Extraction Hardening

README.md

@@ -49,7 +49,7 @@ The shared CLI exposes these project utilities:
 sow-toolkit build
 sow-toolkit build-module
 sow-toolkit build-haks [--hak <name>] [--archive <name>] [--source-manifest <path>] [--plan-only] [--skip-music] [--music-dataset <id>]
-sow-toolkit extract [<module-or-hak-file> ...]
+sow-toolkit extract [<build-relative-archive-or-glob> ...]
 sow-toolkit validate
 sow-toolkit compare
 sow-toolkit apply-hak-manifest [<manifest-path>]
@@ -185,10 +185,12 @@ scripts:
 
 extract:
   layout: nwn_canonical_json
-  hak_discovery: build_glob      # or configured_haks
+  archives:
+    - "{module.resref}.mod"
   cleanup_stale: true
+  consume_archives: false
   ignore_extensions: []
-  delete_module_archive_after_success: false
+  delete_module_archive_after_success: false # deprecated compatibility alias
 
 topdata:
   build: "{paths.build}/topdata"
@@ -222,11 +224,29 @@ cache paths must stay relative to the repository root unless a specific setting
 documents otherwise.
 
 Extraction is deliberately guarded because it writes source files from binary
-archives. Module resources require a configured `paths.source`; HAK assets
+archives. `extract.archives` is a list of glob patterns relative to
-require a configured `paths.assets`; both roots must resolve to safe subtrees.
+`paths.build`; the default is only the configured module `.mod`. HAK extraction
-If a root is unset or resolves to the repository root, extraction fails before
+is opt-in, for example:
-writing instead of creating extension directories like `mdl/` or `png/` at the
+
-repo top level. Stale cleanup uses the same safe-root rule.
+```yaml
+extract:
+  archives:
+    - "{module.resref}.mod"
+    - "sow_over*.hak"
+```
+
+Module resources require a configured `paths.source`; HAK assets require a
+configured `paths.assets`; both roots must resolve to safe subtrees. If a root
+is unset or resolves to the repository root, extraction fails before writing
+instead of creating extension directories like `mdl/` or `png/` at the repo top
+level. `.erf` files are rejected by extraction because they do not contain
+`module.ifo`, so they cannot safely update module-level area membership.
+
+`consume_archives: true` deletes the selected build archive files only after the
+entire extraction succeeds. The older `delete_module_archive_after_success`
+setting remains as a module-only compatibility alias when `consume_archives` is
+not set. Stale cleanup is scoped to roots actually touched by extracted
+resources, so the default module-only extraction does not prune assets.
 
 Music defaults (used when not explicitly configured):
 

internal/app/app.go

@@ -1006,8 +1006,12 @@ func configValidationRules(key string) []string {
 		return []string{"generated output names must use the configured extension and must not escape the repository root"}
 	case key == "extract.layout":
 		return []string{"supported value: nwn_canonical_json"}
+	case key == "extract.archives":
+		return []string{"build-relative .mod/.hak glob patterns; .erf is intentionally rejected"}
+	case key == "extract.consume_archives":
+		return []string{"when true, selected build archives are deleted only after successful extraction"}
 	case key == "extract.hak_discovery":
-		return []string{"supported values: build_glob, configured_haks"}
+		return []string{"deprecated; use extract.archives instead"}
 	case key == "autogen.cache.max_age":
 		return []string{"Go duration string, for example 1h or 30m"}
 	case strings.HasPrefix(key, "topdata.package_hak"):

internal/pipeline/extract.go

@@ -25,71 +25,60 @@ type ExtractResult struct {
 	Skipped             int
 }
 
+type extractionArchive struct {
+	Path      string
+	Rel       string
+	Extension string
+}
+
+type extractionScope struct {
+	Source bool
+	Assets bool
+}
+
 func Extract(p *project.Project, files ...string) (ExtractResult, error) {
 	var result ExtractResult
 	var failures []error
 	desired := map[string]struct{}{}
+	scope := extractionScope{}
 
-	allowed := make(map[string]bool)
+	archives, err := resolveExtractionArchives(p, files)
-	for _, f := range files {
-		allowed[f] = true
-	}
-
-	shouldExtractMod := len(allowed) == 0 || allowed[p.Config.Module.ResRef+".mod"]
-	if shouldExtractMod {
-		modulePath := p.ModuleArchivePath()
-		input, err := os.Open(modulePath)
 	if err != nil {
-			return ExtractResult{}, fmt.Errorf("open module archive: %w", err)
+		return result, err
+	}
+	for _, archivePath := range archives {
+		input, err := os.Open(archivePath.Path)
+		if err != nil {
+			failures = append(failures, fmt.Errorf("open archive %s: %w", archivePath.Path, err))
+			continue
 		}
-		defer input.Close()
-
 		archive, err := erf.Read(input)
-		if err != nil {
-			return ExtractResult{}, fmt.Errorf("read module archive: %w", err)
-		}
-
-		result.ModulePath = modulePath
-		written, overwritten, skipped, errs := extractArchiveResources(p, archive, desired)
-		result.Written += written
-		result.Overwritten += overwritten
-		result.Skipped += skipped
-		failures = append(failures, errs...)
-	}
-
-	hakPaths, err := extractHAKPaths(p)
-	if err != nil {
-		return result, fmt.Errorf("scan hak archives: %w", err)
-	}
-	slices.Sort(hakPaths)
-	for _, hakPath := range hakPaths {
-		filename := filepath.Base(hakPath)
-		if len(allowed) > 0 && !allowed[filename] {
-			continue
-		}
-
-		input, err := os.Open(hakPath)
-		if err != nil {
-			failures = append(failures, fmt.Errorf("open hak archive %s: %w", hakPath, err))
-			continue
-		}
-		hakArchive, err := erf.Read(input)
 		input.Close()
 		if err != nil {
-			failures = append(failures, fmt.Errorf("read hak archive %s: %w", hakPath, err))
+			failures = append(failures, fmt.Errorf("read archive %s: %w", archivePath.Path, err))
 			continue
 		}
 
-		result.HAKPaths = append(result.HAKPaths, hakPath)
+		switch archivePath.Extension {
-		written, overwritten, skipped, errs := extractArchiveResources(p, hakArchive, desired)
+		case ".mod":
+			if result.ModulePath == "" {
+				result.ModulePath = archivePath.Path
+			}
+		case ".hak":
+			result.HAKPaths = append(result.HAKPaths, archivePath.Path)
+		}
+
+		written, overwritten, skipped, extractedScope, errs := extractArchiveResources(p, archive, desired)
 		result.Written += written
 		result.Overwritten += overwritten
 		result.Skipped += skipped
+		scope.Source = scope.Source || extractedScope.Source
+		scope.Assets = scope.Assets || extractedScope.Assets
 		failures = append(failures, errs...)
 	}
 
 	if p.EffectiveConfig().Extract.CleanupStale == nil || *p.EffectiveConfig().Extract.CleanupStale {
-		removed, errs := cleanupStaleFiles(p, desired)
+		removed, errs := cleanupStaleFiles(p, desired, scope)
 		result.Removed = removed
 		failures = append(failures, errs...)
 	}
@@ -97,20 +86,27 @@ func Extract(p *project.Project, files ...string) (ExtractResult, error) {
 	if len(failures) > 0 {
 		return result, errors.Join(failures...)
 	}
-	if p.Config.Extract.DeleteModuleArchiveAfterSuccess && result.ModulePath != "" {
+	consumeAll, consumeModules := extractionConsumePolicy(p)
-		if err := os.Remove(result.ModulePath); err != nil {
+	if consumeAll || consumeModules {
-			return result, fmt.Errorf("delete consumed module archive %s: %w", result.ModulePath, err)
+		for _, archivePath := range archives {
+			if !consumeAll && archivePath.Extension != ".mod" {
+				continue
+			}
+			if err := os.Remove(archivePath.Path); err != nil {
+				return result, fmt.Errorf("delete consumed archive %s: %w", archivePath.Path, err)
+			}
+			result.DeletedArchivePaths = append(result.DeletedArchivePaths, archivePath.Path)
 		}
-		result.DeletedArchivePaths = append(result.DeletedArchivePaths, result.ModulePath)
 	}
 	return result, nil
 }
 
-func extractArchiveResources(p *project.Project, archive erf.Archive, desired map[string]struct{}) (int, int, int, []error) {
+func extractArchiveResources(p *project.Project, archive erf.Archive, desired map[string]struct{}) (int, int, int, extractionScope, []error) {
 	var failures []error
 	writtenCount := 0
 	overwrittenCount := 0
 	skippedCount := 0
+	scope := extractionScope{}
 
 	ignored := make(map[string]bool)
 	for _, ext := range p.Config.Extract.IgnoreExtensions {
@@ -135,6 +131,8 @@ func extractArchiveResources(p *project.Project, archive erf.Archive, desired ma
 			continue
 		}
 		desired[target] = struct{}{}
+		scope.Source = scope.Source || pathWithinRoot(target, p.SourceDir())
+		scope.Assets = scope.Assets || pathWithinRoot(target, p.AssetsDir())
 
 		state, err := writeManagedFile(target, data)
 		if err != nil {
@@ -151,7 +149,138 @@ func extractArchiveResources(p *project.Project, archive erf.Archive, desired ma
 		}
 	}
 
-	return writtenCount, overwrittenCount, skippedCount, failures
+	return writtenCount, overwrittenCount, skippedCount, scope, failures
+}
+
+func resolveExtractionArchives(p *project.Project, overrides []string) ([]extractionArchive, error) {
+	patterns := p.EffectiveConfig().Extract.Archives
+	if len(overrides) > 0 {
+		patterns = overrides
+	}
+	if len(patterns) == 0 {
+		return nil, fmt.Errorf("extract.archives must contain at least one build-relative archive pattern")
+	}
+
+	buildRoot := filepath.Clean(p.BuildDir())
+	byPath := map[string]extractionArchive{}
+	for _, rawPattern := range patterns {
+		pattern, err := normalizeExtractionArchivePattern(p, rawPattern)
+		if err != nil {
+			return nil, err
+		}
+		matched, err := matchExtractionArchives(buildRoot, pattern)
+		if err != nil {
+			return nil, err
+		}
+		if len(matched) == 0 {
+			return nil, fmt.Errorf("extract archive pattern %q matched no files under %s", rawPattern, buildRoot)
+		}
+		for _, archive := range matched {
+			byPath[archive.Path] = archive
+		}
+	}
+
+	archives := make([]extractionArchive, 0, len(byPath))
+	for _, archive := range byPath {
+		archives = append(archives, archive)
+	}
+	slices.SortFunc(archives, func(a, b extractionArchive) int {
+		return strings.Compare(a.Rel, b.Rel)
+	})
+	return archives, nil
+}
+
+func normalizeExtractionArchivePattern(p *project.Project, raw string) (string, error) {
+	pattern := strings.TrimSpace(raw)
+	if pattern == "" {
+		return "", fmt.Errorf("extract archive pattern must not be empty")
+	}
+	pattern = strings.NewReplacer(
+		"{module.resref}", strings.TrimSpace(p.Config.Module.ResRef),
+	).Replace(pattern)
+	if strings.Contains(pattern, "\x00") {
+		return "", fmt.Errorf("extract archive pattern %q must not contain NUL bytes", raw)
+	}
+	if filepath.IsAbs(pattern) {
+		return "", fmt.Errorf("extract archive pattern %q must be relative to paths.build", raw)
+	}
+	clean := filepath.Clean(filepath.FromSlash(pattern))
+	if clean == "." || clean == ".." || strings.HasPrefix(clean, ".."+string(filepath.Separator)) {
+		return "", fmt.Errorf("extract archive pattern %q must not escape paths.build", raw)
+	}
+	return filepath.ToSlash(clean), nil
+}
+
+func matchExtractionArchives(buildRoot, pattern string) ([]extractionArchive, error) {
+	var archives []extractionArchive
+	err := filepath.WalkDir(buildRoot, func(path string, entry os.DirEntry, err error) error {
+		if err != nil {
+			return err
+		}
+		if entry.IsDir() {
+			return nil
+		}
+		rel, err := filepath.Rel(buildRoot, path)
+		if err != nil {
+			return err
+		}
+		rel = filepath.ToSlash(rel)
+		if !matchPathPattern(rel, pattern) {
+			return nil
+		}
+		archive, err := validateExtractionArchive(buildRoot, rel)
+		if err != nil {
+			return err
+		}
+		archives = append(archives, archive)
+		return nil
+	})
+	if err != nil {
+		if errors.Is(err, os.ErrNotExist) {
+			return nil, fmt.Errorf("scan build directory %s: %w", buildRoot, err)
+		}
+		return nil, err
+	}
+	return archives, nil
+}
+
+func validateExtractionArchive(buildRoot, rel string) (extractionArchive, error) {
+	cleanRel := filepath.Clean(filepath.FromSlash(rel))
+	if cleanRel == "." || cleanRel == ".." || strings.HasPrefix(cleanRel, ".."+string(filepath.Separator)) {
+		return extractionArchive{}, fmt.Errorf("extract archive %q must not escape paths.build", rel)
+	}
+	path := filepath.Join(buildRoot, cleanRel)
+	if !pathWithinRoot(path, buildRoot) {
+		return extractionArchive{}, fmt.Errorf("extract archive %q resolves outside paths.build", rel)
+	}
+	extension := strings.ToLower(filepath.Ext(cleanRel))
+	switch extension {
+	case ".mod", ".hak":
+		return extractionArchive{Path: path, Rel: filepath.ToSlash(cleanRel), Extension: extension}, nil
+	case ".erf":
+		return extractionArchive{}, fmt.Errorf("extract archive %q is an .erf; ERF extraction is unsafe because ERFs lack module.ifo context", rel)
+	default:
+		return extractionArchive{}, fmt.Errorf("extract archive %q uses unsupported extension %q", rel, extension)
+	}
+}
+
+func extractionConsumePolicy(p *project.Project) (consumeAll bool, consumeModules bool) {
+	if configured := p.Config.Extract.ConsumeArchives; configured != nil {
+		return *configured, false
+	}
+	return false, p.Config.Extract.DeleteModuleArchiveAfterSuccess
+}
+
+func pathWithinRoot(path, root string) bool {
+	root = filepath.Clean(root)
+	if root == "" || root == "." {
+		return false
+	}
+	rel, err := filepath.Rel(root, path)
+	if err != nil {
+		return false
+	}
+	return rel != ".." && !strings.HasPrefix(rel, ".."+string(filepath.Separator)) && !filepath.IsAbs(rel)
 }
 
 func extractedFile(p *project.Project, resource erf.Resource, extension string) (string, []byte, error) {
@@ -211,30 +340,6 @@ func extractionTarget(p *project.Project, field, configured, root string, parts
 	return target, nil
 }
 
-func extractHAKPaths(p *project.Project) ([]string, error) {
-	switch p.EffectiveConfig().Extract.HAKDiscovery {
-	case "configured_haks":
-		paths := make([]string, 0, len(p.Config.HAKs))
-		for _, hak := range p.Config.HAKs {
-			path := p.HAKArchivePath(hak.Name)
-			if _, err := os.Stat(path); err == nil {
-				paths = append(paths, path)
-			} else if err != nil && !errors.Is(err, os.ErrNotExist) {
-				return nil, err
-			}
-		}
-		slices.Sort(paths)
-		return paths, nil
-	default:
-		hakPaths, err := filepath.Glob(filepath.Join(p.BuildDir(), "*.hak"))
-		if err != nil {
-			return nil, err
-		}
-		slices.Sort(hakPaths)
-		return hakPaths, nil
-	}
-}
-
 type writeState int
 
 const (
@@ -268,9 +373,9 @@ func writeManagedFile(path string, data []byte) (writeState, error) {
 	return writeNew, nil
 }
 
-func cleanupStaleFiles(p *project.Project, desired map[string]struct{}) (int, []error) {
+func cleanupStaleFiles(p *project.Project, desired map[string]struct{}, scope extractionScope) (int, []error) {
 	candidates := make([]string, 0, len(p.Inventory.SourceFiles)+len(p.Inventory.ScriptFiles)+len(p.Inventory.AssetFiles))
-	if safeCleanupRoot(p.SourceDir(), p.Root) {
+	if scope.Source && safeCleanupRoot(p.SourceDir(), p.Root) {
 		for _, rel := range p.Inventory.SourceFiles {
 			candidates = append(candidates, filepath.Join(p.SourceDir(), filepath.FromSlash(rel)))
 		}
@@ -278,7 +383,7 @@ func cleanupStaleFiles(p *project.Project, desired map[string]struct{}) (int, []
 			candidates = append(candidates, filepath.Join(p.SourceDir(), filepath.FromSlash(rel)))
 		}
 	}
-	if safeCleanupRoot(p.AssetsDir(), p.Root) {
+	if scope.Assets && safeCleanupRoot(p.AssetsDir(), p.Root) {
 		for _, rel := range p.Inventory.AssetFiles {
 			candidates = append(candidates, filepath.Join(p.AssetsDir(), filepath.FromSlash(rel)))
 		}

internal/pipeline/pipeline_test.go

@@ -120,6 +120,9 @@ func TestExtractReadsHAKAssets(t *testing.T) {
     "source": "src",
     "assets": "assets",
     "build": "build"
+  },
+  "extract": {
+    "archives": ["*.hak"]
   }
 }
 `)
@@ -175,6 +178,281 @@ func TestExtractReadsHAKAssets(t *testing.T) {
 	}
 }
 
+func TestExtractDefaultOnlyReadsConfiguredModuleArchive(t *testing.T) {
+	root := t.TempDir()
+	mustMkdir(t, filepath.Join(root, "src", "module"))
+	mustMkdir(t, filepath.Join(root, "assets", "mdl"))
+	mustMkdir(t, filepath.Join(root, "build"))
+
+	mustWriteFile(t, filepath.Join(root, "nwn-tool.yaml"), `
+module:
+  name: Test Module
+  resref: testmod
+paths:
+  source: src
+  assets: assets
+  build: build
+`)
+	mustWriteFile(t, filepath.Join(root, "src", "module", "module.ifo.json"), `{
+  "file_type": "IFO ",
+  "file_version": "V3.2",
+  "root": {
+    "struct_type": 0,
+    "fields": [
+      {
+        "label": "Mod_Name",
+        "type": "CExoString",
+        "value": "Original Module"
+      }
+    ]
+  }
+}
+`)
+	mustWriteFile(t, filepath.Join(root, "assets", "mdl", "stale_asset.mdl"), "keep")
+
+	p, err := project.Load(root)
+	if err != nil {
+		t.Fatalf("load project: %v", err)
+	}
+	if err := p.ValidateLayout(); err != nil {
+		t.Fatalf("validate layout: %v", err)
+	}
+	if err := p.Scan(); err != nil {
+		t.Fatalf("scan: %v", err)
+	}
+	if _, err := BuildModule(p); err != nil {
+		t.Fatalf("build module: %v", err)
+	}
+	if err := os.Remove(filepath.Join(root, "src", "module", "module.ifo.json")); err != nil {
+		t.Fatalf("remove source file: %v", err)
+	}
+	if err := p.Scan(); err != nil {
+		t.Fatalf("rescan before extract: %v", err)
+	}
+
+	hakFile, err := os.Create(filepath.Join(root, "build", "vfx.hak"))
+	if err != nil {
+		t.Fatalf("create hak: %v", err)
+	}
+	if err := erf.Write(hakFile, erf.New("HAK ", []erf.Resource{
+		{Name: "test_vfx", Type: 0x07D2, Data: []byte("mdl-data")},
+	})); err != nil {
+		t.Fatalf("write hak: %v", err)
+	}
+	if err := hakFile.Close(); err != nil {
+		t.Fatalf("close hak: %v", err)
+	}
+
+	result, err := Extract(p)
+	if err != nil {
+		t.Fatalf("extract: %v", err)
+	}
+	if len(result.HAKPaths) != 0 {
+		t.Fatalf("expected default extract to ignore haks, got %#v", result.HAKPaths)
+	}
+	if _, err := os.Stat(filepath.Join(root, "src", "module", "module.ifo.json")); err != nil {
+		t.Fatalf("expected module source to be restored: %v", err)
+	}
+	if _, err := os.Stat(filepath.Join(root, "assets", "mdl", "stale_asset.mdl")); err != nil {
+		t.Fatalf("expected asset cleanup to be skipped for module-only extract: %v", err)
+	}
+	if _, err := os.Stat(filepath.Join(root, "assets", "mdl", "test_vfx.mdl")); !os.IsNotExist(err) {
+		t.Fatalf("expected hak asset not to be extracted by default, err=%v", err)
+	}
+}
+
+func TestExtractRejectsERFArchiveTargets(t *testing.T) {
+	root := t.TempDir()
+	mustMkdir(t, filepath.Join(root, "src"))
+	mustMkdir(t, filepath.Join(root, "assets"))
+	mustMkdir(t, filepath.Join(root, "build"))
+
+	mustWriteFile(t, filepath.Join(root, "nwn-tool.yaml"), `
+module:
+  name: Test Module
+  resref: testmod
+paths:
+  source: src
+  assets: assets
+  build: build
+extract:
+  archives:
+    - imported.erf
+`)
+
+	erfFile, err := os.Create(filepath.Join(root, "build", "imported.erf"))
+	if err != nil {
+		t.Fatalf("create erf: %v", err)
+	}
+	if err := erf.Write(erfFile, erf.New("ERF ", nil)); err != nil {
+		t.Fatalf("write erf: %v", err)
+	}
+	if err := erfFile.Close(); err != nil {
+		t.Fatalf("close erf: %v", err)
+	}
+
+	p, err := project.Load(root)
+	if err != nil {
+		t.Fatalf("load project: %v", err)
+	}
+	_, err = Extract(p)
+	if err == nil {
+		t.Fatal("expected erf extraction to fail")
+	}
+	if !strings.Contains(err.Error(), "ERF extraction is unsafe") {
+		t.Fatalf("expected unsafe erf error, got %v", err)
+	}
+}
+
+func TestExtractRejectsEscapingArchiveOverride(t *testing.T) {
+	root := t.TempDir()
+	mustMkdir(t, filepath.Join(root, "src"))
+	mustMkdir(t, filepath.Join(root, "assets"))
+	mustMkdir(t, filepath.Join(root, "build"))
+
+	mustWriteFile(t, filepath.Join(root, "nwn-tool.yaml"), `
+module:
+  name: Test Module
+  resref: testmod
+paths:
+  source: src
+  assets: assets
+  build: build
+`)
+
+	p, err := project.Load(root)
+	if err != nil {
+		t.Fatalf("load project: %v", err)
+	}
+	_, err = Extract(p, "../outside.mod")
+	if err == nil {
+		t.Fatal("expected escaping archive override to fail")
+	}
+	if !strings.Contains(err.Error(), "must not escape paths.build") {
+		t.Fatalf("expected escaping archive error, got %v", err)
+	}
+}
+
+func TestExtractConsumesConfiguredHAKArchiveAfterSuccess(t *testing.T) {
+	root := t.TempDir()
+	mustMkdir(t, filepath.Join(root, "src"))
+	mustMkdir(t, filepath.Join(root, "assets"))
+	mustMkdir(t, filepath.Join(root, "build"))
+
+	mustWriteFile(t, filepath.Join(root, "nwn-tool.yaml"), `
+module:
+  name: Test Module
+  resref: testmod
+paths:
+  source: src
+  assets: assets
+  build: build
+extract:
+  archives:
+    - vfx.hak
+  consume_archives: true
+`)
+
+	hakPath := filepath.Join(root, "build", "vfx.hak")
+	hakFile, err := os.Create(hakPath)
+	if err != nil {
+		t.Fatalf("create hak: %v", err)
+	}
+	if err := erf.Write(hakFile, erf.New("HAK ", []erf.Resource{
+		{Name: "test_vfx", Type: 0x07D2, Data: []byte("mdl-data")},
+	})); err != nil {
+		t.Fatalf("write hak: %v", err)
+	}
+	if err := hakFile.Close(); err != nil {
+		t.Fatalf("close hak: %v", err)
+	}
+
+	p, err := project.Load(root)
+	if err != nil {
+		t.Fatalf("load project: %v", err)
+	}
+	result, err := Extract(p)
+	if err != nil {
+		t.Fatalf("extract: %v", err)
+	}
+	if len(result.DeletedArchivePaths) != 1 || result.DeletedArchivePaths[0] != hakPath {
+		t.Fatalf("expected consumed hak path, got %#v", result.DeletedArchivePaths)
+	}
+	if _, err := os.Stat(hakPath); !os.IsNotExist(err) {
+		t.Fatalf("expected hak archive to be deleted, stat err=%v", err)
+	}
+	if _, err := os.Stat(filepath.Join(root, "assets", "mdl", "test_vfx.mdl")); err != nil {
+		t.Fatalf("expected extracted hak asset: %v", err)
+	}
+}
+
+func TestExtractConsumeArchivesFalseOverridesLegacyDeleteFlag(t *testing.T) {
+	root := t.TempDir()
+	mustMkdir(t, filepath.Join(root, "src", "module"))
+	mustMkdir(t, filepath.Join(root, "assets"))
+	mustMkdir(t, filepath.Join(root, "build"))
+
+	mustWriteFile(t, filepath.Join(root, "nwn-tool.yaml"), `
+module:
+  name: Test Module
+  resref: testmod
+paths:
+  source: src
+  assets: assets
+  build: build
+extract:
+  consume_archives: false
+  delete_module_archive_after_success: true
+`)
+	mustWriteFile(t, filepath.Join(root, "src", "module", "module.ifo.json"), `{
+  "file_type": "IFO ",
+  "file_version": "V3.2",
+  "root": {
+    "struct_type": 0,
+    "fields": [
+      {
+        "label": "Mod_Name",
+        "type": "CExoString",
+        "value": "Original Module"
+      }
+    ]
+  }
+}
+`)
+
+	p, err := project.Load(root)
+	if err != nil {
+		t.Fatalf("load project: %v", err)
+	}
+	if err := p.ValidateLayout(); err != nil {
+		t.Fatalf("validate layout: %v", err)
+	}
+	if err := p.Scan(); err != nil {
+		t.Fatalf("scan: %v", err)
+	}
+	buildResult, err := BuildModule(p)
+	if err != nil {
+		t.Fatalf("build module: %v", err)
+	}
+	if err := os.Remove(filepath.Join(root, "src", "module", "module.ifo.json")); err != nil {
+		t.Fatalf("remove source file: %v", err)
+	}
+	if err := p.Scan(); err != nil {
+		t.Fatalf("rescan before extract: %v", err)
+	}
+
+	result, err := Extract(p)
+	if err != nil {
+		t.Fatalf("extract: %v", err)
+	}
+	if len(result.DeletedArchivePaths) != 0 {
+		t.Fatalf("expected no deleted archives, got %#v", result.DeletedArchivePaths)
+	}
+	if _, err := os.Stat(buildResult.ModulePath); err != nil {
+		t.Fatalf("expected module archive to remain, stat err=%v", err)
+	}
+}
+
 func TestExtractRefusesAssetsWhenAssetsPathIsUnset(t *testing.T) {
 	root := t.TempDir()
 	t.Chdir(root)
@@ -226,7 +504,7 @@ paths:
 		t.Fatalf("close hak: %v", err)
 	}
 
-	_, err = Extract(p)
+	_, err = Extract(p, "vfx.hak")
 	if err == nil {
 		t.Fatal("expected extract to fail")
 	}
@@ -285,7 +563,7 @@ paths:
 		t.Fatalf("close hak: %v", err)
 	}
 
-	_, err = Extract(p)
+	_, err = Extract(p, "vfx.hak")
 	if err == nil {
 		t.Fatal("expected extract to fail")
 	}
@@ -357,7 +635,8 @@ paths:
   assets: assets
   build: build
 extract:
-  hak_discovery: configured_haks
+  archives:
+    - wanted.hak
 haks:
   - name: wanted
     priority: 1

internal/project/effective.go

@@ -231,10 +231,17 @@ func (p *Project) EffectiveConfig() EffectiveConfig {
 	extract := p.Config.Extract
 	extract.Layout = defaultString(extract.Layout, DefaultExtractLayout)
 	extract.HAKDiscovery = defaultString(extract.HAKDiscovery, DefaultExtractHAKDiscovery)
+	if len(extract.Archives) == 0 {
+		extract.Archives = []string{outputs.ModuleArchiveName(p.Config.Module)}
+	}
 	if extract.CleanupStale == nil {
 		defaultCleanup := true
 		extract.CleanupStale = &defaultCleanup
 	}
+	if extract.ConsumeArchives == nil {
+		defaultConsume := false
+		extract.ConsumeArchives = &defaultConsume
+	}
 
 	musicStageRoot := expandPathTemplate(DefaultMusicStageRoot, paths)
 	musicCreditsRoot := expandPathTemplate(DefaultCreditsRoot, paths)
@@ -435,7 +442,9 @@ func markMissingDefaults(provenance ConfigProvenance) {
 		"topdata.wiki.deploy_edit_summary":         DefaultWikiDeployEditSummary,
 		"extract.layout":                           DefaultExtractLayout,
 		"extract.hak_discovery":                    DefaultExtractHAKDiscovery,
+		"extract.archives":                         DefaultModuleArchiveTemplate,
 		"extract.cleanup_stale":                    "true",
+		"extract.consume_archives":                 "false",
 		"autogen.cache.root":                       DefaultAutogenCacheRoot,
 		"autogen.cache.max_age":                    DefaultAutogenCacheMaxAge.String(),
 		"autogen.cache.refresh_env":                DefaultAutogenRefreshEnv,

internal/project/project.go

@@ -214,9 +214,11 @@ type TopDataWikiConfig struct {
 
 type ExtractConfig struct {
 	IgnoreExtensions                []string `json:"ignore_extensions" yaml:"ignore_extensions"`
+	Archives                        []string `json:"archives,omitempty" yaml:"archives,omitempty"`
 	Layout                          string   `json:"layout" yaml:"layout"`
 	HAKDiscovery                    string   `json:"hak_discovery" yaml:"hak_discovery"`
 	CleanupStale                    *bool    `json:"cleanup_stale,omitempty" yaml:"cleanup_stale,omitempty"`
+	ConsumeArchives                 *bool    `json:"consume_archives,omitempty" yaml:"consume_archives,omitempty"`
 	DeleteModuleArchiveAfterSuccess bool     `json:"delete_module_archive_after_success,omitempty" yaml:"delete_module_archive_after_success,omitempty"`
 }
 
@@ -506,6 +508,7 @@ func (p *Project) ValidateLayout() error {
 	default:
 		failures = append(failures, fmt.Errorf("extract.hak_discovery %q is not supported", effective.Extract.HAKDiscovery))
 	}
+	failures = append(failures, validateGlobList("extract.archives", effective.Extract.Archives)...)
 	if _, err := time.ParseDuration(effective.Autogen.Cache.MaxAge); err != nil {
 		failures = append(failures, fmt.Errorf("autogen.cache.max_age must be a duration: %w", err))
 	}
@@ -963,6 +966,7 @@ func normalizeConfig(cfg *Config) {
 		cfg.HAKs[i].Include = normalizeStringSlice(cfg.HAKs[i].Include)
 	}
 	cfg.Extract.IgnoreExtensions = normalizeStringSlice(cfg.Extract.IgnoreExtensions)
+	cfg.Extract.Archives = normalizeStringSlice(cfg.Extract.Archives)
 	for i := range cfg.Autogen.Producers {
 		cfg.Autogen.Producers[i].Include = normalizeStringSlice(cfg.Autogen.Producers[i].Include)
 	}

internal/project/project_test.go

@@ -66,6 +66,12 @@ paths:
 	if got, want := effective.Outputs.HAKManifest, "haks.json"; got != want {
 		t.Fatalf("expected default HAK manifest %q, got %q", want, got)
 	}
+	if got, want := strings.Join(effective.Extract.Archives, ","), "testmod.mod"; got != want {
+		t.Fatalf("expected default extract archives %q, got %q", want, got)
+	}
+	if effective.Extract.ConsumeArchives == nil || *effective.Extract.ConsumeArchives {
+		t.Fatalf("expected default extract consume_archives false, got %#v", effective.Extract.ConsumeArchives)
+	}
 	if got, want := effective.TopData.PackageHAK, "sow_top.hak"; got != want {
 		t.Fatalf("expected default topdata HAK %q, got %q", want, got)
 	}