog(resolver)
+			actual, _, err := NewPythonPackageCataloger().Catalog(resolver)
 			if err != nil {
 				t.Fatalf("failed to catalog python package: %+v", err)
 			}
@@ -173,7 +173,7 @@ func TestIgnorePackage(t *testing.T) {
 		t.Run(test.MetadataFixture, func(t *testing.T) {
 			resolver := source.NewMockResolverForPaths(test.MetadataFixture)
 
-			actual, err := NewPythonPackageCataloger().Catalog(resolver)
+			actual, _, err := NewPythonPackageCataloger().Catalog(resolver)
 			if err != nil {
 				t.Fatalf("failed to catalog python package: %+v", err)
 			}

syft/pkg/cataloger/python/parse_pipfile_lock.go

@@ -6,6 +6,7 @@ import (
 	"io"
 	"strings"
 
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/pkg/cataloger/common"
 )
@@ -37,7 +38,7 @@ type Dependency struct {
 var _ common.ParserFn = parsePipfileLock
 
 // parsePipfileLock is a parser function for Pipfile.lock contents, returning "Default" python packages discovered.
-func parsePipfileLock(_ string, reader io.Reader) ([]pkg.Package, error) {
+func parsePipfileLock(_ string, reader io.Reader) ([]pkg.Package, []artifact.Relationship, error) {
 	packages := make([]pkg.Package, 0)
 	dec := json.NewDecoder(reader)
 
@@ -46,7 +47,7 @@ func parsePipfileLock(_ string, reader io.Reader) ([]pkg.Package, error) {
 		if err := dec.Decode(&lock); err == io.EOF {
 			break
 		} else if err != nil {
-			return nil, fmt.Errorf("failed to parse Pipfile.lock file: %w", err)
+			return nil, nil, fmt.Errorf("failed to parse Pipfile.lock file: %w", err)
 		}
 		for name, pkgMeta := range lock.Default {
 			version := strings.TrimPrefix(pkgMeta.Version, "==")
@@ -59,5 +60,5 @@ func parsePipfileLock(_ string, reader io.Reader) ([]pkg.Package, error) {
 		}
 	}
 
-	return packages, nil
+	return packages, nil, nil
 }

syft/pkg/cataloger/python/parse_pipfile_lock_test.go

@@ -39,7 +39,8 @@ func TestParsePipFileLock(t *testing.T) {
 		t.Fatalf("failed to open fixture: %+v", err)
 	}
 
-	actual, err := parsePipfileLock(fixture.Name(), fixture)
+	// TODO: no relationships are under test yet
+	actual, _, err := parsePipfileLock(fixture.Name(), fixture)
 	if err != nil {
 		t.Fatalf("failed to parse requirements: %+v", err)
 	}

syft/pkg/cataloger/python/parse_poetry_lock.go

@@ -4,6 +4,7 @@ import (
 	"fmt"
 	"io"
 
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/pkg/cataloger/common"
 	"github.com/pelletier/go-toml"
@@ -13,17 +14,17 @@ import (
 var _ common.ParserFn = parsePoetryLock
 
 // parsePoetryLock is a parser function for poetry.lock contents, returning all python packages discovered.
-func parsePoetryLock(_ string, reader io.Reader) ([]pkg.Package, error) {
+func parsePoetryLock(_ string, reader io.Reader) ([]pkg.Package, []artifact.Relationship, error) {
 	tree, err := toml.LoadReader(reader)
 	if err != nil {
-		return nil, fmt.Errorf("unable to load poetry.lock for parsing: %v", err)
+		return nil, nil, fmt.Errorf("unable to load poetry.lock for parsing: %v", err)
 	}
 
 	metadata := PoetryMetadata{}
 	err = tree.Unmarshal(&metadata)
 	if err != nil {
-		return nil, fmt.Errorf("unable to parse poetry.lock: %v", err)
+		return nil, nil, fmt.Errorf("unable to parse poetry.lock: %v", err)
 	}
 
-	return metadata.Pkgs(), nil
+	return metadata.Pkgs(), nil, nil
 }

syft/pkg/cataloger/python/parse_poetry_lock_test.go

@@ -45,7 +45,8 @@ func TestParsePoetryLock(t *testing.T) {
 		t.Fatalf("failed to open fixture: %+v", err)
 	}
 
-	actual, err := parsePoetryLock(fixture.Name(), fixture)
+	// TODO: no relationships are under test yet
+	actual, _, err := parsePoetryLock(fixture.Name(), fixture)
 	if err != nil {
 		t.Error(err)
 	}

syft/pkg/cataloger/python/parse_requirements.go

@@ -6,6 +6,7 @@ import (
 	"io"
 	"strings"
 
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/pkg/cataloger/common"
 )
@@ -15,7 +16,7 @@ var _ common.ParserFn = parseRequirementsTxt
 
 // parseRequirementsTxt takes a Python requirements.txt file, returning all Python packages that are locked to a
 // specific version.
-func parseRequirementsTxt(_ string, reader io.Reader) ([]pkg.Package, error) {
+func parseRequirementsTxt(_ string, reader io.Reader) ([]pkg.Package, []artifact.Relationship, error) {
 	packages := make([]pkg.Package, 0)
 
 	scanner := bufio.NewScanner(reader)
@@ -55,10 +56,10 @@ func parseRequirementsTxt(_ string, reader io.Reader) ([]pkg.Package, error) {
 	}
 
 	if err := scanner.Err(); err != nil {
-		return nil, fmt.Errorf("failed to parse python requirements file: %w", err)
+		return nil, nil, fmt.Errorf("failed to parse python requirements file: %w", err)
 	}
 
-	return packages, nil
+	return packages, nil, nil
 }
 
 // removeTrailingComment takes a requirements.txt line and strips off comment strings.

syft/pkg/cataloger/python/parse_requirements_test.go

@@ -50,7 +50,8 @@ func TestParseRequirementsTxt(t *testing.T) {
 		t.Fatalf("failed to open fixture: %+v", err)
 	}
 
-	actual, err := parseRequirementsTxt(fixture.Name(), fixture)
+	// TODO: no relationships are under test yet
+	actual, _, err := parseRequirementsTxt(fixture.Name(), fixture)
 	if err != nil {
 		t.Fatalf("failed to parse requirements: %+v", err)
 	}

syft/pkg/cataloger/python/parse_setup.go

@@ -6,6 +6,7 @@ import (
 	"regexp"
 	"strings"
 
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/pkg/cataloger/common"
 )
@@ -19,7 +20,7 @@ var _ common.ParserFn = parseSetup
 //		" mypy2 == v0.770", ' mypy3== v0.770',  --> match(name=mypy2 version=v0.770), match(name=mypy3, version=v0.770)
 var pinnedDependency = regexp.MustCompile(`['"]\W?(\w+\W?==\W?[\w\.]*)`)
 
-func parseSetup(_ string, reader io.Reader) ([]pkg.Package, error) {
+func parseSetup(_ string, reader io.Reader) ([]pkg.Package, []artifact.Relationship, error) {
 	packages := make([]pkg.Package, 0)
 
 	scanner := bufio.NewScanner(reader)
@@ -46,5 +47,5 @@ func parseSetup(_ string, reader io.Reader) ([]pkg.Package, error) {
 		}
 	}
 
-	return packages, nil
+	return packages, nil, nil
 }

syft/pkg/cataloger/python/parse_setup_test.go

@@ -45,7 +45,7 @@ func TestParseSetup(t *testing.T) {
 		t.Fatalf("failed to open fixture: %+v", err)
 	}
 
-	actual, err := parseSetup(fixture.Name(), fixture)
+	actual, _, err := parseSetup(fixture.Name(), fixture)
 	if err != nil {
 		t.Fatalf("failed to parse requirements: %+v", err)
 	}

syft/pkg/cataloger/rpmdb/cataloger.go

@@ -8,6 +8,7 @@ import (
 
 	"github.com/anchore/syft/internal"
 
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/source"
 )
@@ -27,24 +28,26 @@ func (c *Cataloger) Name() string {
 }
 
 // Catalog is given an object to resolve file references and content, this function returns any discovered Packages after analyzing rpm db installation.
-func (c *Cataloger) Catalog(resolver source.FileResolver) ([]pkg.Package, error) {
+func (c *Cataloger) Catalog(resolver source.FileResolver) ([]pkg.Package, []artifact.Relationship, error) {
 	fileMatches, err := resolver.FilesByGlob(pkg.RpmDBGlob)
 	if err != nil {
-		return nil, fmt.Errorf("failed to find rpmdb's by glob: %w", err)
+		return nil, nil, fmt.Errorf("failed to find rpmdb's by glob: %w", err)
 	}
 
 	var pkgs []pkg.Package
 	for _, location := range fileMatches {
 		dbContentReader, err := resolver.FileContentsByLocation(location)
 		if err != nil {
-			return nil, err
+			return nil, nil, err
 		}
 
-		pkgs, err = parseRpmDB(resolver, location, dbContentReader)
+		discoveredPkgs, err := parseRpmDB(resolver, location, dbContentReader)
 		internal.CloseAndLogError(dbContentReader, location.VirtualPath)
 		if err != nil {
-			return nil, fmt.Errorf("unable to catalog rpmdb package=%+v: %w", location.RealPath, err)
+			return nil, nil, fmt.Errorf("unable to catalog rpmdb package=%+v: %w", location.RealPath, err)
 		}
+
+		pkgs = append(pkgs, discoveredPkgs...)
 	}
-	return pkgs, nil
+	return pkgs, nil, nil
 }

syft/pkg/cataloger/ruby/parse_gemfile_lock.go

@@ -6,6 +6,7 @@ import (
 	"strings"
 
 	"github.com/anchore/syft/internal"
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/pkg/cataloger/common"
 )
@@ -16,7 +17,7 @@ var _ common.ParserFn = parseGemFileLockEntries
 var sectionsOfInterest = internal.NewStringSetFromSlice([]string{"GEM"})
 
 // parseGemFileLockEntries is a parser function for Gemfile.lock contents, returning all Gems discovered.
-func parseGemFileLockEntries(_ string, reader io.Reader) ([]pkg.Package, error) {
+func parseGemFileLockEntries(_ string, reader io.Reader) ([]pkg.Package, []artifact.Relationship, error) {
 	pkgs := make([]pkg.Package, 0)
 	scanner := bufio.NewScanner(reader)
 
@@ -49,9 +50,9 @@ func parseGemFileLockEntries(_ string, reader io.Reader) ([]pkg.Package, error)
 		}
 	}
 	if err := scanner.Err(); err != nil {
-		return nil, err
+		return nil, nil, err
 	}
-	return pkgs, nil
+	return pkgs, nil, nil
 }
 
 func isDependencyLine(line string) bool {

syft/pkg/cataloger/ruby/parse_gemfile_lock_test.go

@@ -68,7 +68,8 @@ func TestParseGemfileLockEntries(t *testing.T) {
 		t.Fatalf("failed to open fixture: %+v", err)
 	}
 
-	actual, err := parseGemFileLockEntries(fixture.Name(), fixture)
+	// TODO: no relationships are under test yet
+	actual, _, err := parseGemFileLockEntries(fixture.Name(), fixture)
 	if err != nil {
 		t.Fatalf("failed to parse gemfile lock: %+v", err)
 	}

syft/pkg/cataloger/ruby/parse_gemspec.go

@@ -12,6 +12,7 @@ import (
 
 	"github.com/mitchellh/mapstructure"
 
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/pkg/cataloger/common"
 )
@@ -60,7 +61,7 @@ func processList(s string) []string {
 	return results
 }
 
-func parseGemSpecEntries(_ string, reader io.Reader) ([]pkg.Package, error) {
+func parseGemSpecEntries(_ string, reader io.Reader) ([]pkg.Package, []artifact.Relationship, error) {
 	var pkgs []pkg.Package
 	var fields = make(map[string]interface{})
 	scanner := bufio.NewScanner(reader)
@@ -93,7 +94,7 @@ func parseGemSpecEntries(_ string, reader io.Reader) ([]pkg.Package, error) {
 	if fields["name"] != "" && fields["version"] != "" {
 		var metadata pkg.GemMetadata
 		if err := mapstructure.Decode(fields, &metadata); err != nil {
-			return nil, fmt.Errorf("unable to decode gem metadata: %w", err)
+			return nil, nil, fmt.Errorf("unable to decode gem metadata: %w", err)
 		}
 
 		pkgs = append(pkgs, pkg.Package{
@@ -107,7 +108,7 @@ func parseGemSpecEntries(_ string, reader io.Reader) ([]pkg.Package, error) {
 		})
 	}
 
-	return pkgs, nil
+	return pkgs, nil, nil
 }
 
 // renderUtf8 takes any string escaped string sub-sections from the ruby string and replaces those sections with the UTF8 runes.

syft/pkg/cataloger/ruby/parse_gemspec_test.go

@@ -31,7 +31,8 @@ func TestParseGemspec(t *testing.T) {
 		t.Fatalf("failed to open fixture: %+v", err)
 	}
 
-	actual, err := parseGemSpecEntries(fixture.Name(), fixture)
+	// TODO: no relationships are under test yet
+	actual, _, err := parseGemSpecEntries(fixture.Name(), fixture)
 	if err != nil {
 		t.Fatalf("failed to parse gemspec: %+v", err)
 	}

syft/pkg/cataloger/rust/parse_cargo_lock.go

@@ -4,6 +4,7 @@ import (
 	"fmt"
 	"io"
 
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/pkg/cataloger/common"
 	"github.com/pelletier/go-toml"
@@ -13,17 +14,17 @@ import (
 var _ common.ParserFn = parseCargoLock
 
 // parseCargoLock is a parser function for Cargo.lock contents, returning all rust cargo crates discovered.
-func parseCargoLock(_ string, reader io.Reader) ([]pkg.Package, error) {
+func parseCargoLock(_ string, reader io.Reader) ([]pkg.Package, []artifact.Relationship, error) {
 	tree, err := toml.LoadReader(reader)
 	if err != nil {
-		return nil, fmt.Errorf("unable to load Cargo.lock for parsing: %v", err)
+		return nil, nil, fmt.Errorf("unable to load Cargo.lock for parsing: %v", err)
 	}
 
 	metadata := CargoMetadata{}
 	err = tree.Unmarshal(&metadata)
 	if err != nil {
-		return nil, fmt.Errorf("unable to parse Cargo.lock: %v", err)
+		return nil, nil, fmt.Errorf("unable to parse Cargo.lock: %v", err)
 	}
 
-	return metadata.Pkgs(), nil
+	return metadata.Pkgs(), nil, nil
 }

syft/pkg/cataloger/rust/parse_cargo_lock_test.go

@@ -177,7 +177,8 @@ func TestParseCargoLock(t *testing.T) {
 		t.Fatalf("failed to open fixture: %+v", err)
 	}
 
-	actual, err := parseCargoLock(fixture.Name(), fixture)
+	// TODO: no relationships are under test yet
+	actual, _, err := parseCargoLock(fixture.Name(), fixture)
 	if err != nil {
 		t.Error(err)
 	}

syft/pkg/cpe.go

@@ -45,5 +45,8 @@ func MustCPE(cpeStr string) CPE {
 
 func normalizeCpeField(field string) string {
 	// keep dashes and forward slashes unescaped
+	if field == "*" {
+		return wfn.Any
+	}
 	return strings.ReplaceAll(wfn.StripSlashes(field), `\/`, "/")
 }

syft/pkg/cpe_test.go

@@ -1,6 +1,10 @@
 package pkg
 
-import "testing"
+import (
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+)
 
 func must(c CPE, e error) CPE {
 	if e != nil {
@@ -46,3 +50,33 @@ func TestNewCPE(t *testing.T) {
 		})
 	}
 }
+
+func Test_normalizeCpeField(t *testing.T) {
+
+	tests := []struct {
+		field    string
+		expected string
+	}{
+		{
+			field:    "something",
+			expected: "something",
+		},
+		{
+			field:    "some\\thing",
+			expected: `some\thing`,
+		},
+		{
+			field:    "*",
+			expected: "",
+		},
+		{
+			field:    "",
+			expected: "",
+		},
+	}
+	for _, test := range tests {
+		t.Run(test.field, func(t *testing.T) {
+			assert.Equal(t, test.expected, normalizeCpeField(test.field))
+		})
+	}
+}

syft/pkg/id.go

@@ -1,4 +0,0 @@
-package pkg
-
-// ID represents a unique value for each package added to a package catalog.
-type ID string

syft/pkg/java_metadata.go

@@ -21,7 +21,7 @@ type JavaMetadata struct {
 	Manifest      *JavaManifest  `mapstructure:"Manifest" json:"manifest,omitempty"`
 	PomProperties *PomProperties `mapstructure:"PomProperties" json:"pomProperties,omitempty"`
 	PomProject    *PomProject    `mapstructure:"PomProject" json:"pomProject,omitempty"`
-	Parent        *Package       `json:"-"`
+	Parent        *Package       `hash:"ignore" json:"-"` // note: the parent cannot be included in the minimal definition of uniqueness since this field is not reproducible in an encode-decode cycle (is lossy).
 }
 
 // PomProperties represents the fields of interest extracted from a Java archive's pom.properties file.

syft/pkg/package.go

@@ -6,41 +6,39 @@ package pkg
 import (
 	"fmt"
 
+	"github.com/anchore/syft/internal/log"
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/source"
-	"github.com/mitchellh/hashstructure"
 )
 
 // Package represents an application or library that has been bundled into a distributable format.
 // TODO: if we ignore FoundBy for ID generation should we merge the field to show it was found in two places?
 type Package struct {
-	ID        ID                `hash:"ignore"` // uniquely identifies a package, set by the cataloger
+	Name         string            // the package name
-	Name      string            // the package name
+	Version      string            // the version of the package
-	Version   string            // the version of the package
+	FoundBy      string            // the specific cataloger that discovered this package
-	FoundBy   string            // the specific cataloger that discovered this package
+	Locations    []source.Location // the locations that lead to the discovery of this package (note: this is not necessarily the locations that make up this package)
-	Locations []source.Location // the locations that lead to the discovery of this package (note: this is not necessarily the locations that make up this package)
+	Licenses     []string          // licenses discovered with the package metadata
-	// TODO: should we move licenses into metadata?
+	Language     Language          // the language ecosystem this package belongs to (e.g. JavaScript, Python, etc)
-	Licenses     []string     // licenses discovered with the package metadata
+	Type         Type              // the package type (e.g. Npm, Yarn, Python, Rpm, Deb, etc)
-	Language     Language     // the language ecosystem this package belongs to (e.g. JavaScript, Python, etc)
+	CPEs         []CPE             // all possible Common Platform Enumerators
-	Type         Type         // the package type (e.g. Npm, Yarn, Python, Rpm, Deb, etc)
+	PURL         string            // the Package URL (see https://github.com/package-url/purl-spec)
-	CPEs         []CPE        // all possible Common Platform Enumerators
+	MetadataType MetadataType      // the shape of the additional data in the "metadata" field
-	PURL         string       // the Package URL (see https://github.com/package-url/purl-spec)
+	Metadata     interface{}       // additional data found while parsing the package source
-	MetadataType MetadataType // the shape of the additional data in the "metadata" field
+}
-	Metadata     interface{}  // additional data found while parsing the package source
+
+func (p Package) ID() artifact.ID {
+	f, err := artifact.IDFromHash(p)
+	if err != nil {
+		// TODO: what to do in this case?
+		log.Warnf("unable to get fingerprint of package=%s@%s: %+v", p.Name, p.Version, err)
+		return ""
+	}
+
+	return f
 }
 
 // Stringer to represent a package.
 func (p Package) String() string {
 	return fmt.Sprintf("Pkg(type=%s, name=%s, version=%s)", p.Type, p.Name, p.Version)
 }
-
-func (p Package) Fingerprint() (string, error) {
-	f, err := hashstructure.Hash(p, &hashstructure.HashOptions{
-		ZeroNil:      true,
-		SlicesAsSets: true,
-	})
-	if err != nil {
-		return "", fmt.Errorf("could not build package fingerprint for: %s version: %s", p.Name, p.Version)
-	}
-
-	return fmt.Sprint(f), nil
-}

syft/pkg/package_test.go

@@ -9,7 +9,6 @@ import (
 
 func TestFingerprint(t *testing.T) {
 	originalPkg := Package{
-		ID:      "π",
 		Name:    "pi",
 		Version: "3.14",
 		FoundBy: "Archimedes",
@@ -190,10 +189,10 @@ func TestFingerprint(t *testing.T) {
 	for _, test := range tests {
 		t.Run(test.name, func(t *testing.T) {
 			transformedPkg := test.transform(originalPkg)
-			originalFingerprint, err := originalPkg.Fingerprint()
+			originalFingerprint := originalPkg.ID()
-			assert.NoError(t, err, "expected no error on package fingerprint")
+			assert.NotEmpty(t, originalFingerprint)
-			transformedFingerprint, err := transformedPkg.Fingerprint()
+			transformedFingerprint := transformedPkg.ID()
-			assert.NoError(t, err, "expected no error on package fingerprint")
+			assert.NotEmpty(t, transformedFingerprint)
 
 			if test.expectIdentical {
 				assert.Equal(t, originalFingerprint, transformedFingerprint)

syft/pkg/relationship.go

@@ -1,20 +0,0 @@
-package pkg
-
-const (
-	// OwnershipByFileOverlapRelationship indicates that the parent package owns the child package made evident by the set of provided files
-	OwnershipByFileOverlapRelationship RelationshipType = "ownership-by-file-overlap"
-)
-
-type RelationshipType string
-
-type Relationship struct {
-	Parent   ID
-	Child    ID
-	Type     RelationshipType
-	Metadata interface{}
-}
-
-// TODO: as more relationships are added, this function signature will probably accommodate selection
-func NewRelationships(catalog *Catalog) []Relationship {
-	return ownershipByFilesRelationships(catalog)
-}

syft/pkg/relationships.go

@@ -0,0 +1,8 @@
+package pkg
+
+import "github.com/anchore/syft/syft/artifact"
+
+// TODO: as more relationships are added, this function signature will probably accommodate selection
+func NewRelationships(catalog *Catalog) []artifact.Relationship {
+	return RelationshipsByFileOwnership(catalog)
+}

syft/pkg/relationships_by_file_ownership.go

@@ -2,6 +2,7 @@ package pkg
 
 import (
 	"github.com/anchore/syft/internal/log"
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/bmatcuk/doublestar/v2"
 	"github.com/scylladb/go-set/strset"
 )
@@ -20,17 +21,19 @@ type ownershipByFilesMetadata struct {
 	Files []string `json:"files"`
 }
 
-func ownershipByFilesRelationships(catalog *Catalog) []Relationship {
+// RelationshipsByFileOwnership creates a package-to-package relationship based on discovering which packages have
+// evidence locations that overlap with ownership claim from another package's package manager metadata.
+func RelationshipsByFileOwnership(catalog *Catalog) []artifact.Relationship {
 	var relationships = findOwnershipByFilesRelationships(catalog)
 
-	var edges []Relationship
+	var edges []artifact.Relationship
 	for parent, children := range relationships {
 		for child, files := range children {
-			edges = append(edges, Relationship{
+			edges = append(edges, artifact.Relationship{
-				Parent: parent,
+				From: catalog.byID[parent],
-				Child:  child,
+				To:   catalog.byID[child],
-				Type:   OwnershipByFileOverlapRelationship,
+				Type: artifact.OwnershipByFileOverlapRelationship,
-				Metadata: ownershipByFilesMetadata{
+				Data: ownershipByFilesMetadata{
 					Files: files.List(),
 				},
 			})
@@ -42,14 +45,15 @@ func ownershipByFilesRelationships(catalog *Catalog) []Relationship {
 
 // findOwnershipByFilesRelationships find overlaps in file ownership with a file that defines another package. Specifically, a .Location.Path of
 // a package is found to be owned by another (from the owner's .Metadata.Files[]).
-func findOwnershipByFilesRelationships(catalog *Catalog) map[ID]map[ID]*strset.Set {
+func findOwnershipByFilesRelationships(catalog *Catalog) map[artifact.ID]map[artifact.ID]*strset.Set {
-	var relationships = make(map[ID]map[ID]*strset.Set)
+	var relationships = make(map[artifact.ID]map[artifact.ID]*strset.Set)
 
 	if catalog == nil {
 		return relationships
 	}
 
 	for _, candidateOwnerPkg := range catalog.Sorted() {
+		id := candidateOwnerPkg.ID()
 		if candidateOwnerPkg.Metadata == nil {
 			continue
 		}
@@ -68,17 +72,18 @@ func findOwnershipByFilesRelationships(catalog *Catalog) map[ID]map[ID]*strset.S
 
 			// look for package(s) in the catalog that may be owned by this package and mark the relationship
 			for _, subPackage := range catalog.PackagesByPath(ownedFilePath) {
-				if subPackage.ID == candidateOwnerPkg.ID {
+				subID := subPackage.ID()
+				if subID == id {
 					continue
 				}
-				if _, exists := relationships[candidateOwnerPkg.ID]; !exists {
+				if _, exists := relationships[id]; !exists {
-					relationships[candidateOwnerPkg.ID] = make(map[ID]*strset.Set)
+					relationships[id] = make(map[artifact.ID]*strset.Set)
 				}
 
-				if _, exists := relationships[candidateOwnerPkg.ID][subPackage.ID]; !exists {
+				if _, exists := relationships[id][subID]; !exists {
-					relationships[candidateOwnerPkg.ID][subPackage.ID] = strset.New()
+					relationships[id][subID] = strset.New()
 				}
-				relationships[candidateOwnerPkg.ID][subPackage.ID].Add(ownedFilePath)
+				relationships[id][subID].Add(ownedFilePath)
 			}
 		}
 	}

syft/pkg/relationships_by_file_ownership_test.go

@@ -3,21 +3,21 @@ package pkg
 import (
 	"testing"
 
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/source"
-	"github.com/go-test/deep"
+	"github.com/stretchr/testify/assert"
 )
 
 func TestOwnershipByFilesRelationship(t *testing.T) {
+
 	tests := []struct {
-		name              string
+		name  string
-		pkgs              []Package
+		setup func(t testing.TB) ([]Package, []artifact.Relationship)
-		expectedRelations []Relationship
 	}{
 		{
 			name: "owns-by-real-path",
-			pkgs: []Package{
+			setup: func(t testing.TB) ([]Package, []artifact.Relationship) {
-				{
+				parent := Package{
-					ID: "parent",
 					Locations: []source.Location{
 						{
 							RealPath:    "/a/path",
@@ -37,9 +37,9 @@ func TestOwnershipByFilesRelationship(t *testing.T) {
 							{Path: "/d/path"},
 						},
 					},
-				},
+				}
-				{
+
-					ID: "child",
+				child := Package{
 					Locations: []source.Location{
 						{
 							RealPath:    "/c/path",
@@ -51,26 +51,26 @@ func TestOwnershipByFilesRelationship(t *testing.T) {
 						},
 					},
 					Type: NpmPkg,
-				},
+				}
-			},
+
-			expectedRelations: []Relationship{
+				relationship := artifact.Relationship{
-				{
+					From: parent,
-					Parent: "parent",
+					To:   child,
-					Child:  "child",
+					Type: artifact.OwnershipByFileOverlapRelationship,
-					Type:   OwnershipByFileOverlapRelationship,
+					Data: ownershipByFilesMetadata{
-					Metadata: ownershipByFilesMetadata{
 						Files: []string{
 							"/d/path",
 						},
 					},
-				},
+				}
+
+				return []Package{parent, child}, []artifact.Relationship{relationship}
 			},
 		},
 		{
 			name: "owns-by-virtual-path",
-			pkgs: []Package{
+			setup: func(t testing.TB) ([]Package, []artifact.Relationship) {
-				{
+				parent := Package{
-					ID: "parent",
 					Locations: []source.Location{
 						{
 							RealPath:    "/a/path",
@@ -90,9 +90,9 @@ func TestOwnershipByFilesRelationship(t *testing.T) {
 							{Path: "/another/path"},
 						},
 					},
-				},
+				}
-				{
+
-					ID: "child",
+				child := Package{
 					Locations: []source.Location{
 						{
 							RealPath:    "/c/path",
@@ -104,26 +104,25 @@ func TestOwnershipByFilesRelationship(t *testing.T) {
 						},
 					},
 					Type: NpmPkg,
-				},
+				}
-			},
+
-			expectedRelations: []Relationship{
+				relationship := artifact.Relationship{
-				{
+					From: parent,
-					Parent: "parent",
+					To:   child,
-					Child:  "child",
+					Type: artifact.OwnershipByFileOverlapRelationship,
-					Type:   OwnershipByFileOverlapRelationship,
+					Data: ownershipByFilesMetadata{
-					Metadata: ownershipByFilesMetadata{
 						Files: []string{
 							"/another/path",
 						},
 					},
-				},
+				}
+				return []Package{parent, child}, []artifact.Relationship{relationship}
 			},
 		},
 		{
 			name: "ignore-empty-path",
-			pkgs: []Package{
+			setup: func(t testing.TB) ([]Package, []artifact.Relationship) {
-				{
+				parent := Package{
-					ID: "parent",
 					Locations: []source.Location{
 						{
 							RealPath:    "/a/path",
@@ -143,9 +142,9 @@ func TestOwnershipByFilesRelationship(t *testing.T) {
 							{Path: ""},
 						},
 					},
-				},
+				}
-				{
+
-					ID: "child",
+				child := Package{
 					Locations: []source.Location{
 						{
 							RealPath:    "/c/path",
@@ -157,18 +156,26 @@ func TestOwnershipByFilesRelationship(t *testing.T) {
 						},
 					},
 					Type: NpmPkg,
-				},
+				}
+
+				return []Package{parent, child}, nil
 			},
 		},
 	}
 
 	for _, test := range tests {
 		t.Run(test.name, func(t *testing.T) {
-			c := NewCatalog(test.pkgs...)
+			pkgs, expectedRelations := test.setup(t)
-			relationships := ownershipByFilesRelationships(c)
+			c := NewCatalog(pkgs...)
+			relationships := RelationshipsByFileOwnership(c)
 
-			for _, d := range deep.Equal(test.expectedRelations, relationships) {
+			assert.Len(t, relationships, len(expectedRelations))
-				t.Errorf("diff: %+v", d)
+			for idx, expectedRelationship := range expectedRelations {
+				actualRelationship := relationships[idx]
+				assert.Equal(t, expectedRelationship.From.ID(), actualRelationship.From.ID())
+				assert.Equal(t, expectedRelationship.To.ID(), actualRelationship.To.ID())
+				assert.Equal(t, expectedRelationship.Type, actualRelationship.Type)
+				assert.Equal(t, expectedRelationship.Data, actualRelationship.Data)
 			}
 		})
 	}

syft/sbom/sbom.go

@@ -1,6 +1,7 @@
 package sbom
 
 import (
+	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/distro"
 	"github.com/anchore/syft/syft/file"
 	"github.com/anchore/syft/syft/pkg"
@@ -8,8 +9,9 @@ import (
 )
 
 type SBOM struct {
-	Artifacts Artifacts
+	Artifacts     Artifacts
-	Source    source.Metadata
+	Relationships []artifact.Relationship
+	Source        source.Metadata
 }
 
 type Artifacts struct {

syft/source/location.go

@@ -3,19 +3,22 @@ package source
 import (
 	"fmt"
 
-	"github.com/anchore/syft/internal/log"
-
 	"github.com/anchore/stereoscope/pkg/file"
 	"github.com/anchore/stereoscope/pkg/image"
+	"github.com/anchore/syft/internal/log"
+	"github.com/anchore/syft/syft/artifact"
 )
 
 // Location represents a path relative to a particular filesystem resolved to a specific file.Reference. This struct is used as a key
-// in content fetching to uniquely identify a file relative to a request (the VirtualPath).
+// in content fetching to uniquely identify a file relative to a request (the VirtualPath). Note that the VirtualPath
+// and ref are ignored fields when using github.com/mitchellh/hashstructure. The reason for this is to ensure that
+// only the minimally expressible fields of a location are baked into the uniqueness of a Location. Since VirutalPath
+// and ref are not captured in JSON output they cannot be included in this minimal definition.
 type Location struct {
 	RealPath     string         `json:"path"`              // The path where all path ancestors have no hardlinks / symlinks
-	VirtualPath  string         `json:"-"`                 // The path to the file which may or may not have hardlinks / symlinks
+	VirtualPath  string         `hash:"ignore" json:"-"`   // The path to the file which may or may not have hardlinks / symlinks
 	FileSystemID string         `json:"layerID,omitempty"` // An ID representing the filesystem. For container images this is a layer digest, directories or root filesystem this is blank.
-	ref          file.Reference // The file reference relative to the stereoscope.FileCatalog that has more information about this location.
+	ref          file.Reference `hash:"ignore"`            // The file reference relative to the stereoscope.FileCatalog that has more information about this location.
 }
 
 // NewLocation creates a new Location representing a path without denoting a filesystem or FileCatalog reference.
@@ -70,3 +73,14 @@ func (l Location) String() string {
 	}
 	return fmt.Sprintf("Location<%s>", str)
 }
+
+func (l Location) ID() artifact.ID {
+	f, err := artifact.IDFromHash(l)
+	if err != nil {
+		// TODO: what to do in this case?
+		log.Warnf("unable to get fingerprint of location=%+v: %+v", l, err)
+		return ""
+	}
+
+	return f
+}

syft/test-fixtures/pkgs/project/package-lock.json

@@ -1,52 +0,0 @@
-{
-  "name": "npm-lock",
-  "version": "1.0.0",
-  "lockfileVersion": 1,
-  "requires": true,
-  "dependencies": {
-    "collapse-white-space": {
-      "version": "2.0.0",
-      "resolved": "https://registry.npmjs.org/collapse-white-space/-/collapse-white-space-2.0.0.tgz",
-      "integrity": "sha512-eh9krktAIMDL0KHuN7WTBJ/0PMv8KUvfQRBkIlGmW61idRM2DJjgd1qXEPr4wyk2PimZZeNww3RVYo6CMvDGlg=="
-    },
-    "end-of-stream": {
-      "version": "1.4.4",
-      "resolved": "https://registry.npmjs.org/end-of-stream/-/end-of-stream-1.4.4.tgz",
-      "integrity": "sha512-+uw1inIHVPQoaVuHzRyXd21icM+cnt4CzD5rW+NC1wjOUSTOs+Te7FOv7AhN7vS9x/oIyhLP5PR1H+phQAHu5Q==",
-      "dev": true,
-      "requires": {
-        "once": "^1.4.0"
-      }
-    },
-    "insert-css": {
-      "version": "2.0.0",
-      "resolved": "https://registry.npmjs.org/insert-css/-/insert-css-2.0.0.tgz",
-      "integrity": "sha1-610Ql7dUL0x56jBg067gfQU4gPQ="
-    },
-    "once": {
-      "version": "1.4.0",
-      "resolved": "https://registry.npmjs.org/once/-/once-1.4.0.tgz",
-      "integrity": "sha1-WDsap3WWHUsROsF9nFC6753Xa9E=",
-      "dev": true,
-      "requires": {
-        "wrappy": "1"
-      }
-    },
-    "pump": {
-      "version": "3.0.0",
-      "resolved": "https://registry.npmjs.org/pump/-/pump-3.0.0.tgz",
-      "integrity": "sha512-LwZy+p3SFs1Pytd/jYct4wpv49HiYCqd9Rlc5ZVdk0V+8Yzv6jR5Blk3TRmPL1ft69TxP0IMZGJ+WPFU2BFhww==",
-      "dev": true,
-      "requires": {
-        "end-of-stream": "^1.1.0",
-        "once": "^1.3.1"
-      }
-    },
-    "wrappy": {
-      "version": "1.0.2",
-      "resolved": "https://registry.npmjs.org/wrappy/-/wrappy-1.0.2.tgz",
-      "integrity": "sha1-tSQ9jz7BqjXxNkYFvA0QNuMKtp8=",
-      "dev": true
-    }
-  }
-}

test/cli/trait_assertions_test.go

@@ -2,13 +2,11 @@ package cli
 
 import (
 	"encoding/json"
-	"fmt"
 	"regexp"
 	"strings"
 	"testing"
 
 	"github.com/acarl005/stripansi"
-	"github.com/anchore/syft/syft/source"
 )
 
 type traitAssertion func(tb testing.TB, stdout, stderr string, rc int)
@@ -29,17 +27,17 @@ func assertTableReport(tb testing.TB, stdout, _ string, _ int) {
 	}
 }
 
-func assertScope(scope source.Scope) traitAssertion {
+//func assertScope(scope source.Scope) traitAssertion {
-	return func(tb testing.TB, stdout, stderr string, rc int) {
+//	return func(tb testing.TB, stdout, stderr string, rc int) {
-		tb.Helper()
+//		tb.Helper()
-		// we can only verify source with the json report
+//		// we can only verify source with the json report
-		assertJsonReport(tb, stdout, stderr, rc)
+//		assertJsonReport(tb, stdout, stderr, rc)
-
+//
-		if !strings.Contains(stdout, fmt.Sprintf(`"scope": "%s"`, scope.String())) {
+//		if !strings.Contains(stdout, fmt.Sprintf(`"scope": "%s"`, scope.String())) {
-			tb.Errorf("JSON report did not indicate the %q scope", scope)
+//			tb.Errorf("JSON report did not indicate the %q scope", scope)
-		}
+//		}
-	}
+//	}
-}
+//}
 
 func assertLoggingLevel(level string) traitAssertion {
 	// match examples:

test/integration/catalog_packages_test.go

@@ -37,7 +37,7 @@ func BenchmarkImagePackageCatalogers(b *testing.B) {
 
 		b.Run(c.Name(), func(b *testing.B) {
 			for i := 0; i < b.N; i++ {
-				pc, err = cataloger.Catalog(resolver, theDistro, c)
+				pc, _, err = cataloger.Catalog(resolver, theDistro, c)
 				if err != nil {
 					b.Fatalf("failure during benchmark: %+v", err)
 				}
@@ -49,7 +49,7 @@ func BenchmarkImagePackageCatalogers(b *testing.B) {
 }
 
 func TestPkgCoverageImage(t *testing.T) {
-	catalog, _, _ := catalogFixtureImage(t, "image-pkg-coverage")
+	sbom, _ := catalogFixtureImage(t, "image-pkg-coverage")
 
 	observedLanguages := internal.NewStringSet()
 	definedLanguages := internal.NewStringSet()
@@ -82,7 +82,7 @@ func TestPkgCoverageImage(t *testing.T) {
 		t.Run(c.name, func(t *testing.T) {
 			pkgCount := 0
 
-			for a := range catalog.Enumerate(c.pkgType) {
+			for a := range sbom.Artifacts.PackageCatalog.Enumerate(c.pkgType) {
 
 				if a.Language.String() != "" {
 					observedLanguages.Add(a.Language.String())
@@ -110,7 +110,7 @@ func TestPkgCoverageImage(t *testing.T) {
 
 			if pkgCount != len(c.pkgInfo)+c.duplicates {
 				t.Logf("Discovered packages of type %+v", c.pkgType)
-				for a := range catalog.Enumerate(c.pkgType) {
+				for a := range sbom.Artifacts.PackageCatalog.Enumerate(c.pkgType) {
 					t.Log("   ", a)
 				}
 				t.Fatalf("unexpected package count: %d!=%d", pkgCount, len(c.pkgInfo))
@@ -137,7 +137,7 @@ func TestPkgCoverageImage(t *testing.T) {
 }
 
 func TestPkgCoverageDirectory(t *testing.T) {
-	catalog, _, _ := catalogDirectory(t, "test-fixtures/image-pkg-coverage")
+	sbom, _ := catalogDirectory(t, "test-fixtures/image-pkg-coverage")
 
 	observedLanguages := internal.NewStringSet()
 	definedLanguages := internal.NewStringSet()
@@ -159,7 +159,7 @@ func TestPkgCoverageDirectory(t *testing.T) {
 		t.Run(test.name, func(t *testing.T) {
 			actualPkgCount := 0
 
-			for actualPkg := range catalog.Enumerate(test.pkgType) {
+			for actualPkg := range sbom.Artifacts.PackageCatalog.Enumerate(test.pkgType) {
 
 				observedLanguages.Add(actualPkg.Language.String())
 				observedPkgs.Add(string(actualPkg.Type))
@@ -184,7 +184,7 @@ func TestPkgCoverageDirectory(t *testing.T) {
 			}
 
 			if actualPkgCount != len(test.pkgInfo)+test.duplicates {
-				for actualPkg := range catalog.Enumerate(test.pkgType) {
+				for actualPkg := range sbom.Artifacts.PackageCatalog.Enumerate(test.pkgType) {
 					t.Log("   ", actualPkg)
 				}
 				t.Fatalf("unexpected package count: %d!=%d", actualPkgCount, len(test.pkgInfo))

test/integration/distro_test.go

@@ -8,14 +8,14 @@ import (
 )
 
 func TestDistroImage(t *testing.T) {
-	_, actualDistro, _ := catalogFixtureImage(t, "image-distro-id")
+	sbom, _ := catalogFixtureImage(t, "image-distro-id")
 
 	expected, err := distro.NewDistro(distro.Busybox, "1.31.1", "")
 	if err != nil {
 		t.Fatalf("could not create distro: %+v", err)
 	}
 
-	for _, d := range deep.Equal(actualDistro, &expected) {
+	for _, d := range deep.Equal(sbom.Artifacts.Distro, &expected) {
 		t.Errorf("found distro difference: %+v", d)
 	}
 

test/integration/encode_decode_cycle_test.go

@@ -1,15 +1,14 @@
-package syft
+package integration
 
 import (
 	"bytes"
 	"testing"
 
-	"github.com/anchore/syft/syft/sbom"
+	"github.com/anchore/syft/syft"
 
-	"github.com/go-test/deep"
+	"github.com/sergi/go-diff/diffmatchpatch"
 
 	"github.com/anchore/syft/syft/format"
-	"github.com/anchore/syft/syft/source"
 	"github.com/stretchr/testify/assert"
 )
 
@@ -20,7 +19,6 @@ import (
 // to do an object-to-object comparison. For this reason this test focuses on a bytes-to-bytes comparison after an
 // encode-decode-encode loop which will detect lossy behavior in both directions.
 func TestEncodeDecodeEncodeCycleComparison(t *testing.T) {
-	testImage := "image-simple"
 	tests := []struct {
 		format format.Option
 	}{
@@ -29,35 +27,25 @@ func TestEncodeDecodeEncodeCycleComparison(t *testing.T) {
 		},
 	}
 	for _, test := range tests {
-		t.Run(testImage, func(t *testing.T) {
+		t.Run(string(test.format), func(t *testing.T) {
 
-			src, err := source.NewFromDirectory("./test-fixtures/pkgs")
+			originalSBOM, _ := catalogFixtureImage(t, "image-pkg-coverage")
-			if err != nil {
-				t.Fatalf("cant get dir")
-			}
-			originalCatalog, d, err := CatalogPackages(&src, source.SquashedScope)
 
-			originalSBOM := sbom.SBOM{
+			by1, err := syft.Encode(originalSBOM, test.format)
-				Artifacts: sbom.Artifacts{
-					PackageCatalog: originalCatalog,
-					Distro:         d,
-				},
-				Source: src.Metadata,
-			}
-
-			by1, err := Encode(originalSBOM, test.format)
 			assert.NoError(t, err)
 
-			newSBOM, newFormat, err := Decode(bytes.NewReader(by1))
+			newSBOM, newFormat, err := syft.Decode(bytes.NewReader(by1))
 			assert.NoError(t, err)
 			assert.Equal(t, test.format, newFormat)
 
-			by2, err := Encode(*newSBOM, test.format)
+			by2, err := syft.Encode(*newSBOM, test.format)
 			assert.NoError(t, err)
-			for _, diff := range deep.Equal(by1, by2) {
+
-				t.Errorf(diff)
+			if !assert.True(t, bytes.Equal(by1, by2)) {
+				dmp := diffmatchpatch.New()
+				diffs := dmp.DiffMain(string(by1), string(by2), true)
+				t.Errorf("diff: %s", dmp.DiffPrettyText(diffs))
 			}
-			assert.True(t, bytes.Equal(by1, by2))
 		})
 	}
 }

test/integration/node_packages_test.go

@@ -9,11 +9,11 @@ import (
 )
 
 func TestNpmPackageLockDirectory(t *testing.T) {
-	catalog, _, _ := catalogDirectory(t, "test-fixtures/npm-lock")
+	sbom, _ := catalogDirectory(t, "test-fixtures/npm-lock")
 
 	foundPackages := internal.NewStringSet()
 
-	for actualPkg := range catalog.Enumerate(pkg.NpmPkg) {
+	for actualPkg := range sbom.Artifacts.PackageCatalog.Enumerate(pkg.NpmPkg) {
 		for _, actualLocation := range actualPkg.Locations {
 			if strings.Contains(actualLocation.RealPath, "node_modules") {
 				t.Errorf("found packages from package-lock.json in node_modules: %s", actualLocation)
@@ -30,11 +30,11 @@ func TestNpmPackageLockDirectory(t *testing.T) {
 }
 
 func TestYarnPackageLockDirectory(t *testing.T) {
-	catalog, _, _ := catalogDirectory(t, "test-fixtures/yarn-lock")
+	sbom, _ := catalogDirectory(t, "test-fixtures/yarn-lock")
 
 	foundPackages := internal.NewStringSet()
 
-	for actualPkg := range catalog.Enumerate(pkg.NpmPkg) {
+	for actualPkg := range sbom.Artifacts.PackageCatalog.Enumerate(pkg.NpmPkg) {
 		for _, actualLocation := range actualPkg.Locations {
 			if strings.Contains(actualLocation.RealPath, "node_modules") {
 				t.Errorf("found packages from yarn.lock in node_modules: %s", actualLocation)

test/integration/package_ownership_relationship_test.go

@@ -7,7 +7,6 @@ import (
 
 	"github.com/anchore/syft/internal/formats/syftjson"
 	syftjsonModel "github.com/anchore/syft/internal/formats/syftjson/model"
-	"github.com/anchore/syft/syft/sbom"
 )
 
 func TestPackageOwnershipRelationships(t *testing.T) {
@@ -23,15 +22,9 @@ func TestPackageOwnershipRelationships(t *testing.T) {
 
 	for _, test := range tests {
 		t.Run(test.fixture, func(t *testing.T) {
-			catalog, d, src := catalogFixtureImage(t, test.fixture)
+			sbom, _ := catalogFixtureImage(t, test.fixture)
 
-			p := syftjson.Format().Presenter(sbom.SBOM{
+			p := syftjson.Format().Presenter(sbom)
-				Artifacts: sbom.Artifacts{
-					PackageCatalog: catalog,
-					Distro:         d,
-				},
-				Source: src.Metadata,
-			})
 			if p == nil {
 				t.Fatal("unable to get presenter")
 			}

test/integration/regression_apk_scanner_buffer_size_test.go

@@ -9,11 +9,11 @@ import (
 func TestRegression212ApkBufferSize(t *testing.T) {
 	// This is a regression test for issue #212 (https://github.com/anchore/syft/issues/212) in which the apk db could
 	// not be processed due to a scanner buffer that was too small
-	catalog, _, _ := catalogFixtureImage(t, "image-large-apk-data")
+	sbom, _ := catalogFixtureImage(t, "image-large-apk-data")
 
 	expectedPkgs := 58
 	actualPkgs := 0
-	for range catalog.Enumerate(pkg.ApkPkg) {
+	for range sbom.Artifacts.PackageCatalog.Enumerate(pkg.ApkPkg) {
 		actualPkgs += 1
 	}
 

test/integration/regression_go_bin_scanner_arch_test.go

@@ -15,11 +15,11 @@ func TestRegressionGoArchDiscovery(t *testing.T) {
 	)
 	// This is a regression test to make sure the way we detect go binary packages
 	// stays consistent and reproducible as the tool chain evolves
-	catalog, _, _ := catalogFixtureImage(t, "image-go-bin-arch-coverage")
+	sbom, _ := catalogFixtureImage(t, "image-go-bin-arch-coverage")
 
 	var actualELF, actualWIN, actualMACOS int
 
-	for p := range catalog.Enumerate(pkg.GoModulePkg) {
+	for p := range sbom.Artifacts.PackageCatalog.Enumerate(pkg.GoModulePkg) {
 		for _, l := range p.Locations {
 			switch {
 			case strings.Contains(l.RealPath, "elf"):

test/integration/utils_test.go

@@ -3,14 +3,14 @@ package integration
 import (
 	"testing"
 
+	"github.com/anchore/syft/syft/sbom"
+
 	"github.com/anchore/stereoscope/pkg/imagetest"
 	"github.com/anchore/syft/syft"
-	"github.com/anchore/syft/syft/distro"
-	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/source"
 )
 
-func catalogFixtureImage(t *testing.T, fixtureImageName string) (*pkg.Catalog, *distro.Distro, *source.Source) {
+func catalogFixtureImage(t *testing.T, fixtureImageName string) (sbom.SBOM, *source.Source) {
 	imagetest.GetFixtureImage(t, "docker-archive", fixtureImageName)
 	tarPath := imagetest.GetFixtureImageTarPath(t, fixtureImageName)
 
@@ -20,25 +20,39 @@ func catalogFixtureImage(t *testing.T, fixtureImageName string) (*pkg.Catalog, *
 		t.Fatalf("unable to get source: %+v", err)
 	}
 
-	pkgCatalog, actualDistro, err := syft.CatalogPackages(theSource, source.SquashedScope)
+	pkgCatalog, relationships, actualDistro, err := syft.CatalogPackages(theSource, source.SquashedScope)
 	if err != nil {
 		t.Fatalf("failed to catalog image: %+v", err)
 	}
 
-	return pkgCatalog, actualDistro, theSource
+	return sbom.SBOM{
+		Artifacts: sbom.Artifacts{
+			PackageCatalog: pkgCatalog,
+			Distro:         actualDistro,
+		},
+		Relationships: relationships,
+		Source:        theSource.Metadata,
+	}, theSource
 }
 
-func catalogDirectory(t *testing.T, dir string) (*pkg.Catalog, *distro.Distro, *source.Source) {
+func catalogDirectory(t *testing.T, dir string) (sbom.SBOM, *source.Source) {
 	theSource, cleanupSource, err := source.New("dir:"+dir, nil)
 	t.Cleanup(cleanupSource)
 	if err != nil {
 		t.Fatalf("unable to get source: %+v", err)
 	}
 
-	pkgCatalog, actualDistro, err := syft.CatalogPackages(theSource, source.AllLayersScope)
+	pkgCatalog, relationships, actualDistro, err := syft.CatalogPackages(theSource, source.AllLayersScope)
 	if err != nil {
 		t.Fatalf("failed to catalog image: %+v", err)
 	}
 
-	return pkgCatalog, actualDistro, theSource
+	return sbom.SBOM{
+		Artifacts: sbom.Artifacts{
+			PackageCatalog: pkgCatalog,
+			Distro:         actualDistro,
+		},
+		Relationships: relationships,
+		Source:        theSource.Metadata,
+	}, theSource
 }