Follow convention for naming catalogers (#2277)

* follow convention for naming catalogers

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

* fix cataloger name example

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

---------

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

syft/pkg/cataloger/gentoo/cataloger.go

@@ -0,0 +1,14 @@
+/*
+Package gentoo provides a concrete Cataloger implementation related to packages within the Gentoo linux ecosystem.
+*/
+package gentoo
+
+import (
+	"github.com/anchore/syft/syft/pkg/cataloger/generic"
+)
+
+// NewPortageCataloger returns a new cataloger object initialized for Gentoo Portage package manager files (a flat-file store).
+func NewPortageCataloger() *generic.Cataloger {
+	return generic.NewCataloger("portage-cataloger").
+		WithParserByGlobs(parsePortageContents, "**/var/db/pkg/*/*/CONTENTS")
+}

syft/pkg/cataloger/gentoo/cataloger_test.go

@@ -0,0 +1,96 @@
+package gentoo
+
+import (
+	"testing"
+
+	"github.com/anchore/syft/syft/artifact"
+	"github.com/anchore/syft/syft/file"
+	"github.com/anchore/syft/syft/pkg"
+	"github.com/anchore/syft/syft/pkg/cataloger/internal/pkgtest"
+)
+
+func TestPortageCataloger(t *testing.T) {
+	expectedLicenseLocation := file.NewLocation("var/db/pkg/app-containers/skopeo-1.5.1/LICENSE")
+	expectedPkgs := []pkg.Package{
+		{
+			Name:    "app-containers/skopeo",
+			Version: "1.5.1",
+			FoundBy: "portage-cataloger",
+			PURL:    "pkg:ebuild/app-containers/skopeo@1.5.1",
+			Locations: file.NewLocationSet(
+				file.NewLocation("var/db/pkg/app-containers/skopeo-1.5.1/CONTENTS"),
+				file.NewLocation("var/db/pkg/app-containers/skopeo-1.5.1/SIZE"),
+				expectedLicenseLocation,
+			),
+			Licenses: pkg.NewLicenseSet(pkg.NewLicensesFromLocation(expectedLicenseLocation, "Apache-2.0", "BSD", "BSD-2", "CC-BY-SA-4.0", "ISC", "MIT")...),
+			Type:     pkg.PortagePkg,
+			Metadata: pkg.PortageEntry{
+				InstalledSize: 27937835,
+				Files: []pkg.PortageFileRecord{
+					{
+						Path: "/usr/bin/skopeo",
+						Digest: &file.Digest{
+							Algorithm: "md5",
+							Value:     "376c02bd3b22804df8fdfdc895e7dbfb",
+						},
+					},
+					{
+						Path: "/etc/containers/policy.json",
+						Digest: &file.Digest{
+							Algorithm: "md5",
+							Value:     "c01eb6950f03419e09d4fc88cb42ff6f",
+						},
+					},
+					{
+						Path: "/etc/containers/registries.d/default.yaml",
+						Digest: &file.Digest{
+							Algorithm: "md5",
+							Value:     "e6e66cd3c24623e0667f26542e0e08f6",
+						},
+					},
+					{
+						Path: "/var/lib/atomic/sigstore/.keep_app-containers_skopeo-0",
+						Digest: &file.Digest{
+							Algorithm: "md5",
+							Value:     "d41d8cd98f00b204e9800998ecf8427e",
+						},
+					},
+				},
+			},
+		},
+	}
+
+	// TODO: relationships are not under test yet
+	var expectedRelationships []artifact.Relationship
+
+	pkgtest.NewCatalogTester().
+		FromDirectory(t, "test-fixtures/image-portage").
+		Expects(expectedPkgs, expectedRelationships).
+		TestCataloger(t, NewPortageCataloger())
+
+}
+
+func TestCataloger_Globs(t *testing.T) {
+	tests := []struct {
+		name     string
+		fixture  string
+		expected []string
+	}{
+		{
+			name:    "obtain portage contents file",
+			fixture: "test-fixtures/glob-paths",
+			expected: []string{
+				"var/db/pkg/x/y/CONTENTS",
+			},
+		},
+	}
+
+	for _, test := range tests {
+		t.Run(test.name, func(t *testing.T) {
+			pkgtest.NewCatalogTester().
+				FromDirectory(t, test.fixture).
+				ExpectsResolverContentQueries(test.expected).
+				TestCataloger(t, NewPortageCataloger())
+		})
+	}
+}

syft/pkg/cataloger/gentoo/parse_portage_contents.go

@@ -0,0 +1,155 @@
+package gentoo
+
+import (
+	"bufio"
+	"fmt"
+	"path"
+	"path/filepath"
+	"regexp"
+	"strconv"
+	"strings"
+
+	"github.com/scylladb/go-set/strset"
+
+	"github.com/anchore/syft/internal/log"
+	"github.com/anchore/syft/syft/artifact"
+	"github.com/anchore/syft/syft/file"
+	"github.com/anchore/syft/syft/pkg"
+	"github.com/anchore/syft/syft/pkg/cataloger/generic"
+)
+
+var (
+	cpvRe                = regexp.MustCompile(`/([^/]*/[\w+][\w+-]*)-((\d+)((\.\d+)*)([a-z]?)((_(pre|p|beta|alpha|rc)\d*)*)(-r\d+)?)/CONTENTS$`)
+	_     generic.Parser = parsePortageContents
+)
+
+// parses individual CONTENTS files from the portage flat-file store (e.g. /var/db/pkg/*/*/CONTENTS).
+func parsePortageContents(resolver file.Resolver, _ *generic.Environment, reader file.LocationReadCloser) ([]pkg.Package, []artifact.Relationship, error) {
+	cpvMatch := cpvRe.FindStringSubmatch(reader.Location.RealPath)
+	if cpvMatch == nil {
+		return nil, nil, fmt.Errorf("failed to match package and version in %s", reader.Location.RealPath)
+	}
+
+	name, version := cpvMatch[1], cpvMatch[2]
+	if name == "" || version == "" {
+		log.WithFields("path", reader.Location.RealPath).Warnf("failed to parse portage name and version")
+		return nil, nil, nil
+	}
+
+	p := pkg.Package{
+		Name:    name,
+		Version: version,
+		PURL:    packageURL(name, version),
+		Locations: file.NewLocationSet(
+			reader.Location.WithAnnotation(pkg.EvidenceAnnotationKey, pkg.PrimaryEvidenceAnnotation),
+		),
+		Type: pkg.PortagePkg,
+		Metadata: pkg.PortageEntry{
+			// ensure the default value for a collection is never nil since this may be shown as JSON
+			Files: make([]pkg.PortageFileRecord, 0),
+		},
+	}
+	addLicenses(resolver, reader.Location, &p)
+	addSize(resolver, reader.Location, &p)
+	addFiles(resolver, reader.Location, &p)
+
+	p.SetID()
+
+	return []pkg.Package{p}, nil, nil
+}
+
+func addFiles(resolver file.Resolver, dbLocation file.Location, p *pkg.Package) {
+	contentsReader, err := resolver.FileContentsByLocation(dbLocation)
+	if err != nil {
+		log.WithFields("path", dbLocation.RealPath).Warnf("failed to fetch portage contents (package=%s): %+v", p.Name, err)
+		return
+	}
+
+	entry, ok := p.Metadata.(pkg.PortageEntry)
+	if !ok {
+		return
+	}
+
+	scanner := bufio.NewScanner(contentsReader)
+	for scanner.Scan() {
+		line := strings.Trim(scanner.Text(), "\n")
+		fields := strings.Split(line, " ")
+
+		if fields[0] == "obj" {
+			record := pkg.PortageFileRecord{
+				Path: fields[1],
+			}
+			record.Digest = &file.Digest{
+				Algorithm: "md5",
+				Value:     fields[2],
+			}
+			entry.Files = append(entry.Files, record)
+		}
+	}
+
+	p.Metadata = entry
+	p.Locations.Add(dbLocation)
+}
+
+func addLicenses(resolver file.Resolver, dbLocation file.Location, p *pkg.Package) {
+	parentPath := filepath.Dir(dbLocation.RealPath)
+
+	location := resolver.RelativeFileByPath(dbLocation, path.Join(parentPath, "LICENSE"))
+
+	if location == nil {
+		return
+	}
+
+	licenseReader, err := resolver.FileContentsByLocation(*location)
+	if err != nil {
+		log.WithFields("path", dbLocation.RealPath).Warnf("failed to fetch portage LICENSE: %+v", err)
+		return
+	}
+
+	findings := strset.New()
+	scanner := bufio.NewScanner(licenseReader)
+	scanner.Split(bufio.ScanWords)
+	for scanner.Scan() {
+		token := scanner.Text()
+		if token != "||" && token != "(" && token != ")" {
+			findings.Add(token)
+		}
+	}
+
+	licenseCandidates := findings.List()
+	p.Licenses = pkg.NewLicenseSet(pkg.NewLicensesFromLocation(*location, licenseCandidates...)...)
+	p.Locations.Add(location.WithAnnotation(pkg.EvidenceAnnotationKey, pkg.SupportingEvidenceAnnotation))
+}
+
+func addSize(resolver file.Resolver, dbLocation file.Location, p *pkg.Package) {
+	parentPath := filepath.Dir(dbLocation.RealPath)
+
+	location := resolver.RelativeFileByPath(dbLocation, path.Join(parentPath, "SIZE"))
+
+	if location == nil {
+		return
+	}
+
+	entry, ok := p.Metadata.(pkg.PortageEntry)
+	if !ok {
+		return
+	}
+
+	sizeReader, err := resolver.FileContentsByLocation(*location)
+	if err != nil {
+		log.WithFields("name", p.Name).Warnf("failed to fetch portage SIZE: %+v", err)
+		return
+	}
+
+	scanner := bufio.NewScanner(sizeReader)
+	for scanner.Scan() {
+		line := strings.Trim(scanner.Text(), "\n")
+		size, err := strconv.Atoi(line)
+		if err == nil {
+			entry.InstalledSize = size
+		}
+	}
+
+	p.Metadata = entry
+	p.Locations.Add(location.WithAnnotation(pkg.EvidenceAnnotationKey, pkg.SupportingEvidenceAnnotation))
+}

syft/pkg/cataloger/gentoo/purl.go

@@ -0,0 +1,18 @@
+package gentoo
+
+import (
+	"github.com/anchore/packageurl-go"
+)
+
+func packageURL(name, version string) string {
+	var qualifiers packageurl.Qualifiers
+
+	return packageurl.NewPackageURL(
+		"ebuild", // currently this is the proposed type for portage packages at https://github.com/package-url/purl-spec/blob/master/PURL-TYPES.rst
+		"",
+		name,
+		version,
+		qualifiers,
+		"",
+	).ToString()
+}

syft/pkg/cataloger/gentoo/purl_test.go

@@ -0,0 +1,28 @@
+package gentoo
+
+import (
+	"fmt"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+)
+
+func Test_packageURL(t *testing.T) {
+
+	tests := []struct {
+		name    string
+		version string
+		want    string
+	}{
+		{
+			"app-admin/eselect",
+			"1.4.15",
+			"pkg:ebuild/app-admin/eselect@1.4.15",
+		},
+	}
+	for _, tt := range tests {
+		t.Run(fmt.Sprintf("%s@%s", tt.name, tt.version), func(t *testing.T) {
+			assert.Equal(t, tt.want, packageURL(tt.name, tt.version))
+		})
+	}
+}

syft/pkg/cataloger/gentoo/test-fixtures/glob-paths/var/db/pkg/x/y/CONTENTS

@@ -0,0 +1 @@
+bogus contents

syft/pkg/cataloger/gentoo/test-fixtures/image-portage/var/db/pkg/app-containers/skopeo-1.5.1/CONTENTS

@@ -0,0 +1,13 @@
+dir /usr
+dir /usr/bin
+obj /usr/bin/skopeo 376c02bd3b22804df8fdfdc895e7dbfb 1649284374
+dir /etc
+dir /etc/containers
+obj /etc/containers/policy.json c01eb6950f03419e09d4fc88cb42ff6f 1649284375
+dir /etc/containers/registries.d
+obj /etc/containers/registries.d/default.yaml e6e66cd3c24623e0667f26542e0e08f6 1649284375
+dir /var
+dir /var/lib
+dir /var/lib/atomic
+dir /var/lib/atomic/sigstore
+obj /var/lib/atomic/sigstore/.keep_app-containers_skopeo-0 d41d8cd98f00b204e9800998ecf8427e 1649284375

syft/pkg/cataloger/gentoo/test-fixtures/image-portage/var/db/pkg/app-containers/skopeo-1.5.1/LICENSE

@@ -0,0 +1 @@
+Apache-2.0 BSD BSD-2 CC-BY-SA-4.0 ISC MIT

syft/pkg/cataloger/gentoo/test-fixtures/image-portage/var/db/pkg/app-containers/skopeo-1.5.1/SIZE

@@ -0,0 +1 @@
+27937835