1948-filter-pkg-by-type (#2011)

---------

Signed-off-by: Christopher Phillips <christopher.phillips@anchore.com>

README.md

@@ -484,7 +484,7 @@ exclude: []
 
 # allows users to exclude synthetic binary packages from the sbom
 # these packages are removed if an overlap with a non-synthetic package is found
-exclude-overlap-by-ownership: true
+exclude-binary-overlap-by-ownership: true
 
 # os and/or architecture to use when referencing container images (e.g. "windows/armv6" or "arm64")
 # same as --platform; SYFT_PLATFORM env var

syft/lib.go

@@ -79,7 +79,7 @@ func CatalogPackages(src source.Source, cfg cataloger.Config) (*pkg.Collection,
 	// https://github.com/anchore/syft/issues/931
 	if cfg.ExcludeBinaryOverlapByOwnership {
 		for _, r := range relationships {
-			if cataloger.Exclude(r, catalog) {
+			if cataloger.ExcludeBinaryByFileOwnershipOverlap(r, catalog) {
 				catalog.Delete(r.To.ID())
 				relationships = removeRelationshipsByID(relationships, r.To.ID())
 			}

syft/pkg/cataloger/alpm/cataloger.go

@@ -5,9 +5,9 @@ import (
 	"github.com/anchore/syft/syft/pkg/cataloger/generic"
 )
 
-const CatalogerName = "alpmdb-cataloger"
+const catalogerName = "alpmdb-cataloger"
 
 func NewAlpmdbCataloger() *generic.Cataloger {
-	return generic.NewCataloger(CatalogerName).
+	return generic.NewCataloger(catalogerName).
 		WithParserByGlobs(parseAlpmDB, pkg.AlpmDBGlob)
 }

syft/pkg/cataloger/apkdb/cataloger.go

@@ -8,10 +8,10 @@ import (
 	"github.com/anchore/syft/syft/pkg/cataloger/generic"
 )
 
-const CatalogerName = "apkdb-cataloger"
+const catalogerName = "apkdb-cataloger"
 
 // NewApkdbCataloger returns a new Alpine DB cataloger object.
 func NewApkdbCataloger() *generic.Cataloger {
-	return generic.NewCataloger(CatalogerName).
+	return generic.NewCataloger(catalogerName).
 		WithParserByGlobs(parseApkDB, pkg.ApkDBGlob)
 }

syft/pkg/cataloger/binary/cataloger.go

@@ -7,7 +7,7 @@ import (
 	"github.com/anchore/syft/syft/pkg"
 )
 
-const CatalogerName = "binary-cataloger"
+const catalogerName = "binary-cataloger"
 
 func NewCataloger() *Cataloger {
 	return &Cataloger{}
@@ -22,7 +22,7 @@ type Cataloger struct{}
 
 // Name returns a string that uniquely describes the Cataloger
 func (c Cataloger) Name() string {
-	return CatalogerName
+	return catalogerName
 }
 
 // Catalog is given an object to resolve file references and content, this function returns any discovered Packages

syft/pkg/cataloger/binary/package.go

@@ -31,7 +31,7 @@ func newPackage(classifier classifier, location file.Location, matchMetadata map
 		),
 		Type:         pkg.BinaryPkg,
 		CPEs:         cpes,
-		FoundBy:      CatalogerName,
+		FoundBy:      catalogerName,
 		MetadataType: pkg.BinaryMetadataType,
 		Metadata: pkg.BinaryMetadata{
 			Matches: []pkg.ClassifierMatch{

syft/pkg/cataloger/deb/cataloger.go

@@ -7,11 +7,11 @@ import (
 	"github.com/anchore/syft/syft/pkg/cataloger/generic"
 )
 
-const CatalogerName = "dpkgdb-cataloger"
+const catalogerName = "dpkgdb-cataloger"
 
 // NewDpkgdbCataloger returns a new Deb package cataloger capable of parsing DPKG status DB files.
 func NewDpkgdbCataloger() *generic.Cataloger {
-	return generic.NewCataloger(CatalogerName).
+	return generic.NewCataloger(catalogerName).
 		// note: these globs have been intentionally split up in order to improve search performance,
 		// please do NOT combine into: "**/var/lib/dpkg/{status,status.d/*}"
 		WithParserByGlobs(parseDpkgDB, "**/var/lib/dpkg/status", "**/var/lib/dpkg/status.d/*", "**/lib/opkg/info/*.control", "**/lib/opkg/status")

syft/pkg/cataloger/nix/cataloger.go

@@ -12,7 +12,7 @@ import (
 )
 
 const (
-	CatalogerName = "nix-store-cataloger"
+	catalogerName = "nix-store-cataloger"
 	nixStoreGlob  = "**/nix/store/*"
 )
 
@@ -24,7 +24,7 @@ func NewStoreCataloger() *StoreCataloger {
 }
 
 func (c *StoreCataloger) Name() string {
-	return CatalogerName
+	return catalogerName
 }
 
 func (c *StoreCataloger) Catalog(resolver file.Resolver) ([]pkg.Package, []artifact.Relationship, error) {

syft/pkg/cataloger/nix/cataloger_test.go

@@ -24,7 +24,7 @@ func TestCataloger_Catalog(t *testing.T) {
 					Version:      "2.34-210",
 					PURL:         "pkg:nix/glibc@2.34-210?output=bin&outputhash=h0cnbmfcn93xm5dg2x27ixhag1cwndga",
 					Locations:    file.NewLocationSet(file.NewLocation("nix/store/h0cnbmfcn93xm5dg2x27ixhag1cwndga-glibc-2.34-210-bin")),
-					FoundBy:      CatalogerName,
+					FoundBy:      catalogerName,
 					Type:         pkg.NixPkg,
 					MetadataType: pkg.NixStoreMetadataType,
 					Metadata: pkg.NixStoreMetadata{

syft/pkg/cataloger/nix/package.go

@@ -10,7 +10,7 @@ func newNixStorePackage(storePath nixStorePath, locations ...file.Location) pkg.
 	p := pkg.Package{
 		Name:         storePath.name,
 		Version:      storePath.version,
-		FoundBy:      CatalogerName,
+		FoundBy:      catalogerName,
 		Locations:    file.NewLocationSet(locations...),
 		Type:         pkg.NixPkg,
 		PURL:         packageURL(storePath),

syft/pkg/cataloger/package_exclusions.go

@@ -5,33 +5,29 @@ import (
 
 	"github.com/anchore/syft/syft/artifact"
 	"github.com/anchore/syft/syft/pkg"
-	"github.com/anchore/syft/syft/pkg/cataloger/alpm"
-	"github.com/anchore/syft/syft/pkg/cataloger/apkdb"
-	"github.com/anchore/syft/syft/pkg/cataloger/binary"
-	"github.com/anchore/syft/syft/pkg/cataloger/deb"
-	"github.com/anchore/syft/syft/pkg/cataloger/nix"
-	"github.com/anchore/syft/syft/pkg/cataloger/rpm"
 )
 
 var (
-	osCatalogerTypes = []string{
+	osCatalogerTypes = []pkg.Type{
-		apkdb.CatalogerName,
+		pkg.AlpmPkg,
-		alpm.CatalogerName,
+		pkg.ApkPkg,
-		deb.CatalogerName,
+		pkg.DebPkg,
-		nix.CatalogerName,
+		pkg.NixPkg,
-		rpm.DBCatalogerName,
+		pkg.PortagePkg,
-		rpm.FileCatalogerName,
+		pkg.RpmPkg,
+	}
+	binaryCatalogerTypes = []pkg.Type{
+		pkg.BinaryPkg,
 	}
-	binaryCatalogerTypes = []string{binary.CatalogerName}
 )
 
-// Exclude will remove packages from a collection given the following properties are true
+// ExcludeBinaryByFileOwnershipOverlap will remove packages from a collection given the following properties are true
 // 1) the relationship between packages is OwnershipByFileOverlap
 // 2) the parent is an "os" package
 // 3) the child is a synthetic package generated by the binary cataloger
 // 4) the package names are identical
-// This exclude was implemented as a way to help resolve: https://github.com/anchore/syft/issues/931
+// This was implemented as a way to help resolve: https://github.com/anchore/syft/issues/931
-func Exclude(r artifact.Relationship, c *pkg.Collection) bool {
+func ExcludeBinaryByFileOwnershipOverlap(r artifact.Relationship, c *pkg.Collection) bool {
 	if artifact.OwnershipByFileOverlapRelationship != r.Type {
 		return false
 	}
@@ -41,7 +37,7 @@ func Exclude(r artifact.Relationship, c *pkg.Collection) bool {
 		return false
 	}
 
-	parentInExclusion := slices.Contains(osCatalogerTypes, parent.FoundBy)
+	parentInExclusion := slices.Contains(osCatalogerTypes, parent.Type)
 	if !parentInExclusion {
 		return false
 	}
@@ -51,5 +47,5 @@ func Exclude(r artifact.Relationship, c *pkg.Collection) bool {
 		return false
 	}
 
-	return slices.Contains(binaryCatalogerTypes, child.FoundBy)
+	return slices.Contains(binaryCatalogerTypes, child.Type)
 }

syft/pkg/cataloger/package_exclusions_test.go

@@ -5,15 +5,13 @@ 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/binary"
 )
 
 func TestExclude(t *testing.T) {
-	packageA := pkg.Package{Name: "package-a", Type: pkg.ApkPkg, FoundBy: apkdb.CatalogerName}
+	packageA := pkg.Package{Name: "package-a", Type: pkg.ApkPkg}
-	packageB := pkg.Package{Name: "package-a", Type: pkg.PythonPkg, FoundBy: "language-cataloger"}
+	packageB := pkg.Package{Name: "package-a", Type: pkg.PythonPkg}
-	packageC := pkg.Package{Name: "package-a", Type: pkg.BinaryPkg, FoundBy: binary.CatalogerName}
+	packageC := pkg.Package{Name: "package-a", Type: pkg.BinaryPkg}
-	packageD := pkg.Package{Name: "package-d", Type: pkg.BinaryPkg, FoundBy: binary.CatalogerName}
+	packageD := pkg.Package{Name: "package-d", Type: pkg.BinaryPkg}
 	for _, p := range []*pkg.Package{&packageA, &packageB, &packageC, &packageD} {
 		p := p
 		p.SetID()
@@ -69,7 +67,7 @@ func TestExclude(t *testing.T) {
 
 	for _, test := range tests {
 		t.Run(test.name, func(t *testing.T) {
-			if !Exclude(test.relationship, test.packages) && test.shouldExclude {
+			if !ExcludeBinaryByFileOwnershipOverlap(test.relationship, test.packages) && test.shouldExclude {
 				t.Errorf("expected to exclude relationship %+v", test.relationship)
 			}
 		})

syft/pkg/cataloger/rpm/cataloger.go

@@ -12,8 +12,8 @@ import (
 )
 
 const (
-	DBCatalogerName   = "rpm-db-cataloger"
+	dbCatalogerName   = "rpm-db-cataloger"
-	FileCatalogerName = "rpm-file-cataloger"
+	fileCatalogerName = "rpm-file-cataloger"
 )
 
 // NewRpmDBCataloger returns a new RPM DB cataloger object.
@@ -23,14 +23,14 @@ func NewRpmDBCataloger() *generic.Cataloger {
 		log.Warnf("sqlite driver is not available, newer RPM databases might not be cataloged")
 	}
 
-	return generic.NewCataloger(DBCatalogerName).
+	return generic.NewCataloger(dbCatalogerName).
 		WithParserByGlobs(parseRpmDB, pkg.RpmDBGlob).
 		WithParserByGlobs(parseRpmManifest, pkg.RpmManifestGlob)
 }
 
 // NewFileCataloger returns a new RPM file cataloger object.
 func NewFileCataloger() *generic.Cataloger {
-	return generic.NewCataloger(FileCatalogerName).
+	return generic.NewCataloger(fileCatalogerName).
 		WithParserByGlobs(parseRpm, "**/*.rpm")
 }