Create single license scanner for all catalogers (#3348)

* add single license scanner instance

Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>

* rename testing license scanner

Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>

---------

Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
This commit is contained in:
Alex Goodman
2024-10-21 16:17:12 +00:00
committed by GitHub
parent 14355aac21
commit e4e985b9b0
16 changed files with 277 additions and 148 deletions
+16 -14
View File
@@ -3,6 +3,7 @@ package golang
import (
"archive/zip"
"bytes"
"context"
"fmt"
"io"
"io/fs"
@@ -79,9 +80,9 @@ func remotesForModule(proxies []string, noProxy []string, module string) []strin
return proxies
}
func (c *goLicenseResolver) getLicenses(resolver file.Resolver, moduleName, moduleVersion string) ([]pkg.License, error) {
func (c *goLicenseResolver) getLicenses(ctx context.Context, scanner licenses.Scanner, resolver file.Resolver, moduleName, moduleVersion string) ([]pkg.License, error) {
// search the scan target first, ignoring local and remote sources
goLicenses, err := c.findLicensesInSource(resolver,
goLicenses, err := c.findLicensesInSource(ctx, scanner, resolver,
fmt.Sprintf(`**/go/pkg/mod/%s@%s/*`, processCaps(moduleName), moduleVersion),
)
if err != nil || len(goLicenses) > 0 {
@@ -90,7 +91,7 @@ func (c *goLicenseResolver) getLicenses(resolver file.Resolver, moduleName, modu
// look in the local host mod directory...
if c.opts.SearchLocalModCacheLicenses {
goLicenses, err = c.getLicensesFromLocal(moduleName, moduleVersion)
goLicenses, err = c.getLicensesFromLocal(ctx, scanner, moduleName, moduleVersion)
if err != nil || len(goLicenses) > 0 {
return toPkgLicenses(goLicenses), err
}
@@ -98,13 +99,13 @@ func (c *goLicenseResolver) getLicenses(resolver file.Resolver, moduleName, modu
// download from remote sources
if c.opts.SearchRemoteLicenses {
goLicenses, err = c.getLicensesFromRemote(moduleName, moduleVersion)
goLicenses, err = c.getLicensesFromRemote(ctx, scanner, moduleName, moduleVersion)
}
return toPkgLicenses(goLicenses), err
}
func (c *goLicenseResolver) getLicensesFromLocal(moduleName, moduleVersion string) ([]goLicense, error) {
func (c *goLicenseResolver) getLicensesFromLocal(ctx context.Context, scanner licenses.Scanner, moduleName, moduleVersion string) ([]goLicense, error) {
if c.localModCacheDir == nil {
return nil, nil
}
@@ -120,10 +121,10 @@ func (c *goLicenseResolver) getLicensesFromLocal(moduleName, moduleVersion strin
// if we're running against a directory on the filesystem, it may not include the
// user's homedir / GOPATH, so we defer to using the localModCacheResolver
// we use $GOPATH/pkg/mod to avoid leaking information about the user's system
return c.findLicensesInFS("file://$GOPATH/pkg/mod/"+subdir+"/", dir)
return c.findLicensesInFS(ctx, scanner, "file://$GOPATH/pkg/mod/"+subdir+"/", dir)
}
func (c *goLicenseResolver) getLicensesFromRemote(moduleName, moduleVersion string) ([]goLicense, error) {
func (c *goLicenseResolver) getLicensesFromRemote(ctx context.Context, scanner licenses.Scanner, moduleName, moduleVersion string) ([]goLicense, error) {
return c.licenseCache.Resolve(fmt.Sprintf("%s/%s", moduleName, moduleVersion), func() ([]goLicense, error) {
proxies := remotesForModule(c.opts.Proxies, c.opts.NoProxy, moduleName)
@@ -132,11 +133,11 @@ func (c *goLicenseResolver) getLicensesFromRemote(moduleName, moduleVersion stri
return nil, err
}
return c.findLicensesInFS(urlPrefix, fsys)
return c.findLicensesInFS(ctx, scanner, urlPrefix, fsys)
})
}
func (c *goLicenseResolver) findLicensesInFS(urlPrefix string, fsys fs.FS) ([]goLicense, error) {
func (c *goLicenseResolver) findLicensesInFS(ctx context.Context, scanner licenses.Scanner, urlPrefix string, fsys fs.FS) ([]goLicense, error) {
var out []goLicense
err := fs.WalkDir(fsys, ".", func(filePath string, d fs.DirEntry, err error) error {
if err != nil {
@@ -156,7 +157,8 @@ func (c *goLicenseResolver) findLicensesInFS(urlPrefix string, fsys fs.FS) ([]go
return nil
}
defer internal.CloseAndLogError(rdr, filePath)
parsed, err := licenses.Parse(rdr, file.NewLocation(filePath))
parsed, err := licenses.Search(ctx, scanner, file.NewLocationReadCloser(file.NewLocation(filePath), rdr))
if err != nil {
log.Debugf("error parsing license file %s: %v", filePath, err)
return nil
@@ -174,7 +176,7 @@ func (c *goLicenseResolver) findLicensesInFS(urlPrefix string, fsys fs.FS) ([]go
return out, err
}
func (c *goLicenseResolver) findLicensesInSource(resolver file.Resolver, globMatch string) ([]goLicense, error) {
func (c *goLicenseResolver) findLicensesInSource(ctx context.Context, scanner licenses.Scanner, resolver file.Resolver, globMatch string) ([]goLicense, error) {
var out []goLicense
locations, err := resolver.FilesByGlob(globMatch)
if err != nil {
@@ -182,7 +184,7 @@ func (c *goLicenseResolver) findLicensesInSource(resolver file.Resolver, globMat
}
for _, l := range locations {
parsed, err := c.parseLicenseFromLocation(l, resolver)
parsed, err := c.parseLicenseFromLocation(ctx, scanner, l, resolver)
if err != nil {
return nil, err
}
@@ -200,7 +202,7 @@ func (c *goLicenseResolver) findLicensesInSource(resolver file.Resolver, globMat
return out, nil
}
func (c *goLicenseResolver) parseLicenseFromLocation(l file.Location, resolver file.Resolver) ([]goLicense, error) {
func (c *goLicenseResolver) parseLicenseFromLocation(ctx context.Context, scanner licenses.Scanner, l file.Location, resolver file.Resolver) ([]goLicense, error) {
var out []goLicense
fileName := path.Base(l.RealPath)
if c.lowerLicenseFileNames.Has(strings.ToLower(fileName)) {
@@ -209,7 +211,7 @@ func (c *goLicenseResolver) parseLicenseFromLocation(l file.Location, resolver f
return nil, err
}
defer internal.CloseAndLogError(contents, l.RealPath)
parsed, err := licenses.Parse(contents, l)
parsed, err := licenses.Search(ctx, scanner, file.NewLocationReadCloser(l, contents))
if err != nil {
return nil, err
}
+16 -8
View File
@@ -3,6 +3,7 @@ package golang
import (
"archive/zip"
"bytes"
"context"
"fmt"
"io/fs"
"net/http"
@@ -15,6 +16,7 @@ import (
"github.com/stretchr/testify/require"
"github.com/anchore/syft/internal/licenses"
"github.com/anchore/syft/syft/file"
"github.com/anchore/syft/syft/internal/fileresolver"
"github.com/anchore/syft/syft/license"
@@ -26,6 +28,8 @@ func Test_LocalLicenseSearch(t *testing.T) {
loc2 := file.NewLocation("github.com/!cap!o!r!g/!cap!project@v4.111.5/LICENSE.txt")
loc3 := file.NewLocation("github.com/someorg/strangelicense@v1.2.3/LiCeNsE.tXt")
licenseScanner := licenses.TestingOnlyScanner()
tests := []struct {
name string
version string
@@ -78,12 +82,12 @@ func Test_LocalLicenseSearch(t *testing.T) {
LocalModCacheDir: filepath.Join(wd, "test-fixtures", "licenses", "pkg", "mod"),
},
)
licenses, err := l.getLicenses(fileresolver.Empty{}, test.name, test.version)
lics, err := l.getLicenses(context.Background(), licenseScanner, fileresolver.Empty{}, test.name, test.version)
require.NoError(t, err)
require.Len(t, licenses, 1)
require.Len(t, lics, 1)
require.Equal(t, test.expected, licenses[0])
require.Equal(t, test.expected, lics[0])
})
}
}
@@ -92,6 +96,8 @@ func Test_RemoteProxyLicenseSearch(t *testing.T) {
loc1 := file.NewLocation("github.com/someorg/somename@v0.3.2/LICENSE")
loc2 := file.NewLocation("github.com/!cap!o!r!g/!cap!project@v4.111.5/LICENSE.txt")
licenseScanner := licenses.TestingOnlyScanner()
server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
buf := &bytes.Buffer{}
uri := strings.TrimPrefix(strings.TrimSuffix(r.RequestURI, ".zip"), "/")
@@ -169,12 +175,12 @@ func Test_RemoteProxyLicenseSearch(t *testing.T) {
},
)
licenses, err := l.getLicenses(fileresolver.Empty{}, test.name, test.version)
lics, err := l.getLicenses(context.Background(), licenseScanner, fileresolver.Empty{}, test.name, test.version)
require.NoError(t, err)
require.Len(t, licenses, 1)
require.Len(t, lics, 1)
require.Equal(t, test.expected, licenses[0])
require.Equal(t, test.expected, lics[0])
})
}
}
@@ -248,7 +254,7 @@ func Test_findVersionPath(t *testing.T) {
func Test_walkDirErrors(t *testing.T) {
resolver := newGoLicenseResolver("", CatalogerConfig{})
_, err := resolver.findLicensesInFS("somewhere", badFS{})
_, err := resolver.findLicensesInFS(context.Background(), licenses.TestingOnlyScanner(), "somewhere", badFS{})
require.Error(t, err)
}
@@ -266,6 +272,8 @@ func Test_noLocalGoModDir(t *testing.T) {
validTmp := t.TempDir()
require.NoError(t, os.MkdirAll(filepath.Join(validTmp, "mod@ver"), 0700|os.ModeDir))
licenseScanner := licenses.TestingOnlyScanner()
tests := []struct {
name string
dir string
@@ -299,7 +307,7 @@ func Test_noLocalGoModDir(t *testing.T) {
SearchLocalModCacheLicenses: true,
LocalModCacheDir: test.dir,
})
_, err := resolver.getLicensesFromLocal("mod", "ver")
_, err := resolver.getLicensesFromLocal(context.Background(), licenseScanner, "mod", "ver")
test.wantErr(t, err)
})
}
+1 -7
View File
@@ -5,21 +5,15 @@ import (
"strings"
"github.com/anchore/packageurl-go"
"github.com/anchore/syft/internal/log"
"github.com/anchore/syft/syft/file"
"github.com/anchore/syft/syft/pkg"
)
func (c *goBinaryCataloger) newGoBinaryPackage(resolver file.Resolver, dep *debug.Module, mainModule, goVersion, architecture string, buildSettings pkg.KeyValues, cryptoSettings, experiments []string, locations ...file.Location) pkg.Package {
func (c *goBinaryCataloger) newGoBinaryPackage(dep *debug.Module, mainModule, goVersion, architecture string, buildSettings pkg.KeyValues, cryptoSettings, experiments []string, licenses []pkg.License, locations ...file.Location) pkg.Package {
if dep.Replace != nil {
dep = dep.Replace
}
licenses, err := c.licenseResolver.getLicenses(resolver, dep.Path, dep.Version)
if err != nil {
log.Tracef("error getting licenses for golang package: %s %v", dep.Path, err)
}
p := pkg.Package{
Name: dep.Path,
Version: dep.Version,
+21 -7
View File
@@ -18,6 +18,7 @@ import (
"golang.org/x/mod/module"
"github.com/anchore/syft/internal"
"github.com/anchore/syft/internal/licenses"
"github.com/anchore/syft/internal/log"
"github.com/anchore/syft/syft/artifact"
"github.com/anchore/syft/syft/file"
@@ -59,9 +60,11 @@ func newGoBinaryCataloger(opts CatalogerConfig) *goBinaryCataloger {
}
// parseGoBinary catalogs packages found in the "buildinfo" section of a binary built by the go compiler.
func (c *goBinaryCataloger) parseGoBinary(_ context.Context, resolver file.Resolver, _ *generic.Environment, reader file.LocationReadCloser) ([]pkg.Package, []artifact.Relationship, error) {
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)
unionReader, err := unionreader.GetUnionReader(reader.ReadCloser)
if err != nil {
return nil, nil, err
@@ -73,7 +76,7 @@ func (c *goBinaryCataloger) parseGoBinary(_ context.Context, resolver file.Resol
var rels []artifact.Relationship
for _, mod := range mods {
var depPkgs []pkg.Package
mainPkg, depPkgs := c.buildGoPkgInfo(resolver, reader.Location, mod, mod.arch, unionReader)
mainPkg, depPkgs := c.buildGoPkgInfo(ctx, licenseScanner, resolver, reader.Location, mod, mod.arch, unionReader)
if mainPkg != nil {
rels = createModuleRelationships(*mainPkg, depPkgs)
pkgs = append(pkgs, *mainPkg)
@@ -101,7 +104,7 @@ func createModuleRelationships(main pkg.Package, deps []pkg.Package) []artifact.
var emptyModule debug.Module
var moduleFromPartialPackageBuild = debug.Module{Path: "command-line-arguments"}
func (c *goBinaryCataloger) buildGoPkgInfo(resolver file.Resolver, location file.Location, mod *extendedBuildInfo, arch string, reader io.ReadSeekCloser) (*pkg.Package, []pkg.Package) {
func (c *goBinaryCataloger) buildGoPkgInfo(ctx context.Context, licenseScanner licenses.Scanner, resolver file.Resolver, location file.Location, mod *extendedBuildInfo, arch string, reader io.ReadSeekCloser) (*pkg.Package, []pkg.Package) {
if mod == nil {
return nil, nil
}
@@ -116,9 +119,13 @@ func (c *goBinaryCataloger) buildGoPkgInfo(resolver file.Resolver, location file
continue
}
lics, err := c.licenseResolver.getLicenses(ctx, licenseScanner, resolver, dep.Path, dep.Version)
if err != nil {
log.Tracef("error getting licenses for golang package: %s %v", dep.Path, err)
}
gover, experiments := getExperimentsFromVersion(mod.GoVersion)
p := c.newGoBinaryPackage(
resolver,
dep,
mod.Main.Path,
gover,
@@ -126,6 +133,7 @@ func (c *goBinaryCataloger) buildGoPkgInfo(resolver file.Resolver, location file
nil,
mod.cryptoSettings,
experiments,
lics,
location.WithAnnotation(pkg.EvidenceAnnotationKey, pkg.PrimaryEvidenceAnnotation),
)
if pkg.IsValid(&p) {
@@ -137,7 +145,7 @@ func (c *goBinaryCataloger) buildGoPkgInfo(resolver file.Resolver, location file
return nil, pkgs
}
main := c.makeGoMainPackage(resolver, mod, arch, location, reader)
main := c.makeGoMainPackage(ctx, licenseScanner, resolver, mod, arch, location, reader)
return &main, pkgs
}
@@ -152,11 +160,16 @@ func missingMainModule(mod *extendedBuildInfo) bool {
return mod.Main == moduleFromPartialPackageBuild
}
func (c *goBinaryCataloger) makeGoMainPackage(resolver file.Resolver, mod *extendedBuildInfo, arch string, location file.Location, reader io.ReadSeekCloser) pkg.Package {
func (c *goBinaryCataloger) makeGoMainPackage(ctx context.Context, licenseScanner licenses.Scanner, resolver file.Resolver, mod *extendedBuildInfo, arch string, location file.Location, reader io.ReadSeekCloser) pkg.Package {
gbs := getBuildSettings(mod.Settings)
lics, err := c.licenseResolver.getLicenses(ctx, licenseScanner, resolver, mod.Main.Path, mod.Main.Version)
if err != nil {
log.Tracef("error getting licenses for golang package: %s %v", mod.Main.Path, err)
}
gover, experiments := getExperimentsFromVersion(mod.GoVersion)
main := c.newGoBinaryPackage(
resolver,
&mod.Main,
mod.Main.Path,
gover,
@@ -164,6 +177,7 @@ func (c *goBinaryCataloger) makeGoMainPackage(resolver file.Resolver, mod *exten
gbs,
mod.cryptoSettings,
experiments,
lics,
location.WithAnnotation(pkg.EvidenceAnnotationKey, pkg.PrimaryEvidenceAnnotation),
)
@@ -3,6 +3,7 @@ package golang
import (
"bufio"
"bytes"
"context"
"errors"
"io"
"os"
@@ -17,6 +18,7 @@ import (
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
"github.com/anchore/syft/internal/licenses"
"github.com/anchore/syft/syft/file"
"github.com/anchore/syft/syft/internal/fileresolver"
"github.com/anchore/syft/syft/internal/unionreader"
@@ -167,6 +169,8 @@ func TestBuildGoPkgInfo(t *testing.T) {
},
}
licenseScanner := licenses.TestingOnlyScanner()
tests := []struct {
name string
mod *extendedBuildInfo
@@ -1053,7 +1057,7 @@ func TestBuildGoPkgInfo(t *testing.T) {
c := newGoBinaryCataloger(DefaultCatalogerConfig())
reader, err := unionreader.GetUnionReader(io.NopCloser(strings.NewReader(test.binaryContent)))
require.NoError(t, err)
mainPkg, pkgs := c.buildGoPkgInfo(fileresolver.Empty{}, location, test.mod, test.mod.arch, reader)
mainPkg, pkgs := c.buildGoPkgInfo(context.Background(), licenseScanner, fileresolver.Empty{}, location, test.mod, test.mod.arch, reader)
if mainPkg != nil {
pkgs = append(pkgs, *mainPkg)
}
+8 -5
View File
@@ -11,6 +11,7 @@ import (
"golang.org/x/mod/modfile"
"github.com/anchore/syft/internal"
"github.com/anchore/syft/internal/licenses"
"github.com/anchore/syft/internal/log"
"github.com/anchore/syft/syft/artifact"
"github.com/anchore/syft/syft/file"
@@ -31,9 +32,11 @@ func newGoModCataloger(opts CatalogerConfig) *goModCataloger {
// parseGoModFile takes a go.mod and lists all packages discovered.
//
//nolint:funlen
func (c *goModCataloger) parseGoModFile(_ context.Context, resolver file.Resolver, _ *generic.Environment, reader file.LocationReadCloser) ([]pkg.Package, []artifact.Relationship, error) {
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)
contents, err := io.ReadAll(reader)
if err != nil {
return nil, nil, fmt.Errorf("failed to read go module: %w", err)
@@ -50,7 +53,7 @@ func (c *goModCataloger) parseGoModFile(_ context.Context, resolver file.Resolve
}
for _, m := range f.Require {
licenses, err := c.licenseResolver.getLicenses(resolver, m.Mod.Path, m.Mod.Version)
lics, err := c.licenseResolver.getLicenses(ctx, licenseScanner, resolver, m.Mod.Path, m.Mod.Version)
if err != nil {
log.Tracef("error getting licenses for package: %s %v", m.Mod.Path, err)
}
@@ -58,7 +61,7 @@ func (c *goModCataloger) parseGoModFile(_ context.Context, resolver file.Resolve
packages[m.Mod.Path] = pkg.Package{
Name: m.Mod.Path,
Version: m.Mod.Version,
Licenses: pkg.NewLicenseSet(licenses...),
Licenses: pkg.NewLicenseSet(lics...),
Locations: file.NewLocationSet(reader.Location.WithAnnotation(pkg.EvidenceAnnotationKey, pkg.PrimaryEvidenceAnnotation)),
PURL: packageURL(m.Mod.Path, m.Mod.Version),
Language: pkg.Go,
@@ -71,7 +74,7 @@ func (c *goModCataloger) parseGoModFile(_ context.Context, resolver file.Resolve
// remove any old packages and replace with new ones...
for _, m := range f.Replace {
licenses, err := c.licenseResolver.getLicenses(resolver, m.New.Path, m.New.Version)
lics, err := c.licenseResolver.getLicenses(ctx, licenseScanner, resolver, m.New.Path, m.New.Version)
if err != nil {
log.Tracef("error getting licenses for package: %s %v", m.New.Path, err)
}
@@ -83,7 +86,7 @@ func (c *goModCataloger) parseGoModFile(_ context.Context, resolver file.Resolve
packages[m.New.Path] = pkg.Package{
Name: m.New.Path,
Version: m.New.Version,
Licenses: pkg.NewLicenseSet(licenses...),
Licenses: pkg.NewLicenseSet(lics...),
Locations: file.NewLocationSet(reader.Location.WithAnnotation(pkg.EvidenceAnnotationKey, pkg.PrimaryEvidenceAnnotation)),
PURL: packageURL(m.New.Path, m.New.Version),
Language: pkg.Go,