Performance improvements around package ID (#698)

* set package ID in catalogers and improve hashing performance

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

* update setting ID + tests

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

syft/pkg/cataloger/golang/parse_go_bin.go

@@ -17,6 +17,28 @@ const (
 
 type exeOpener func(file io.ReadCloser) ([]exe, error)
 
+func newGoBinaryPackage(name, version, h1Digest, goVersion, architecture string, location source.Location) pkg.Package {
+	p := pkg.Package{
+		Name:     name,
+		Version:  version,
+		Language: pkg.Go,
+		Type:     pkg.GoModulePkg,
+		Locations: []source.Location{
+			location,
+		},
+		MetadataType: pkg.GolangBinMetadataType,
+		Metadata: pkg.GolangBinMetadata{
+			GoCompiledVersion: goVersion,
+			H1Digest:          h1Digest,
+			Architecture:      architecture,
+		},
+	}
+
+	p.SetID()
+
+	return p
+}
+
 func parseGoBin(location source.Location, reader io.ReadCloser, opener exeOpener) (pkgs []pkg.Package, err error) {
 	var exes []exe
 	// it has been found that there are stdlib paths within openExe that can panic. We want to prevent this behavior
@@ -54,21 +76,10 @@ func buildGoPkgInfo(location source.Location, mod, goVersion, arch string) []pkg
 		}
 
 		if fields[0] == packageIdentifier || fields[0] == replaceIdentifier {
-			pkgsSlice = append(pkgsSlice, pkg.Package{
-				Name:     fields[1],
-				Version:  fields[2],
-				Language: pkg.Go,
-				Type:     pkg.GoModulePkg,
-				Locations: []source.Location{
-					location,
-				},
-				MetadataType: pkg.GolangBinMetadataType,
-				Metadata: pkg.GolangBinMetadata{
-					GoCompiledVersion: goVersion,
-					H1Digest:          fields[3],
-					Architecture:      arch,
-				},
-			})
+			name := fields[1]
+			version := fields[2]
+			h1Digest := fields[3]
+			pkgsSlice = append(pkgsSlice, newGoBinaryPackage(name, version, h1Digest, goVersion, arch, location))
 		}
 	}
 

syft/pkg/cataloger/golang/parse_go_bin_test.go

@@ -147,6 +147,10 @@ func TestBuildGoPkgInfo(t *testing.T) {
 
 	for _, test := range tests {
 		t.Run(test.name, func(t *testing.T) {
+			for i := range test.expected {
+				p := &test.expected[i]
+				p.SetID()
+			}
 			location := source.Location{
 				Coordinates: source.Coordinates{
 					RealPath:     "/a-path",

syft/pkg/cataloger/golang/parse_go_mod.go

@@ -12,8 +12,8 @@ import (
 )
 
 // parseGoMod takes a go.mod and lists all packages discovered.
-func parseGoMod(path string, reader io.Reader) ([]pkg.Package, []artifact.Relationship, error) {
-	packages := make(map[string]pkg.Package)
+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 {
@@ -26,7 +26,7 @@ func parseGoMod(path string, reader io.Reader) ([]pkg.Package, []artifact.Relati
 	}
 
 	for _, m := range file.Require {
-		packages[m.Mod.Path] = pkg.Package{
+		packages[m.Mod.Path] = &pkg.Package{
 			Name:     m.Mod.Path,
 			Version:  m.Mod.Version,
 			Language: pkg.Go,
@@ -36,7 +36,7 @@ func parseGoMod(path string, reader io.Reader) ([]pkg.Package, []artifact.Relati
 
 	// remove any old packages and replace with new ones...
 	for _, m := range file.Replace {
-		packages[m.New.Path] = pkg.Package{
+		packages[m.New.Path] = &pkg.Package{
 			Name:     m.New.Path,
 			Version:  m.New.Version,
 			Language: pkg.Go,
@@ -49,7 +49,7 @@ func parseGoMod(path string, reader io.Reader) ([]pkg.Package, []artifact.Relati
 		delete(packages, m.Mod.Path)
 	}
 
-	pkgsSlice := make([]pkg.Package, len(packages))
+	pkgsSlice := make([]*pkg.Package, len(packages))
 	idx := 0
 	for _, p := range packages {
 		pkgsSlice[idx] = p

syft/pkg/cataloger/golang/parse_go_mod_test.go

@@ -87,7 +87,7 @@ func TestParseGoMod(t *testing.T) {
 					continue
 				}
 
-				diffs := deep.Equal(a, e)
+				diffs := deep.Equal(a, &e)
 				if len(diffs) > 0 {
 					t.Errorf("diffs found for %q", a.Name)
 					for _, d := range diffs {