Add nested archive cataloging (#5260)

Signed-off-by: Kudryavcev Nikolay <kydry.nikolau@gmail.com>
Signed-off-by: Thomas Bechtold <thomasbechtold@jpberlin.de>
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
Signed-off-by: Keith Zantow <kzantow@gmail.com>
Co-authored-by: Kudryavcev Nikolay <kydry.nikolau@gmail.com>
Co-authored-by: Thomas Bechtold <thomasbechtold@jpberlin.de>
Co-authored-by: Alex Goodman <wagoodman@users.noreply.github.com>

internal/capabilities/generate/discover_cataloger_configs.go

@@ -9,6 +9,7 @@ import (
 	"os"
 	"path/filepath"
 	"regexp"
+	"slices"
 	"strings"
 
 	"github.com/anchore/syft/internal/capabilities/internal"
@@ -31,6 +32,10 @@ type ConfigInfo struct {
 
 var appConfigAnnotationPattern = regexp.MustCompile(`^//\s*app-config:\s*(.+)$`)
 
+// embeddedFieldsAnnotationPattern matches "// capabilities-fields: A, B" on an embedded struct, which limits
+// the embedded fields listed to the ones named (for a shared type the cataloger reads only part of).
+var embeddedFieldsAnnotationPattern = regexp.MustCompile(`^//\s*capabilities-fields:\s*(.+)$`)
+
 // DiscoverConfigs walks the cataloger directory and discovers all configuration structs
 // Returns map where key is "packageName.StructName" (e.g., "golang.CatalogerConfig")
 func DiscoverConfigs(repoRoot string) (map[string]ConfigInfo, error) {
@@ -160,6 +165,9 @@ func extractCatalogerConfigFieldsRecursive(structType *ast.StructType, visitedTy
 		if len(field.Names) == 0 {
 			// this is an embedded field - resolve and extract its fields
 			embeddedFields := resolveEmbeddedStructFields(field.Type, visitedTypes)
+			if only := embeddedFieldAllowlist(field.Doc); only != nil {
+				embeddedFields = slices.DeleteFunc(embeddedFields, func(f ConfigField) bool { return !only[f.Name] })
+			}
 			fields = append(fields, embeddedFields...)
 			continue
 		}
@@ -195,6 +203,26 @@ func extractCatalogerConfigFieldsRecursive(structType *ast.StructType, visitedTy
 	return fields
 }
 
+// embeddedFieldAllowlist returns the field names from a "capabilities-fields:" annotation, or nil when there
+// is none (every exported embedded field is listed).
+func embeddedFieldAllowlist(commentGroup *ast.CommentGroup) map[string]bool {
+	if commentGroup == nil {
+		return nil
+	}
+	for _, comment := range commentGroup.List {
+		matches := embeddedFieldsAnnotationPattern.FindStringSubmatch(comment.Text)
+		if len(matches) < 2 {
+			continue
+		}
+		only := make(map[string]bool)
+		for _, name := range strings.Split(matches[1], ",") {
+			only[strings.TrimSpace(name)] = true
+		}
+		return only
+	}
+	return nil
+}
+
 // resolveEmbeddedStructFields resolves an embedded struct type and extracts its fields recursively
 func resolveEmbeddedStructFields(fieldType ast.Expr, visitedTypes map[string]bool) []ConfigField {
 	// extract the type name from the expression

internal/capabilities/generate/discover_cataloger_configs_test.go

@@ -457,3 +457,32 @@ func TestDiscoverAllowedConfigStructs(t *testing.T) {
 		t.Logf("  - %s", configType)
 	}
 }
+
+func TestEmbeddedFieldAllowlist(t *testing.T) {
+	comments := func(lines ...string) *ast.CommentGroup {
+		g := &ast.CommentGroup{}
+		for _, l := range lines {
+			g.List = append(g.List, &ast.Comment{Text: l})
+		}
+		return g
+	}
+
+	tests := []struct {
+		name string
+		doc  *ast.CommentGroup
+		want map[string]bool
+	}{
+		{name: "no doc lists every field", doc: nil, want: nil},
+		{name: "no annotation lists every field", doc: comments("// Deprecated: use something else."), want: nil},
+		{
+			name: "annotation limits to the named fields",
+			doc:  comments("// Deprecated: use something else.", "// capabilities-fields: IncludeIndexedArchives, IncludeUnindexedArchives"),
+			want: map[string]bool{"IncludeIndexedArchives": true, "IncludeUnindexedArchives": true},
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			require.Equal(t, tt.want, embeddedFieldAllowlist(tt.doc))
+		})
+	}
+}