datastore: add package_db to package table

rh-pre-commit.version: 2.3.2
rh-pre-commit.check-secrets: ENABLED

Author: RTann

datastore/postgres/indexpackage.go

@@ -48,10 +48,10 @@ var (
 // particular scanner. See the LayerScanned method for more details.
 func (s *IndexerStore) IndexPackages(ctx context.Context, pkgs []*claircore.Package, layer *claircore.Layer, scnr indexer.VersionedScanner) error {
 	const (
-		insert = ` 
-		INSERT INTO package (name, kind, version, norm_kind, norm_version, module, arch)
-		VALUES ($1, $2, $3, $4, $5::int[], $6, $7)
-		ON CONFLICT (name, kind, version, module, arch) DO NOTHING;
+		insert = `
+		INSERT INTO package (name, kind, version, norm_kind, norm_version, module, arch, package_db)
+		VALUES ($1, $2, $3, $4, $5::int[], $6, $7, $8)
+		ON CONFLICT (name, kind, version, module, arch, package_db) DO NOTHING;
 		`
 
 		insertWith = `
@@ -64,27 +64,27 @@ func (s *IndexerStore) IndexPackages(ctx context.Context, pkgs []*claircore.Pack
 			  AND module = $4
 			  AND arch = $5
 		),
-			 binary_package AS (
-				 SELECT id AS package_id
-				 FROM package
-				 WHERE name = $6
-				   AND kind = $7
-				   AND version = $8
-				   AND module = $9
-				   AND arch = $10
-			 ),
-			 scanner AS (
-				 SELECT id AS scanner_id
-				 FROM scanner
-				 WHERE name = $11
-				   AND version = $12
-				   AND kind = $13
-			 ),
-			 layer AS (
-				 SELECT id AS layer_id
-				 FROM layer
-				 WHERE layer.hash = $14
-			 )
+		binary_package AS (
+			SELECT id AS package_id
+			FROM package
+			WHERE name = $6
+			  AND kind = $7
+			  AND version = $8
+			  AND module = $9
+			  AND arch = $10
+		),
+		scanner AS (
+			SELECT id AS scanner_id
+			FROM scanner
+			WHERE name = $11
+			  AND version = $12
+			  AND kind = $13
+		),
+		layer AS (
+			SELECT id AS layer_id
+			FROM layer
+			WHERE layer.hash = $14
+		)
 		INSERT
 		INTO package_scanartifact (layer_id, package_db, repository_hint, filepath, package_id, source_id, scanner_id)
 		VALUES ((SELECT layer_id FROM layer),
@@ -207,7 +207,7 @@ func queueInsert(ctx context.Context, b *microbatch.Insert, stmt string, pkg *cl
 		vNorm = pkg.NormalizedVersion.V[:]
 	}
 	err := b.Queue(ctx, stmt,
-		pkg.Name, pkg.Kind, pkg.Version, vKind, vNorm, pkg.Module, pkg.Arch,
+		pkg.Name, pkg.Kind, pkg.Version, vKind, vNorm, pkg.Module, pkg.Arch, pkg.PackageDB,
 	)
 	if err != nil {
 		return fmt.Errorf("failed to queue insert for package %q: %w", pkg.Name, err)

datastore/postgres/migrations/indexer/10-package-db.sql

@@ -0,0 +1,7 @@
+-- The package table only save one copy for each unique name+version of a package (and a few more fields).
+-- It is possible an image has several packages with the same name+version (and the other fields)
+-- in different locations. Without this migration, the IndexReport only tracks one of those packages
+-- instead of a package for each unique location.
+ALTER TABLE IF EXISTS package ADD COLUMN package_db text;
+DROP INDEX IF EXISTS package_unique_idx;
+CREATE UNIQUE INDEX IF NOT EXISTS package_unique_idx ON package (name, version, kind, module, arch, package_db);

datastore/postgres/migrations/migrations.go

@@ -65,6 +65,10 @@ var IndexerMigrations = []migrate.Migration{
 		ID: 9,
 		Up: runFile("indexer/09-delete-scannerlist.sql"),
 	},
+	{
+		ID: 10,
+		Up: runFile("indexer/10-package-db.sql"),
+	},
 }
 
 var MatcherMigrations = []migrate.Migration{