feat: 3626 add option enable license content; disable by default (#3631)

---------
Signed-off-by: Christopher Phillips <32073428+spiffcs@users.noreply.github.com>

cmd/syft/internal/options/catalog.go

@@ -33,6 +33,7 @@ type Catalog struct {
 	DefaultCatalogers []string            `yaml:"default-catalogers" json:"default-catalogers" mapstructure:"default-catalogers"`
 	SelectCatalogers  []string            `yaml:"select-catalogers" json:"select-catalogers" mapstructure:"select-catalogers"`
 	Package           packageConfig       `yaml:"package" json:"package" mapstructure:"package"`
+	License           licenseConfig       `yaml:"license" json:"license" mapstructure:"license"`
 	File              fileConfig          `yaml:"file" json:"file" mapstructure:"file"`
 	Scope             string              `yaml:"scope" json:"scope" mapstructure:"scope"`
 	Parallelism       int                 `yaml:"parallelism" json:"parallelism" mapstructure:"parallelism"` // the number of catalog workers to run in parallel
@@ -69,6 +70,7 @@ func DefaultCatalog() Catalog {
 		Compliance:    defaultComplianceConfig(),
 		Scope:         source.SquashedScope.String(),
 		Package:       defaultPackageConfig(),
+		License:       defaultLicenseConfig(),
 		LinuxKernel:   defaultLinuxKernelConfig(),
 		Golang:        defaultGolangConfig(),
 		Java:          defaultJavaConfig(),
@@ -89,6 +91,7 @@ func (cfg Catalog) ToSBOMConfig(id clio.Identification) *syft.CreateSBOMConfig {
 		WithUnknownsConfig(cfg.ToUnknownsConfig()).
 		WithSearchConfig(cfg.ToSearchConfig()).
 		WithPackagesConfig(cfg.ToPackagesConfig()).
+		WithLicenseConfig(cfg.ToLicenseConfig()).
 		WithFilesConfig(cfg.ToFilesConfig()).
 		WithCatalogerSelection(
 			cataloging.NewSelectionRequest().
@@ -146,6 +149,13 @@ func (cfg Catalog) ToFilesConfig() filecataloging.Config {
 	}
 }
 
+func (cfg Catalog) ToLicenseConfig() cataloging.LicenseConfig {
+	return cataloging.LicenseConfig{
+		IncludeUnkownLicenseContent: cfg.License.IncludeUnknownLicenseContent,
+		Coverage:                    cfg.License.LicenseCoverage,
+	}
+}
+
 func (cfg Catalog) ToPackagesConfig() pkgcataloging.Config {
 	archiveSearch := cataloging.ArchiveSearchConfig{
 		IncludeIndexedArchives:   cfg.Package.SearchIndexedArchives,

cmd/syft/internal/options/license.go

@@ -0,0 +1,28 @@
+package options
+
+import (
+	"github.com/anchore/clio"
+)
+
+type licenseConfig struct {
+	IncludeUnknownLicenseContent bool    `yaml:"include-unknown-license-content" json:"include-unknown-license-content" mapstructure:"include-unknown-license-content"`
+	LicenseCoverage              float64 `yaml:"license-coverage" json:"license-coverage" mapstructure:"license-coverage"`
+}
+
+var _ interface {
+	clio.FieldDescriber
+} = (*licenseConfig)(nil)
+
+func (o *licenseConfig) DescribeFields(descriptions clio.FieldDescriptionSet) {
+	descriptions.Add(&o.IncludeUnknownLicenseContent, `include the content of a license in the SBOM when syft
+cannot determine a valid SPDX ID for the given license`)
+	descriptions.Add(&o.LicenseCoverage, `adjust the percent as a fraction of the total text, in normalized words, that
+matches any valid license for the given inputs, expressed as a percentage across all of the licenses matched.`)
+}
+
+func defaultLicenseConfig() licenseConfig {
+	return licenseConfig{
+		IncludeUnknownLicenseContent: false,
+		LicenseCoverage:              75,
+	}
+}

internal/licenses/context.go

@@ -10,9 +10,9 @@ func SetContextLicenseScanner(ctx context.Context, s Scanner) context.Context {
 	return context.WithValue(ctx, licenseScannerKey{}, s)
 }
 
-func ContextLicenseScanner(ctx context.Context) Scanner {
+func ContextLicenseScanner(ctx context.Context) (Scanner, error) {
 	if s, ok := ctx.Value(licenseScannerKey{}).(Scanner); ok {
-		return s
+		return s, nil
 	}
 	return NewDefaultScanner()
 }

internal/licenses/scanner.go

@@ -2,66 +2,84 @@ package licenses
 
 import (
 	"context"
+	"fmt"
 	"io"
 
 	"github.com/google/licensecheck"
 
 	"github.com/anchore/syft/internal/log"
+	"github.com/anchore/syft/syft/file"
+	"github.com/anchore/syft/syft/pkg"
 )
 
-const coverageThreshold = 75 // determined by experimentation
+const (
+	DefaultCoverageThreshold     = 75 // determined by experimentation
+	DefaultIncludeLicenseContent = false
+)
 
 type Scanner interface {
 	IdentifyLicenseIDs(context.Context, io.Reader) ([]string, []byte, error)
+	FileSearch(context.Context, file.LocationReadCloser) ([]file.License, error)
+	PkgSearch(context.Context, file.LocationReadCloser) ([]pkg.License, error)
 }
 
 var _ Scanner = (*scanner)(nil)
 
 type scanner struct {
 	coverageThreshold     float64 // between 0 and 100
+	includeLicenseContent bool
 	scanner               func([]byte) licensecheck.Coverage
 }
 
+type ScannerConfig struct {
+	CoverageThreshold     float64
+	IncludeLicenseContent bool
+	Scanner               func([]byte) licensecheck.Coverage
+}
+
+type Option func(*scanner)
+
+func WithCoverage(coverage float64) Option {
+	return func(s *scanner) {
+		s.coverageThreshold = coverage
+	}
+}
+
+func WithIncludeLicenseContent(includeLicenseContent bool) Option {
+	return func(s *scanner) {
+		s.includeLicenseContent = includeLicenseContent
+	}
+}
+
 // NewDefaultScanner returns a scanner that uses a new instance of the default licensecheck package scanner.
-func NewDefaultScanner() Scanner {
+func NewDefaultScanner(o ...Option) (Scanner, error) {
 	s, err := licensecheck.NewScanner(licensecheck.BuiltinLicenses())
 	if err != nil {
 		log.WithFields("error", err).Trace("unable to create default license scanner")
-		s = nil
+		return nil, fmt.Errorf("unable to create default license scanner: %w", err)
 	}
-	return &scanner{
+	newScanner := &scanner{
-		coverageThreshold: coverageThreshold,
+		coverageThreshold:     DefaultCoverageThreshold,
+		includeLicenseContent: DefaultIncludeLicenseContent,
 		scanner:               s.Scan,
 	}
+
+	for _, opt := range o {
+		opt(newScanner)
+	}
+	return newScanner, nil
 }
 
-func NewScanner(scan func([]byte) licensecheck.Coverage, coverage float64) Scanner {
+// NewScanner generates a license Scanner with the given ScannerConfig
-	return scanner{
+// if config is nil NewDefaultScanner is used
-		coverageThreshold: coverage,
+func NewScanner(c *ScannerConfig) (Scanner, error) {
-		scanner:           scan,
+	if c == nil {
+		return NewDefaultScanner()
 	}
-}
+
-
+	return &scanner{
-func (s scanner) IdentifyLicenseIDs(_ context.Context, reader io.Reader) ([]string, []byte, error) {
+		coverageThreshold:     c.CoverageThreshold,
-	if s.scanner == nil {
+		includeLicenseContent: c.IncludeLicenseContent,
-		return nil, nil, nil
+		scanner:               c.Scanner,
-	}
+	}, nil
-
-	content, err := io.ReadAll(reader)
-	if err != nil {
-		return nil, nil, err
-	}
-
-	cov := s.scanner(content)
-	if cov.Percent < s.coverageThreshold {
-		// unknown or no licenses here?
-		// => return binary content
-		return nil, content, nil
-	}
-
-	var ids []string
-	for _, m := range cov.Match {
-		ids = append(ids, m.ID)
-	}
-	return ids, nil, nil
 }

internal/licenses/scanner_test.go

@@ -27,11 +27,11 @@ func TestIdentifyLicenseIDs(t *testing.T) {
 			expected: expectation{
 				yieldError: false,
 				ids:        []string{"Apache-2.0"},
-				content:    []byte{},
+				content:    nil,
 			},
 		},
 		{
-			name: "custom license",
+			name: "custom license includes content for IdentifyLicenseIDs",
 			in:   "test-fixtures/nvidia-software-and-cuda-supplement",
 			expected: expectation{
 				yieldError: false,
@@ -45,7 +45,7 @@ func TestIdentifyLicenseIDs(t *testing.T) {
 		t.Run(test.name, func(t *testing.T) {
 			content, err := os.ReadFile(test.in)
 			require.NoError(t, err)
-			ids, content, err := testScanner().IdentifyLicenseIDs(context.TODO(), bytes.NewReader(content))
+			ids, content, err := testScanner(false).IdentifyLicenseIDs(context.TODO(), bytes.NewReader(content))
 			if test.expected.yieldError {
 				require.Error(t, err)
 			} else {
@@ -66,9 +66,10 @@ func TestIdentifyLicenseIDs(t *testing.T) {
 	}
 }
 
-func testScanner() Scanner {
+func testScanner(includeLicenseContent bool) Scanner {
 	return &scanner{
-		coverageThreshold: coverageThreshold,
+		coverageThreshold:     DefaultCoverageThreshold,
+		includeLicenseContent: includeLicenseContent,
 		scanner:               licensecheck.Scan,
 	}
 }

internal/licenses/search.go

@@ -4,6 +4,7 @@ import (
 	"context"
 	"crypto/sha256"
 	"fmt"
+	"io"
 	"strings"
 
 	"github.com/anchore/syft/syft/file"
@@ -21,11 +22,35 @@ func getCustomLicenseContentHash(contents []byte) string {
 	return fmt.Sprintf("%x", hash[:])
 }
 
-// Search scans the contents of a license file to attempt to determine the type of license it is
+func (s *scanner) IdentifyLicenseIDs(_ context.Context, reader io.Reader) ([]string, []byte, error) {
-func Search(ctx context.Context, scanner Scanner, reader file.LocationReadCloser) (licenses []pkg.License, err error) {
+	if s.scanner == nil {
+		return nil, nil, nil
+	}
+
+	content, err := io.ReadAll(reader)
+	if err != nil {
+		return nil, nil, err
+	}
+
+	cov := s.scanner(content)
+	if cov.Percent < s.coverageThreshold {
+		// unknown or no licenses here
+		// => check return content to Search to process
+		return nil, content, nil
+	}
+
+	var ids []string
+	for _, m := range cov.Match {
+		ids = append(ids, m.ID)
+	}
+	return ids, nil, nil
+}
+
+// PkgSearch scans the contents of a license file to attempt to determine the type of license it is
+func (s *scanner) PkgSearch(ctx context.Context, reader file.LocationReadCloser) (licenses []pkg.License, err error) {
 	licenses = make([]pkg.License, 0)
 
-	ids, content, err := scanner.IdentifyLicenseIDs(ctx, reader)
+	ids, content, err := s.IdentifyLicenseIDs(ctx, reader)
 	if err != nil {
 		return nil, err
 	}
@@ -48,7 +73,47 @@ func Search(ctx context.Context, scanner Scanner, reader file.LocationReadCloser
 
 		lic := pkg.NewLicenseFromLocations(unknownLicenseType, reader.Location)
 		lic.SPDXExpression = UnknownLicensePrefix + getCustomLicenseContentHash(content)
+		if s.includeLicenseContent {
 			lic.Contents = string(content)
+		}
+		lic.Type = license.Declared
+
+		licenses = append(licenses, lic)
+	}
+
+	return licenses, nil
+}
+
+// FileSearch scans the contents of a license file to attempt to determine the type of license it is
+func (s *scanner) FileSearch(ctx context.Context, reader file.LocationReadCloser) (licenses []file.License, err error) {
+	licenses = make([]file.License, 0)
+
+	ids, content, err := s.IdentifyLicenseIDs(ctx, reader)
+	if err != nil {
+		return nil, err
+	}
+
+	// IdentifyLicenseIDs can only return a list of ID or content
+	// These return values are mutually exclusive.
+	// If the scanner threshold for matching scores < 75% then we return the license full content
+	if len(ids) > 0 {
+		for _, id := range ids {
+			lic := file.NewLicense(id)
+			lic.Type = license.Concluded
+
+			licenses = append(licenses, lic)
+		}
+	} else if len(content) > 0 {
+		// harmonize line endings to unix compatible first:
+		// 1. \r\n => \n   (Windows   => UNIX)
+		// 2. \r   => \n   (Macintosh => UNIX)
+		content = []byte(strings.ReplaceAll(strings.ReplaceAll(string(content), "\r\n", "\n"), "\r", "\n"))
+
+		lic := file.NewLicense(unknownLicenseType)
+		lic.SPDXExpression = UnknownLicensePrefix + getCustomLicenseContentHash(content)
+		if s.includeLicenseContent {
+			lic.Contents = string(content)
+		}
 		lic.Type = license.Declared
 
 		licenses = append(licenses, lic)

internal/licenses/search_test.go

@@ -27,15 +27,16 @@ func newBytesReadCloser(data []byte) *bytesReadCloser {
 	}
 }
 
-func TestSearch(t *testing.T) {
+func TestSearchFileLicenses(t *testing.T) {
 	type expectation struct {
 		yieldError bool
-		licenses   []pkg.License
+		licenses   []file.License
 	}
-	testLocation := file.NewLocation("LICENSE")
+
 	tests := []struct {
 		name                        string
 		in                          string
+		includeUnkownLicenseContent bool
 		expected                    expectation
 	}{
 		{
@@ -43,31 +44,11 @@ func TestSearch(t *testing.T) {
 			in:   "test-fixtures/apache-license-2.0",
 			expected: expectation{
 				yieldError: false,
-				licenses: []pkg.License{
+				licenses: []file.License{
 					{
 						Value:          "Apache-2.0",
 						SPDXExpression: "Apache-2.0",
 						Type:           "concluded",
-						URLs:           nil,
-						Locations:      file.NewLocationSet(testLocation),
-						Contents:       "",
-					},
-				},
-			},
-		},
-		{
-			name: "custom license",
-			in:   "test-fixtures/nvidia-software-and-cuda-supplement",
-			expected: expectation{
-				yieldError: false,
-				licenses: []pkg.License{
-					{
-						Value:          "UNKNOWN",
-						SPDXExpression: "UNKNOWN_eebcea3ab1d1a28e671de90119ffcfb35fe86951e4af1b17af52b7a82fcf7d0a",
-						Type:           "declared",
-						URLs:           nil,
-						Locations:      file.NewLocationSet(testLocation),
-						Contents:       string(mustOpen("test-fixtures/nvidia-software-and-cuda-supplement")),
 					},
 				},
 			},
@@ -76,9 +57,11 @@ func TestSearch(t *testing.T) {
 
 	for _, test := range tests {
 		t.Run(test.name, func(t *testing.T) {
+			ctx := context.TODO()
 			content, err := os.ReadFile(test.in)
 			require.NoError(t, err)
-			result, err := Search(context.TODO(), testScanner(), file.NewLocationReadCloser(file.NewLocation("LICENSE"), io.NopCloser(bytes.NewReader(content))))
+			s := testScanner(false)
+			result, err := s.FileSearch(ctx, file.NewLocationReadCloser(file.NewLocation("LICENSE"), io.NopCloser(bytes.NewReader(content))))
 			if test.expected.yieldError {
 				require.Error(t, err)
 			} else {
@@ -93,3 +76,91 @@ func TestSearch(t *testing.T) {
 		})
 	}
 }
+
+func TestSearchPkgLicenses(t *testing.T) {
+	type expectation struct {
+		wantErr  require.ErrorAssertionFunc
+		licenses []pkg.License
+	}
+
+	testLocation := file.NewLocation("LICENSE")
+	tests := []struct {
+		name                        string
+		in                          string
+		includeUnkownLicenseContent bool
+		expected                    expectation
+	}{
+		{
+			name: "apache license 2.0",
+			in:   "test-fixtures/apache-license-2.0",
+			expected: expectation{
+				licenses: []pkg.License{
+					{
+						Value:          "Apache-2.0",
+						SPDXExpression: "Apache-2.0",
+						Type:           "concluded",
+						URLs:           nil,
+						Locations:      file.NewLocationSet(testLocation),
+						Contents:       "",
+					},
+				},
+				wantErr: nil,
+			},
+		},
+		{
+			name: "custom license no content by default",
+			in:   "test-fixtures/nvidia-software-and-cuda-supplement",
+			expected: expectation{
+				licenses: []pkg.License{
+					{
+						Value:          "UNKNOWN",
+						SPDXExpression: "UNKNOWN_eebcea3ab1d1a28e671de90119ffcfb35fe86951e4af1b17af52b7a82fcf7d0a",
+						Type:           "declared",
+						URLs:           nil,
+						Locations:      file.NewLocationSet(testLocation),
+						Contents:       "",
+					},
+				},
+				wantErr: nil,
+			},
+		},
+		{
+			name:                        "custom license with content when scanner has content config",
+			in:                          "test-fixtures/nvidia-software-and-cuda-supplement",
+			includeUnkownLicenseContent: true,
+			expected: expectation{
+				licenses: []pkg.License{
+					{
+						Value:          "UNKNOWN",
+						SPDXExpression: "UNKNOWN_eebcea3ab1d1a28e671de90119ffcfb35fe86951e4af1b17af52b7a82fcf7d0a",
+						Type:           "declared",
+						URLs:           nil,
+						Locations:      file.NewLocationSet(testLocation),
+						Contents:       string(mustOpen("test-fixtures/nvidia-software-and-cuda-supplement")),
+					},
+				},
+				wantErr: nil,
+			},
+		},
+	}
+
+	for _, test := range tests {
+		t.Run(test.name, func(t *testing.T) {
+			ctx := context.TODO()
+			content, err := os.ReadFile(test.in)
+			require.NoError(t, err)
+			s := testScanner(test.includeUnkownLicenseContent)
+			result, err := s.PkgSearch(ctx, file.NewLocationReadCloser(file.NewLocation("LICENSE"), io.NopCloser(bytes.NewReader(content))))
+			if test.expected.wantErr != nil {
+				test.expected.wantErr(t, err)
+			}
+			require.NoError(t, err)
+
+			require.Len(t, result, len(test.expected.licenses))
+
+			if len(test.expected.licenses) > 0 {
+				require.Equal(t, test.expected.licenses, result)
+			}
+		})
+	}
+}

internal/task/cataloging_config.go

@@ -11,6 +11,7 @@ type CatalogingFactoryConfig struct {
 	SearchConfig         cataloging.SearchConfig
 	RelationshipsConfig  cataloging.RelationshipsConfig
 	DataGenerationConfig cataloging.DataGenerationConfig
+	LicenseConfig        cataloging.LicenseConfig
 	PackagesConfig       pkgcataloging.Config
 	FilesConfig          filecataloging.Config
 }
@@ -21,6 +22,7 @@ func DefaultCatalogingFactoryConfig() CatalogingFactoryConfig {
 		SearchConfig:         cataloging.DefaultSearchConfig(),
 		RelationshipsConfig:  cataloging.DefaultRelationshipsConfig(),
 		DataGenerationConfig: cataloging.DefaultDataGenerationConfig(),
+		LicenseConfig:        cataloging.DefaultLicenseConfig(),
 		PackagesConfig:       pkgcataloging.DefaultConfig(),
 		FilesConfig:          filecataloging.DefaultConfig(),
 	}

syft/cataloging/license.go

@@ -0,0 +1,15 @@
+package cataloging
+
+import "github.com/anchore/syft/internal/licenses"
+
+type LicenseConfig struct {
+	IncludeUnkownLicenseContent bool    `json:"include-unknown-license-content" yaml:"include-unknown-license-content" mapstructure:"include-unknown-license-content"`
+	Coverage                    float64 `json:"coverage" yaml:"coverage" mapstructure:"coverage"`
+}
+
+func DefaultLicenseConfig() LicenseConfig {
+	return LicenseConfig{
+		IncludeUnkownLicenseContent: licenses.DefaultIncludeLicenseContent,
+		Coverage:                    licenses.DefaultCoverageThreshold,
+	}
+}

syft/configuration_audit_trail.go

@@ -16,6 +16,7 @@ type configurationAuditTrail struct {
 	DataGeneration cataloging.DataGenerationConfig `json:"data-generation" yaml:"data-generation" mapstructure:"data-generation"`
 	Packages       pkgcataloging.Config            `json:"packages" yaml:"packages" mapstructure:"packages"`
 	Files          filecataloging.Config           `json:"files" yaml:"files" mapstructure:"files"`
+	Licenses       cataloging.LicenseConfig        `json:"licenses" yaml:"licenses" mapstructure:"licenses"`
 	Catalogers     catalogerManifest               `json:"catalogers" yaml:"catalogers" mapstructure:"catalogers"`
 	ExtraConfigs   any                             `json:"extra,omitempty" yaml:"extra" mapstructure:"extra"`
 }

syft/create_sbom.go

@@ -52,6 +52,7 @@ func CreateSBOM(ctx context.Context, src source.Source, cfg *CreateSBOMConfig) (
 				DataGeneration: cfg.DataGeneration,
 				Packages:       cfg.Packages,
 				Files:          cfg.Files,
+				Licenses:       cfg.Licenses,
 				Catalogers:     *audit,
 				ExtraConfigs:   cfg.ToolConfiguration,
 			},
@@ -61,8 +62,15 @@ func CreateSBOM(ctx context.Context, src source.Source, cfg *CreateSBOMConfig) (
 		},
 	}
 
-	// inject a single license scanner for all package cataloging tasks into context
+	// inject a single license scanner and content config for all package cataloging tasks into context
-	ctx = licenses.SetContextLicenseScanner(ctx, licenses.NewDefaultScanner())
+	licenseScanner, err := licenses.NewDefaultScanner(
+		licenses.WithIncludeLicenseContent(cfg.Licenses.IncludeUnkownLicenseContent),
+		licenses.WithCoverage(cfg.Licenses.Coverage),
+	)
+	if err != nil {
+		return nil, fmt.Errorf("could not build licenseScanner for cataloging: %w", err)
+	}
+	ctx = licenses.SetContextLicenseScanner(ctx, licenseScanner)
 
 	catalogingProgress := monitorCatalogingTask(src.ID(), taskGroups)
 	packageCatalogingProgress := monitorPackageCatalogingTask()

syft/create_sbom_config.go

@@ -26,6 +26,7 @@ type CreateSBOMConfig struct {
 	Unknowns           cataloging.UnknownsConfig
 	DataGeneration     cataloging.DataGenerationConfig
 	Packages           pkgcataloging.Config
+	Licenses           cataloging.LicenseConfig
 	Files              filecataloging.Config
 	Parallelism        int
 	CatalogerSelection cataloging.SelectionRequest
@@ -46,6 +47,7 @@ func DefaultCreateSBOMConfig() *CreateSBOMConfig {
 		Relationships:        cataloging.DefaultRelationshipsConfig(),
 		DataGeneration:       cataloging.DefaultDataGenerationConfig(),
 		Packages:             pkgcataloging.DefaultConfig(),
+		Licenses:             cataloging.DefaultLicenseConfig(),
 		Files:                filecataloging.DefaultConfig(),
 		Parallelism:          1,
 		packageTaskFactories: task.DefaultPackageTaskFactories(),
@@ -134,6 +136,12 @@ func (c *CreateSBOMConfig) WithPackagesConfig(cfg pkgcataloging.Config) *CreateS
 	return c
 }
 
+// WithPackagesConfig allows for defining any specific behavior for syft-implemented catalogers.
+func (c *CreateSBOMConfig) WithLicenseConfig(cfg cataloging.LicenseConfig) *CreateSBOMConfig {
+	c.Licenses = cfg
+	return c
+}
+
 // WithFilesConfig allows for defining file-based cataloging parameters.
 func (c *CreateSBOMConfig) WithFilesConfig(cfg filecataloging.Config) *CreateSBOMConfig {
 	c.Files = cfg

syft/file/license.go

@@ -10,6 +10,7 @@ type License struct {
 	SPDXExpression  string
 	Type            license.Type
 	LicenseEvidence *LicenseEvidence // evidence from license classifier
+	Contents        string           `hash:"ignore"`
 }
 
 type LicenseEvidence struct {

syft/pkg/cataloger/golang/licenses.go

@@ -214,7 +214,7 @@ func (c *goLicenseResolver) findLicensesInFS(ctx context.Context, scanner licens
 		}
 		defer internal.CloseAndLogError(rdr, filePath)
 
-		parsed, err := licenses.Search(ctx, scanner, file.NewLocationReadCloser(file.NewLocation(filePath), rdr))
+		parsed, err := scanner.PkgSearch(ctx, file.NewLocationReadCloser(file.NewLocation(filePath), rdr))
 		if err != nil {
 			log.Debugf("error parsing license file %s: %v", filePath, err)
 			return nil
@@ -267,7 +267,7 @@ func (c *goLicenseResolver) parseLicenseFromLocation(ctx context.Context, scanne
 			return nil, err
 		}
 		defer internal.CloseAndLogError(contents, l.RealPath)
-		parsed, err := licenses.Search(ctx, scanner, file.NewLocationReadCloser(l, contents))
+		parsed, err := scanner.PkgSearch(ctx, file.NewLocationReadCloser(l, contents))
 		if err != nil {
 			return nil, err
 		}

syft/pkg/cataloger/golang/licenses_test.go

@@ -71,7 +71,12 @@ func Test_LicenseSearch(t *testing.T) {
 
 	localVendorDir := filepath.Join(wd, "test-fixtures", "licenses-vendor")
 
-	licenseScanner := licenses.NewScanner(licensecheck.Scan, float64(75))
+	sc := &licenses.ScannerConfig{
+		CoverageThreshold: 75,
+		Scanner:           licensecheck.Scan,
+	}
+	licenseScanner, err := licenses.NewScanner(sc)
+	require.NoError(t, err)
 
 	tests := []struct {
 		name     string
@@ -296,7 +301,10 @@ func Test_findVersionPath(t *testing.T) {
 
 func Test_walkDirErrors(t *testing.T) {
 	resolver := newGoLicenseResolver("", CatalogerConfig{})
-	_, err := resolver.findLicensesInFS(context.Background(), licenses.NewScanner(licensecheck.Scan, float64(75)), "somewhere", badFS{})
+	sc := &licenses.ScannerConfig{Scanner: licensecheck.Scan, CoverageThreshold: 75}
+	scanner, err := licenses.NewScanner(sc)
+	require.NoError(t, err)
+	_, err = resolver.findLicensesInFS(context.Background(), scanner, "somewhere", badFS{})
 	require.Error(t, err)
 }
 
@@ -314,7 +322,9 @@ func Test_noLocalGoModDir(t *testing.T) {
 	validTmp := t.TempDir()
 	require.NoError(t, os.MkdirAll(filepath.Join(validTmp, "mod@ver"), 0700|os.ModeDir))
 
-	licenseScanner := licenses.NewScanner(licensecheck.Scan, float64(75))
+	sc := &licenses.ScannerConfig{Scanner: licensecheck.Scan, CoverageThreshold: 75}
+	licenseScanner, err := licenses.NewScanner(sc)
+	require.NoError(t, err)
 	tests := []struct {
 		name    string
 		dir     string

syft/pkg/cataloger/golang/parse_go_binary.go

@@ -63,7 +63,10 @@ func newGoBinaryCataloger(opts CatalogerConfig) *goBinaryCataloger {
 func (c *goBinaryCataloger) parseGoBinary(ctx context.Context, resolver file.Resolver, _ *generic.Environment, reader file.LocationReadCloser) ([]pkg.Package, []artifact.Relationship, error) {
 	var pkgs []pkg.Package
 
-	licenseScanner := licenses.ContextLicenseScanner(ctx)
+	licenseScanner, err := licenses.ContextLicenseScanner(ctx)
+	if err != nil {
+		return nil, nil, err
+	}
 
 	unionReader, err := unionreader.GetUnionReader(reader.ReadCloser)
 	if err != nil {

syft/pkg/cataloger/golang/parse_go_binary_test.go

@@ -170,7 +170,9 @@ func TestBuildGoPkgInfo(t *testing.T) {
 		},
 	}
 
-	licenseScanner := licenses.NewScanner(licensecheck.Scan, float64(75))
+	sc := &licenses.ScannerConfig{Scanner: licensecheck.Scan, CoverageThreshold: 75}
+	licenseScanner, err := licenses.NewScanner(sc)
+	require.NoError(t, err)
 
 	tests := []struct {
 		name          string

syft/pkg/cataloger/golang/parse_go_mod.go

@@ -35,7 +35,10 @@ func newGoModCataloger(opts CatalogerConfig) *goModCataloger {
 func (c *goModCataloger) parseGoModFile(ctx context.Context, resolver file.Resolver, _ *generic.Environment, reader file.LocationReadCloser) ([]pkg.Package, []artifact.Relationship, error) {
 	packages := make(map[string]pkg.Package)
 
-	licenseScanner := licenses.ContextLicenseScanner(ctx)
+	licenseScanner, err := licenses.ContextLicenseScanner(ctx)
+	if err != nil {
+		return nil, nil, fmt.Errorf("unable to create default license scanner: %w", err)
+	}
 
 	contents, err := io.ReadAll(reader)
 	if err != nil {

syft/pkg/cataloger/java/archive_parser.go

@@ -96,7 +96,10 @@ func uniquePkgKey(groupID string, p *pkg.Package) string {
 // newJavaArchiveParser returns a new java archive parser object for the given archive. Can be configured to discover
 // and parse nested archives or ignore them.
 func newJavaArchiveParser(ctx context.Context, reader file.LocationReadCloser, detectNested bool, cfg ArchiveCatalogerConfig) (*archiveParser, func(), error) {
-	licenseScanner := licenses.ContextLicenseScanner(ctx)
+	licenseScanner, err := licenses.ContextLicenseScanner(ctx)
+	if err != nil {
+		return nil, nil, fmt.Errorf("could not build license scanner for java archive parser: %w", err)
+	}
 
 	// fetch the last element of the virtual path
 	virtualElements := strings.Split(reader.Path(), ":")
@@ -506,7 +509,7 @@ func (j *archiveParser) getLicenseFromFileInArchive(ctx context.Context) ([]pkg.
 			for _, licenseMatch := range licenseMatches {
 				licenseContents := contents[licenseMatch]
 				r := strings.NewReader(licenseContents)
-				parsed, err := licenses.Search(ctx, j.licenseScanner, file.NewLocationReadCloser(j.location, io.NopCloser(r)))
+				parsed, err := j.licenseScanner.PkgSearch(ctx, file.NewLocationReadCloser(j.location, io.NopCloser(r)))
 				if err != nil {
 					return nil, err
 				}

syft/pkg/cataloger/java/archive_parser_test.go

@@ -32,8 +32,10 @@ import (
 
 func TestSearchMavenForLicenses(t *testing.T) {
 	url := maventest.MockRepo(t, "internal/maven/test-fixtures/maven-repo")
-
+	sc := &licenses.ScannerConfig{Scanner: licensecheck.Scan, CoverageThreshold: 75}
-	ctx := licenses.SetContextLicenseScanner(context.Background(), licenses.NewScanner(licensecheck.Scan, float64(75)))
+	scanner, err := licenses.NewScanner(sc)
+	require.NoError(t, err)
+	ctx := licenses.SetContextLicenseScanner(context.Background(), scanner)
 
 	tests := []struct {
 		name             string
@@ -92,7 +94,10 @@ func TestSearchMavenForLicenses(t *testing.T) {
 }
 
 func TestParseJar(t *testing.T) {
-	ctx := licenses.SetContextLicenseScanner(context.Background(), licenses.NewScanner(licensecheck.Scan, float64(75)))
+	sc := &licenses.ScannerConfig{Scanner: licensecheck.Scan, CoverageThreshold: 75}
+	scanner, err := licenses.NewScanner(sc)
+	require.NoError(t, err)
+	ctx := licenses.SetContextLicenseScanner(context.Background(), scanner)
 
 	tests := []struct {
 		name         string
@@ -1375,7 +1380,10 @@ func Test_parseJavaArchive_regressions(t *testing.T) {
 }
 
 func Test_deterministicMatchingPomProperties(t *testing.T) {
-	ctx := licenses.SetContextLicenseScanner(context.Background(), licenses.NewScanner(licensecheck.Scan, float64(75)))
+	sc := &licenses.ScannerConfig{Scanner: licensecheck.Scan, CoverageThreshold: 75}
+	scanner, err := licenses.NewScanner(sc)
+	require.NoError(t, err)
+	ctx := licenses.SetContextLicenseScanner(context.Background(), scanner)
 
 	tests := []struct {
 		fixture  string

syft/pkg/cataloger/python/package.go

@@ -109,7 +109,7 @@ func findLicenses(ctx context.Context, scanner licenses.Scanner, resolver file.R
 		if len(found) > 0 {
 			metadataContents, err := resolver.FileContentsByLocation(found[0])
 			if err == nil {
-				parsed, err := licenses.Search(ctx, scanner, file.NewLocationReadCloser(m.LicenseLocation, metadataContents))
+				parsed, err := scanner.PkgSearch(ctx, file.NewLocationReadCloser(m.LicenseLocation, metadataContents))
 				if err != nil {
 					log.WithFields("error", err).Tracef("unable to parse a license from the file in %s", m.LicenseLocation.Path())
 				}

syft/pkg/cataloger/python/parse_wheel_egg.go

@@ -20,8 +20,10 @@ import (
 // parseWheelOrEgg takes the primary metadata file reference and returns the python package it represents. Contained
 // fields are governed by the PyPA core metadata specification (https://packaging.python.org/en/latest/specifications/core-metadata/).
 func parseWheelOrEgg(ctx context.Context, resolver file.Resolver, _ *generic.Environment, reader file.LocationReadCloser) ([]pkg.Package, []artifact.Relationship, error) {
-	licenseScanner := licenses.ContextLicenseScanner(ctx)
+	licenseScanner, err := licenses.ContextLicenseScanner(ctx)
-
+	if err != nil {
+		return nil, nil, err
+	}
 	pd, sources, err := assembleEggOrWheelMetadata(resolver, reader.Location)
 	if err != nil {
 		return nil, nil, err

test/cli/license_test.go

@@ -0,0 +1,45 @@
+package cli
+
+import "testing"
+
+func Test_Licenses(t *testing.T) {
+	testImage := getFixtureImage(t, "image-unknown-licenses")
+	tests := []struct {
+		name       string
+		args       []string
+		env        map[string]string
+		assertions []traitAssertion
+	}{
+		{
+			name: "licenses default with no content",
+			args: []string{"scan", "-o", "json", testImage, "--from", "docker-archive"},
+			env:  map[string]string{"SYFT_FORMAT_PRETTY": "true"},
+			assertions: []traitAssertion{
+				assertJsonReport,
+				assertUnknownLicenseContent(false),
+				assertSuccessfulReturnCode,
+			},
+		},
+
+		{
+			name: "licenses with content",
+			args: []string{"scan", "-o", "json", testImage, "--from", "docker-archive"},
+			env:  map[string]string{"SYFT_FORMAT_PRETTY": "true", "SYFT_LICENSE_INCLUDE_UNKNOWN_LICENSE_CONTENT": "true"},
+			assertions: []traitAssertion{
+				assertJsonReport,
+				assertUnknownLicenseContent(true),
+				assertSuccessfulReturnCode,
+			},
+		},
+	}
+
+	for _, test := range tests {
+		t.Run(test.name, func(t *testing.T) {
+			cmd, stdout, stderr := runSyft(t, test.env, test.args...)
+			for _, traitFn := range test.assertions {
+				traitFn(t, stdout, stderr, cmd.ProcessState.ExitCode())
+			}
+			logOutputOnFailure(t, cmd, stdout, stderr)
+		})
+	}
+}

test/cli/test-fixtures/image-unknown-licenses/Dockerfile

@@ -0,0 +1,3 @@
+FROM alpine@sha256:c5c5fda71656f28e49ac9c5416b3643eaa6a108a8093151d6d1afc9463be8e33
+RUN rm -rf /lib/apk/db/installed
+COPY . /home/files

test/cli/trait_assertions_test.go

@@ -10,7 +10,10 @@ import (
 	"testing"
 
 	"github.com/acarl005/stripansi"
+	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
+
+	"github.com/anchore/syft/syft/format/syftjson/model"
 )
 
 type traitAssertion func(tb testing.TB, stdout, stderr string, rc int)
@@ -148,6 +151,34 @@ func assertPackageCount(length uint) traitAssertion {
 	}
 }
 
+func assertUnknownLicenseContent(required bool) traitAssertion {
+	return func(tb testing.TB, stdout, _ string, _ int) {
+		tb.Helper()
+		type NameAndLicense struct {
+			Name     string          `json:"name"`
+			Licenses []model.License `json:"Licenses"`
+		}
+		type partial struct {
+			Artifacts []NameAndLicense `json:"artifacts"`
+		}
+
+		var data partial
+		if err := json.Unmarshal([]byte(stdout), &data); err != nil {
+			tb.Errorf("expected to find a JSON report, but was unmarshalable: %+v", err)
+		}
+
+		for _, pkg := range data.Artifacts {
+			for _, lic := range pkg.Licenses {
+				if strings.Contains(lic.SPDXExpression, "UNKNOWN") && required {
+					assert.NotZero(tb, len(lic.Contents))
+				} else {
+					assert.Empty(tb, lic.Contents)
+				}
+			}
+		}
+	}
+}
+
 func assertFailingReturnCode(tb testing.TB, _, _ string, rc int) {
 	tb.Helper()
 	if rc == 0 {