detect license ID from full text when incidentally provided as a value (#3876)

---------
Signed-off-by: Christopher Phillips <32073428+spiffcs@users.noreply.github.com>
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
Co-authored-by: Alex Goodman <wagoodman@users.noreply.github.com>

cmd/syft/internal/options/catalog.go

@@ -158,8 +158,8 @@ func (cfg Catalog) ToFilesConfig() filecataloging.Config {
 
 func (cfg Catalog) ToLicenseConfig() cataloging.LicenseConfig {
 	return cataloging.LicenseConfig{
-		IncludeUnkownLicenseContent: cfg.License.IncludeUnknownLicenseContent,
-		Coverage:                    cfg.License.LicenseCoverage,
+		IncludeContent: cfg.License.Content,
+		Coverage:       cfg.License.Coverage,
 	}
 }
 

cmd/syft/internal/options/license.go

@@ -1,12 +1,22 @@
 package options
 
 import (
+	"fmt"
+
 	"github.com/anchore/clio"
+	"github.com/anchore/syft/syft/cataloging"
 )
 
 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"`
+	Content cataloging.LicenseContent `yaml:"content" json:"content" mapstructure:"content"`
+	// Deprecated: please use include-license-content instead
+	IncludeUnknownLicenseContent *bool `yaml:"-" json:"-" mapstructure:"include-unknown-license-content"`
+
+	Coverage float64 `yaml:"coverage" json:"coverage" mapstructure:"coverage"`
+	// Deprecated: please use coverage instead
+	LicenseCoverage *float64 `yaml:"license-coverage" json:"license-coverage" mapstructure:"license-coverage"`
+
+	AvailableLicenseContent []cataloging.LicenseContent `yaml:"-" json:"-" mapstructure:"-"`
 }
 
 var _ interface {
@@ -14,15 +24,56 @@ var _ interface {
 } = (*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
+	descriptions.Add(&o.Content, fmt.Sprintf("include the content of licenses in the SBOM for a given syft scan; valid values are: %s", o.AvailableLicenseContent))
+	descriptions.Add(&o.IncludeUnknownLicenseContent, `deprecated: please use 'license-content' instead`)
+
+	descriptions.Add(&o.Coverage, `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.`)
+	descriptions.Add(&o.LicenseCoverage, `deprecated: please use 'coverage' instead`)
+}
+
+func (o *licenseConfig) PostLoad() error {
+	cfg := cataloging.DefaultLicenseConfig()
+	defaultContent := cfg.IncludeContent
+	defaultCoverage := cfg.Coverage
+
+	// if both legacy and new fields are specified, error out
+	if o.IncludeUnknownLicenseContent != nil && o.Content != defaultContent {
+		return fmt.Errorf("both 'include-unknown-license-content' and 'content' are set, please use only 'content'")
+	}
+
+	if o.LicenseCoverage != nil && o.Coverage != defaultCoverage {
+		return fmt.Errorf("both 'license-coverage' and 'coverage' are set, please use only 'coverage'")
+	}
+
+	// finalize the license content value
+	if o.IncludeUnknownLicenseContent != nil {
+		// convert 'include-unknown-license-content' -> 'license-content'
+		v := cataloging.LicenseContentExcludeAll
+		if *o.IncludeUnknownLicenseContent {
+			v = cataloging.LicenseContentIncludeUnknown
+		}
+		o.Content = v
+	}
+
+	// finalize the coverage value
+	if o.LicenseCoverage != nil {
+		// convert 'license-coverage' -> 'coverage'
+		o.Coverage = *o.LicenseCoverage
+	}
+
+	return nil
 }
 
 func defaultLicenseConfig() licenseConfig {
+	cfg := cataloging.DefaultLicenseConfig()
 	return licenseConfig{
-		IncludeUnknownLicenseContent: false,
-		LicenseCoverage:              75,
+		Content:  cfg.IncludeContent,
+		Coverage: cfg.Coverage,
+		AvailableLicenseContent: []cataloging.LicenseContent{
+			cataloging.LicenseContentIncludeAll,
+			cataloging.LicenseContentIncludeUnknown,
+			cataloging.LicenseContentExcludeAll,
+		},
 	}
 }