Identify task (#1839)

* Add identify task
* Change type naming
* Debounce folder select text input
* Add generic slice comparison function

graphql/documents/data/config.graphql

@@ -81,6 +81,43 @@ fragment ConfigScrapingData on ConfigScrapingResult {
   excludeTagPatterns
 }
 
+fragment IdentifyFieldOptionsData on IdentifyFieldOptions {
+  field
+  strategy
+  createMissing
+}
+
+fragment IdentifyMetadataOptionsData on IdentifyMetadataOptions {
+  fieldOptions {
+    ...IdentifyFieldOptionsData
+  }
+  setCoverImage
+  setOrganized
+  includeMalePerformers
+}
+
+fragment ScraperSourceData on ScraperSource {
+  stash_box_index
+  stash_box_endpoint
+  scraper_id
+}
+
+fragment ConfigDefaultSettingsData on ConfigDefaultSettingsResult {
+  identify {
+    sources {
+      source {
+        ...ScraperSourceData
+      }
+      options {
+        ...IdentifyMetadataOptionsData
+      }
+    }
+    options {
+      ...IdentifyMetadataOptionsData
+    }
+  }
+}
+
 fragment ConfigData on ConfigResult {
   general {
     ...ConfigGeneralData
@@ -94,4 +131,7 @@ fragment ConfigData on ConfigResult {
   scraping {
     ...ConfigScrapingData
   }
+  defaults {
+    ...ConfigDefaultSettingsData
+  }
 }

graphql/documents/mutations/config.graphql

@@ -30,6 +30,12 @@ mutation ConfigureScraping($input: ConfigScrapingInput!) {
   }
 }
 
+mutation ConfigureDefaults($input: ConfigDefaultSettingsInput!) {
+  configureDefaults(input: $input) {
+    ...ConfigDefaultSettingsData
+  }
+}
+
 mutation GenerateAPIKey($input: GenerateAPIKeyInput!) {
   generateAPIKey(input: $input)
 }

graphql/documents/mutations/metadata.graphql

@@ -26,6 +26,10 @@ mutation MetadataAutoTag($input: AutoTagMetadataInput!) {
   metadataAutoTag(input: $input)
 }
 
+mutation MetadataIdentify($input: IdentifyMetadataInput!) {
+  metadataIdentify(input: $input)
+}
+
 mutation MetadataClean($input: CleanMetadataInput!) {
   metadataClean(input: $input)
 }

graphql/schema/schema.graphql

@@ -237,6 +237,7 @@ type Mutation {
   configureInterface(input: ConfigInterfaceInput!): ConfigInterfaceResult!
   configureDLNA(input: ConfigDLNAInput!): ConfigDLNAResult!
   configureScraping(input: ConfigScrapingInput!): ConfigScrapingResult!
+  configureDefaults(input: ConfigDefaultSettingsInput!): ConfigDefaultSettingsResult!
 
   """Generate and set (or clear) API key"""
   generateAPIKey(input: GenerateAPIKeyInput!): String!
@@ -259,6 +260,8 @@ type Mutation {
   metadataAutoTag(input: AutoTagMetadataInput!): ID!
   """Clean metadata. Returns the job ID"""
   metadataClean(input: CleanMetadataInput!): ID!
+  """Identifies scenes using scrapers. Returns the job ID"""
+  metadataIdentify(input: IdentifyMetadataInput!): ID!
   """Migrate generated files for the current hash naming"""
   migrateHashNaming: ID!
 

graphql/schema/types/config.graphql

@@ -302,12 +302,21 @@ type ConfigScrapingResult {
   excludeTagPatterns: [String!]!
 }
 
+type ConfigDefaultSettingsResult {
+  identify: IdentifyMetadataTaskOptions
+}
+
+input ConfigDefaultSettingsInput {
+  identify: IdentifyMetadataInput
+}
+
 """All configuration settings"""
 type ConfigResult {
   general: ConfigGeneralResult!
   interface: ConfigInterfaceResult!
   dlna: ConfigDLNAResult!
   scraping: ConfigScrapingResult!
+  defaults: ConfigDefaultSettingsResult!
 }
 
 """Directory structure of a path"""

graphql/schema/types/metadata.graphql

@@ -67,6 +67,88 @@ input AutoTagMetadataInput {
   tags: [String!]
 }
 
+enum IdentifyFieldStrategy {
+  """Never sets the field value"""
+  IGNORE
+  """
+  For multi-value fields, merge with existing. 
+  For single-value fields, ignore if already set
+  """
+  MERGE
+  """Always replaces the value if a value is found. 
+  For multi-value fields, any existing values are removed and replaced with the
+  scraped values.
+  """
+  OVERWRITE
+}
+
+input IdentifyFieldOptionsInput {
+  field: String!
+  strategy: IdentifyFieldStrategy!
+  """creates missing objects if needed - only applicable for performers, tags and studios"""
+  createMissing: Boolean
+}
+
+input IdentifyMetadataOptionsInput {
+  """any fields missing from here are defaulted to MERGE and createMissing false"""
+  fieldOptions: [IdentifyFieldOptionsInput!]
+  """defaults to true if not provided"""
+  setCoverImage: Boolean
+  setOrganized: Boolean
+  """defaults to true if not provided"""
+  includeMalePerformers: Boolean
+}
+
+input IdentifySourceInput {
+  source: ScraperSourceInput!
+  """Options defined for a source override the defaults"""
+  options: IdentifyMetadataOptionsInput
+}
+
+input IdentifyMetadataInput {
+  """An ordered list of sources to identify items with. Only the first source that finds a match is used."""
+  sources: [IdentifySourceInput!]!
+  """Options defined here override the configured defaults"""
+  options: IdentifyMetadataOptionsInput
+
+  """scene ids to identify"""
+  sceneIDs: [ID!]
+
+  """paths of scenes to identify - ignored if scene ids are set"""
+  paths: [String!]
+}
+
+# types for default options
+type IdentifyFieldOptions {
+  field: String!
+  strategy: IdentifyFieldStrategy!
+  """creates missing objects if needed - only applicable for performers, tags and studios"""
+  createMissing: Boolean
+}
+
+type IdentifyMetadataOptions {
+  """any fields missing from here are defaulted to MERGE and createMissing false"""
+  fieldOptions: [IdentifyFieldOptions!]
+  """defaults to true if not provided"""
+  setCoverImage: Boolean
+  setOrganized: Boolean
+  """defaults to true if not provided"""
+  includeMalePerformers: Boolean
+}
+
+type IdentifySource {
+  source: ScraperSource!
+  """Options defined for a source override the defaults"""
+  options: IdentifyMetadataOptions
+}
+
+type IdentifyMetadataTaskOptions {
+  """An ordered list of sources to identify items with. Only the first source that finds a match is used."""
+  sources: [IdentifySource!]!
+  """Options defined here override the configured defaults"""
+  options: IdentifyMetadataOptions
+}
+
 input ExportObjectTypeInput {
   ids: [String!]
   all: Boolean

graphql/schema/types/scraper.graphql

@@ -96,7 +96,18 @@ input ScrapedGalleryInput {
 
 input ScraperSourceInput {
   """Index of the configured stash-box instance to use. Should be unset if scraper_id is set"""
-  stash_box_index: Int
+  stash_box_index: Int @deprecated(reason: "use stash_box_endpoint")
+  """Stash-box endpoint"""
+  stash_box_endpoint: String
+  """Scraper ID to scrape with. Should be unset if stash_box_index is set"""
+  scraper_id: ID
+}
+
+type ScraperSource {
+  """Index of the configured stash-box instance to use. Should be unset if scraper_id is set"""
+  stash_box_index: Int @deprecated(reason: "use stash_box_endpoint")
+  """Stash-box endpoint"""
+  stash_box_endpoint: String
   """Scraper ID to scrape with. Should be unset if stash_box_index is set"""
   scraper_id: ID
 }

pkg/api/resolver_mutation_configure.go

@@ -352,6 +352,20 @@ func (r *mutationResolver) ConfigureScraping(ctx context.Context, input models.C
 	return makeConfigScrapingResult(), nil
 }
 
+func (r *mutationResolver) ConfigureDefaults(ctx context.Context, input models.ConfigDefaultSettingsInput) (*models.ConfigDefaultSettingsResult, error) {
+	c := config.GetInstance()
+
+	if input.Identify != nil {
+		c.Set(config.DefaultIdentifySettings, input.Identify)
+	}
+
+	if err := c.Write(); err != nil {
+		return makeConfigDefaultsResult(), err
+	}
+
+	return makeConfigDefaultsResult(), nil
+}
+
 func (r *mutationResolver) GenerateAPIKey(ctx context.Context, input models.GenerateAPIKeyInput) (string, error) {
 	c := config.GetInstance()
 

pkg/api/resolver_mutation_metadata.go

@@ -90,6 +90,13 @@ func (r *mutationResolver) MetadataAutoTag(ctx context.Context, input models.Aut
 	return strconv.Itoa(jobID), nil
 }
 
+func (r *mutationResolver) MetadataIdentify(ctx context.Context, input models.IdentifyMetadataInput) (string, error) {
+	t := manager.CreateIdentifyJob(input)
+	jobID := manager.GetInstance().JobManager.Add(ctx, "Identifying...", t)
+
+	return strconv.Itoa(jobID), nil
+}
+
 func (r *mutationResolver) MetadataClean(ctx context.Context, input models.CleanMetadataInput) (string, error) {
 	jobID := manager.GetInstance().Clean(ctx, input)
 	return strconv.Itoa(jobID), nil

pkg/api/resolver_mutation_scene.go

@@ -11,6 +11,7 @@ import (
 	"github.com/stashapp/stash/pkg/manager/config"
 	"github.com/stashapp/stash/pkg/models"
 	"github.com/stashapp/stash/pkg/plugin"
+	"github.com/stashapp/stash/pkg/scene"
 	"github.com/stashapp/stash/pkg/utils"
 )
 
@@ -119,7 +120,7 @@ func (r *mutationResolver) sceneUpdate(ctx context.Context, input models.SceneUp
 	}
 
 	qb := repo.Scene()
-	scene, err := qb.Update(updatedScene)
+	s, err := qb.Update(updatedScene)
 	if err != nil {
 		return nil, err
 	}
@@ -169,13 +170,13 @@ func (r *mutationResolver) sceneUpdate(ctx context.Context, input models.SceneUp
 
 	// only update the cover image if provided and everything else was successful
 	if coverImageData != nil {
-		err = manager.SetSceneScreenshot(scene.GetHash(config.GetInstance().GetVideoFileNamingAlgorithm()), coverImageData)
+		err = scene.SetScreenshot(manager.GetInstance().Paths, s.GetHash(config.GetInstance().GetVideoFileNamingAlgorithm()), coverImageData)
 		if err != nil {
 			return nil, err
 		}
 	}
 
-	return scene, nil
+	return s, nil
 }
 
 func (r *mutationResolver) updateScenePerformers(qb models.SceneReaderWriter, sceneID int, performerIDs []string) error {

pkg/api/resolver_query_configuration.go

@@ -43,6 +43,7 @@ func makeConfigResult() *models.ConfigResult {
 		Interface: makeConfigInterfaceResult(),
 		Dlna:      makeConfigDLNAResult(),
 		Scraping:  makeConfigScrapingResult(),
+		Defaults:  makeConfigDefaultsResult(),
 	}
 }
 
@@ -159,3 +160,11 @@ func makeConfigScrapingResult() *models.ConfigScrapingResult {
 		ExcludeTagPatterns: config.GetScraperExcludeTagPatterns(),
 	}
 }
+
+func makeConfigDefaultsResult() *models.ConfigDefaultSettingsResult {
+	config := config.GetInstance()
+
+	return &models.ConfigDefaultSettingsResult{
+		Identify: config.GetDefaultIdentifySettings(),
+	}
+}

pkg/identify/identify.go

@@ -0,0 +1,264 @@
+package identify
+
+import (
+	"context"
+	"database/sql"
+	"fmt"
+
+	"github.com/stashapp/stash/pkg/logger"
+	"github.com/stashapp/stash/pkg/models"
+	"github.com/stashapp/stash/pkg/scene"
+	"github.com/stashapp/stash/pkg/utils"
+)
+
+type SceneScraper interface {
+	ScrapeScene(sceneID int) (*models.ScrapedScene, error)
+}
+
+type SceneUpdatePostHookExecutor interface {
+	ExecuteSceneUpdatePostHooks(ctx context.Context, input models.SceneUpdateInput, inputFields []string)
+}
+
+type ScraperSource struct {
+	Name       string
+	Options    *models.IdentifyMetadataOptionsInput
+	Scraper    SceneScraper
+	RemoteSite string
+}
+
+type SceneIdentifier struct {
+	DefaultOptions              *models.IdentifyMetadataOptionsInput
+	Sources                     []ScraperSource
+	ScreenshotSetter            scene.ScreenshotSetter
+	SceneUpdatePostHookExecutor SceneUpdatePostHookExecutor
+}
+
+func (t *SceneIdentifier) Identify(ctx context.Context, repo models.Repository, scene *models.Scene) error {
+	result, err := t.scrapeScene(scene)
+	if err != nil {
+		return err
+	}
+
+	if result == nil {
+		logger.Infof("Unable to identify %s", scene.Path)
+		return nil
+	}
+
+	// results were found, modify the scene
+	if err := t.modifyScene(ctx, repo, scene, result); err != nil {
+		return fmt.Errorf("error modifying scene: %v", err)
+	}
+
+	return nil
+}
+
+type scrapeResult struct {
+	result *models.ScrapedScene
+	source ScraperSource
+}
+
+func (t *SceneIdentifier) scrapeScene(scene *models.Scene) (*scrapeResult, error) {
+	// iterate through the input sources
+	for _, source := range t.Sources {
+		// scrape using the source
+		scraped, err := source.Scraper.ScrapeScene(scene.ID)
+		if err != nil {
+			return nil, fmt.Errorf("error scraping from %v: %v", source.Scraper, err)
+		}
+
+		// if results were found then return
+		if scraped != nil {
+			return &scrapeResult{
+				result: scraped,
+				source: source,
+			}, nil
+		}
+	}
+
+	return nil, nil
+}
+
+func (t *SceneIdentifier) getSceneUpdater(ctx context.Context, s *models.Scene, result *scrapeResult, repo models.Repository) (*scene.UpdateSet, error) {
+	ret := &scene.UpdateSet{
+		ID: s.ID,
+	}
+
+	options := []models.IdentifyMetadataOptionsInput{}
+	if result.source.Options != nil {
+		options = append(options, *result.source.Options)
+	}
+	if t.DefaultOptions != nil {
+		options = append(options, *t.DefaultOptions)
+	}
+
+	fieldOptions := getFieldOptions(options)
+
+	setOrganized := false
+	for _, o := range options {
+		if o.SetOrganized != nil {
+			setOrganized = *o.SetOrganized
+			break
+		}
+	}
+
+	scraped := result.result
+
+	rel := sceneRelationships{
+		repo:         repo,
+		scene:        s,
+		result:       result,
+		fieldOptions: fieldOptions,
+	}
+
+	ret.Partial = getScenePartial(s, scraped, fieldOptions, setOrganized)
+
+	studioID, err := rel.studio()
+	if err != nil {
+		return nil, fmt.Errorf("error getting studio: %w", err)
+	}
+
+	if studioID != nil {
+		ret.Partial.StudioID = &sql.NullInt64{
+			Int64: *studioID,
+			Valid: true,
+		}
+	}
+
+	ignoreMale := false
+	for _, o := range options {
+		if o.IncludeMalePerformers != nil {
+			ignoreMale = !*o.IncludeMalePerformers
+			break
+		}
+	}
+
+	ret.PerformerIDs, err = rel.performers(ignoreMale)
+	if err != nil {
+		return nil, err
+	}
+
+	ret.TagIDs, err = rel.tags()
+	if err != nil {
+		return nil, err
+	}
+
+	ret.StashIDs, err = rel.stashIDs()
+	if err != nil {
+		return nil, err
+	}
+
+	setCoverImage := false
+	for _, o := range options {
+		if o.SetCoverImage != nil {
+			setCoverImage = *o.SetCoverImage
+			break
+		}
+	}
+
+	if setCoverImage {
+		ret.CoverImage, err = rel.cover(ctx)
+		if err != nil {
+			return nil, err
+		}
+	}
+
+	return ret, nil
+}
+
+func (t *SceneIdentifier) modifyScene(ctx context.Context, repo models.Repository, scene *models.Scene, result *scrapeResult) error {
+	updater, err := t.getSceneUpdater(ctx, scene, result, repo)
+	if err != nil {
+		return err
+	}
+
+	// don't update anything if nothing was set
+	if updater.IsEmpty() {
+		logger.Infof("Nothing to set for %s", scene.Path)
+		return nil
+	}
+
+	_, err = updater.Update(repo.Scene(), t.ScreenshotSetter)
+	if err != nil {
+		return fmt.Errorf("error updating scene: %w", err)
+	}
+
+	// fire post-update hooks
+	updateInput := updater.UpdateInput()
+	fields := utils.NotNilFields(updateInput, "json")
+	t.SceneUpdatePostHookExecutor.ExecuteSceneUpdatePostHooks(ctx, updateInput, fields)
+
+	as := ""
+	title := updater.Partial.Title
+	if title != nil {
+		as = fmt.Sprintf(" as %s", title.String)
+	}
+	logger.Infof("Successfully identified %s%s using %s", scene.Path, as, result.source.Name)
+
+	return nil
+}
+
+func getFieldOptions(options []models.IdentifyMetadataOptionsInput) map[string]*models.IdentifyFieldOptionsInput {
+	// prefer source-specific field strategies, then the defaults
+	ret := make(map[string]*models.IdentifyFieldOptionsInput)
+	for _, oo := range options {
+		for _, f := range oo.FieldOptions {
+			if _, found := ret[f.Field]; !found {
+				ret[f.Field] = f
+			}
+		}
+	}
+
+	return ret
+}
+
+func getScenePartial(scene *models.Scene, scraped *models.ScrapedScene, fieldOptions map[string]*models.IdentifyFieldOptionsInput, setOrganized bool) models.ScenePartial {
+	partial := models.ScenePartial{
+		ID: scene.ID,
+	}
+
+	if scraped.Title != nil && scene.Title.String != *scraped.Title {
+		if shouldSetSingleValueField(fieldOptions["title"], scene.Title.String != "") {
+			partial.Title = models.NullStringPtr(*scraped.Title)
+		}
+	}
+	if scraped.Date != nil && scene.Date.String != *scraped.Date {
+		if shouldSetSingleValueField(fieldOptions["date"], scene.Date.Valid) {
+			partial.Date = &models.SQLiteDate{
+				String: *scraped.Date,
+				Valid:  true,
+			}
+		}
+	}
+	if scraped.Details != nil && scene.Details.String != *scraped.Details {
+		if shouldSetSingleValueField(fieldOptions["details"], scene.Details.String != "") {
+			partial.Details = models.NullStringPtr(*scraped.Details)
+		}
+	}
+	if scraped.URL != nil && scene.URL.String != *scraped.URL {
+		if shouldSetSingleValueField(fieldOptions["url"], scene.URL.String != "") {
+			partial.URL = models.NullStringPtr(*scraped.URL)
+		}
+	}
+
+	if setOrganized && !scene.Organized {
+		// just reuse the boolean since we know it's true
+		partial.Organized = &setOrganized
+	}
+
+	return partial
+}
+
+func shouldSetSingleValueField(strategy *models.IdentifyFieldOptionsInput, hasExistingValue bool) bool {
+	// if unset then default to MERGE
+	fs := models.IdentifyFieldStrategyMerge
+
+	if strategy != nil && strategy.Strategy.IsValid() {
+		fs = strategy.Strategy
+	}
+
+	if fs == models.IdentifyFieldStrategyIgnore {
+		return false
+	}
+
+	return !hasExistingValue || fs == models.IdentifyFieldStrategyOverwrite
+}

pkg/identify/identify_test.go

@@ -0,0 +1,502 @@
+package identify
+
+import (
+	"context"
+	"errors"
+	"reflect"
+	"testing"
+
+	"github.com/stashapp/stash/pkg/models"
+	"github.com/stashapp/stash/pkg/models/mocks"
+	"github.com/stashapp/stash/pkg/utils"
+	"github.com/stretchr/testify/mock"
+)
+
+type mockSceneScraper struct {
+	errIDs  []int
+	results map[int]*models.ScrapedScene
+}
+
+func (s mockSceneScraper) ScrapeScene(sceneID int) (*models.ScrapedScene, error) {
+	if utils.IntInclude(s.errIDs, sceneID) {
+		return nil, errors.New("scrape scene error")
+	}
+	return s.results[sceneID], nil
+}
+
+type mockHookExecutor struct {
+}
+
+func (s mockHookExecutor) ExecuteSceneUpdatePostHooks(ctx context.Context, input models.SceneUpdateInput, inputFields []string) {
+}
+
+func TestSceneIdentifier_Identify(t *testing.T) {
+	const (
+		errID1 = iota
+		errID2
+		missingID
+		found1ID
+		found2ID
+		errUpdateID
+	)
+
+	var scrapedTitle = "scrapedTitle"
+
+	defaultOptions := &models.IdentifyMetadataOptionsInput{}
+	sources := []ScraperSource{
+		{
+			Scraper: mockSceneScraper{
+				errIDs: []int{errID1},
+				results: map[int]*models.ScrapedScene{
+					found1ID: {
+						Title: &scrapedTitle,
+					},
+				},
+			},
+		},
+		{
+			Scraper: mockSceneScraper{
+				errIDs: []int{errID2},
+				results: map[int]*models.ScrapedScene{
+					found2ID: {
+						Title: &scrapedTitle,
+					},
+					errUpdateID: {
+						Title: &scrapedTitle,
+					},
+				},
+			},
+		},
+	}
+
+	repo := mocks.NewTransactionManager()
+	repo.Scene().(*mocks.SceneReaderWriter).On("Update", mock.MatchedBy(func(partial models.ScenePartial) bool {
+		return partial.ID != errUpdateID
+	})).Return(nil, nil)
+	repo.Scene().(*mocks.SceneReaderWriter).On("Update", mock.MatchedBy(func(partial models.ScenePartial) bool {
+		return partial.ID == errUpdateID
+	})).Return(nil, errors.New("update error"))
+
+	tests := []struct {
+		name    string
+		sceneID int
+		wantErr bool
+	}{
+		{
+			"error scraping",
+			errID1,
+			true,
+		},
+		{
+			"error scraping from second",
+			errID2,
+			true,
+		},
+		{
+			"found in first scraper",
+			found1ID,
+			false,
+		},
+		{
+			"found in second scraper",
+			found2ID,
+			false,
+		},
+		{
+			"not found",
+			missingID,
+			false,
+		},
+		{
+			"error modifying",
+			errUpdateID,
+			true,
+		},
+	}
+
+	identifier := SceneIdentifier{
+		DefaultOptions:              defaultOptions,
+		Sources:                     sources,
+		SceneUpdatePostHookExecutor: mockHookExecutor{},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			scene := &models.Scene{
+				ID: tt.sceneID,
+			}
+			if err := identifier.Identify(context.TODO(), repo, scene); (err != nil) != tt.wantErr {
+				t.Errorf("SceneIdentifier.Identify() error = %v, wantErr %v", err, tt.wantErr)
+			}
+		})
+	}
+}
+
+func TestSceneIdentifier_modifyScene(t *testing.T) {
+	repo := mocks.NewTransactionManager()
+	tr := &SceneIdentifier{}
+
+	type args struct {
+		scene  *models.Scene
+		result *scrapeResult
+	}
+	tests := []struct {
+		name    string
+		args    args
+		wantErr bool
+	}{
+		{
+			"empty update",
+			args{
+				&models.Scene{},
+				&scrapeResult{
+					result: &models.ScrapedScene{},
+				},
+			},
+			false,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if err := tr.modifyScene(context.TODO(), repo, tt.args.scene, tt.args.result); (err != nil) != tt.wantErr {
+				t.Errorf("SceneIdentifier.modifyScene() error = %v, wantErr %v", err, tt.wantErr)
+			}
+		})
+	}
+}
+
+func Test_getFieldOptions(t *testing.T) {
+	const (
+		inFirst  = "inFirst"
+		inSecond = "inSecond"
+		inBoth   = "inBoth"
+	)
+
+	type args struct {
+		options []models.IdentifyMetadataOptionsInput
+	}
+	tests := []struct {
+		name string
+		args args
+		want map[string]*models.IdentifyFieldOptionsInput
+	}{
+		{
+			"simple",
+			args{
+				[]models.IdentifyMetadataOptionsInput{
+					{
+						FieldOptions: []*models.IdentifyFieldOptionsInput{
+							{
+								Field:    inFirst,
+								Strategy: models.IdentifyFieldStrategyIgnore,
+							},
+							{
+								Field:    inBoth,
+								Strategy: models.IdentifyFieldStrategyIgnore,
+							},
+						},
+					},
+					{
+						FieldOptions: []*models.IdentifyFieldOptionsInput{
+							{
+								Field:    inSecond,
+								Strategy: models.IdentifyFieldStrategyMerge,
+							},
+							{
+								Field:    inBoth,
+								Strategy: models.IdentifyFieldStrategyMerge,
+							},
+						},
+					},
+				},
+			},
+			map[string]*models.IdentifyFieldOptionsInput{
+				inFirst: {
+					Field:    inFirst,
+					Strategy: models.IdentifyFieldStrategyIgnore,
+				},
+				inSecond: {
+					Field:    inSecond,
+					Strategy: models.IdentifyFieldStrategyMerge,
+				},
+				inBoth: {
+					Field:    inBoth,
+					Strategy: models.IdentifyFieldStrategyIgnore,
+				},
+			},
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if got := getFieldOptions(tt.args.options); !reflect.DeepEqual(got, tt.want) {
+				t.Errorf("getFieldOptions() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+func Test_getScenePartial(t *testing.T) {
+	var (
+		originalTitle   = "originalTitle"
+		originalDate    = "originalDate"
+		originalDetails = "originalDetails"
+		originalURL     = "originalURL"
+	)
+
+	var (
+		scrapedTitle   = "scrapedTitle"
+		scrapedDate    = "scrapedDate"
+		scrapedDetails = "scrapedDetails"
+		scrapedURL     = "scrapedURL"
+	)
+
+	originalScene := &models.Scene{
+		Title: models.NullString(originalTitle),
+		Date: models.SQLiteDate{
+			String: originalDate,
+			Valid:  true,
+		},
+		Details: models.NullString(originalDetails),
+		URL:     models.NullString(originalURL),
+	}
+
+	organisedScene := *originalScene
+	organisedScene.Organized = true
+
+	emptyScene := &models.Scene{}
+
+	postPartial := models.ScenePartial{
+		Title: models.NullStringPtr(scrapedTitle),
+		Date: &models.SQLiteDate{
+			String: scrapedDate,
+			Valid:  true,
+		},
+		Details: models.NullStringPtr(scrapedDetails),
+		URL:     models.NullStringPtr(scrapedURL),
+	}
+
+	scrapedScene := &models.ScrapedScene{
+		Title:   &scrapedTitle,
+		Date:    &scrapedDate,
+		Details: &scrapedDetails,
+		URL:     &scrapedURL,
+	}
+
+	scrapedUnchangedScene := &models.ScrapedScene{
+		Title:   &originalTitle,
+		Date:    &originalDate,
+		Details: &originalDetails,
+		URL:     &originalURL,
+	}
+
+	makeFieldOptions := func(input *models.IdentifyFieldOptionsInput) map[string]*models.IdentifyFieldOptionsInput {
+		return map[string]*models.IdentifyFieldOptionsInput{
+			"title":   input,
+			"date":    input,
+			"details": input,
+			"url":     input,
+		}
+	}
+
+	overwriteAll := makeFieldOptions(&models.IdentifyFieldOptionsInput{
+		Strategy: models.IdentifyFieldStrategyOverwrite,
+	})
+	ignoreAll := makeFieldOptions(&models.IdentifyFieldOptionsInput{
+		Strategy: models.IdentifyFieldStrategyIgnore,
+	})
+	mergeAll := makeFieldOptions(&models.IdentifyFieldOptionsInput{
+		Strategy: models.IdentifyFieldStrategyMerge,
+	})
+
+	setOrganised := true
+
+	type args struct {
+		scene        *models.Scene
+		scraped      *models.ScrapedScene
+		fieldOptions map[string]*models.IdentifyFieldOptionsInput
+		setOrganized bool
+	}
+	tests := []struct {
+		name string
+		args args
+		want models.ScenePartial
+	}{
+		{
+			"overwrite all",
+			args{
+				originalScene,
+				scrapedScene,
+				overwriteAll,
+				false,
+			},
+			postPartial,
+		},
+		{
+			"ignore all",
+			args{
+				originalScene,
+				scrapedScene,
+				ignoreAll,
+				false,
+			},
+			models.ScenePartial{},
+		},
+		{
+			"merge (existing values)",
+			args{
+				originalScene,
+				scrapedScene,
+				mergeAll,
+				false,
+			},
+			models.ScenePartial{},
+		},
+		{
+			"merge (empty values)",
+			args{
+				emptyScene,
+				scrapedScene,
+				mergeAll,
+				false,
+			},
+			postPartial,
+		},
+		{
+			"unchanged",
+			args{
+				originalScene,
+				scrapedUnchangedScene,
+				overwriteAll,
+				false,
+			},
+			models.ScenePartial{},
+		},
+		{
+			"set organized",
+			args{
+				originalScene,
+				scrapedUnchangedScene,
+				overwriteAll,
+				true,
+			},
+			models.ScenePartial{
+				Organized: &setOrganised,
+			},
+		},
+		{
+			"set organized unchanged",
+			args{
+				&organisedScene,
+				scrapedUnchangedScene,
+				overwriteAll,
+				true,
+			},
+			models.ScenePartial{},
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if got := getScenePartial(tt.args.scene, tt.args.scraped, tt.args.fieldOptions, tt.args.setOrganized); !reflect.DeepEqual(got, tt.want) {
+				t.Errorf("getScenePartial() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+func Test_shouldSetSingleValueField(t *testing.T) {
+	const invalid = "invalid"
+
+	type args struct {
+		strategy         *models.IdentifyFieldOptionsInput
+		hasExistingValue bool
+	}
+	tests := []struct {
+		name string
+		args args
+		want bool
+	}{
+		{
+			"ignore",
+			args{
+				&models.IdentifyFieldOptionsInput{
+					Strategy: models.IdentifyFieldStrategyIgnore,
+				},
+				false,
+			},
+			false,
+		},
+		{
+			"merge existing",
+			args{
+				&models.IdentifyFieldOptionsInput{
+					Strategy: models.IdentifyFieldStrategyMerge,
+				},
+				true,
+			},
+			false,
+		},
+		{
+			"merge absent",
+			args{
+				&models.IdentifyFieldOptionsInput{
+					Strategy: models.IdentifyFieldStrategyMerge,
+				},
+				false,
+			},
+			true,
+		},
+		{
+			"overwrite",
+			args{
+				&models.IdentifyFieldOptionsInput{
+					Strategy: models.IdentifyFieldStrategyOverwrite,
+				},
+				true,
+			},
+			true,
+		},
+		{
+			"nil (merge) existing",
+			args{
+				&models.IdentifyFieldOptionsInput{},
+				true,
+			},
+			false,
+		},
+		{
+			"nil (merge) absent",
+			args{
+				&models.IdentifyFieldOptionsInput{},
+				false,
+			},
+			true,
+		},
+		{
+			"invalid (merge) existing",
+			args{
+				&models.IdentifyFieldOptionsInput{
+					Strategy: invalid,
+				},
+				true,
+			},
+			false,
+		},
+		{
+			"invalid (merge) absent",
+			args{
+				&models.IdentifyFieldOptionsInput{
+					Strategy: invalid,
+				},
+				false,
+			},
+			true,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if got := shouldSetSingleValueField(tt.args.strategy, tt.args.hasExistingValue); got != tt.want {
+				t.Errorf("shouldSetSingleValueField() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}

pkg/identify/performer.go

@@ -0,0 +1,108 @@
+package identify
+
+import (
+	"database/sql"
+	"fmt"
+	"strconv"
+	"time"
+
+	"github.com/stashapp/stash/pkg/models"
+	"github.com/stashapp/stash/pkg/utils"
+)
+
+func getPerformerID(endpoint string, r models.Repository, p *models.ScrapedPerformer, createMissing bool) (*int, error) {
+	if p.StoredID != nil {
+		// existing performer, just add it
+		performerID, err := strconv.Atoi(*p.StoredID)
+		if err != nil {
+			return nil, fmt.Errorf("error converting performer ID %s: %w", *p.StoredID, err)
+		}
+
+		return &performerID, nil
+	} else if createMissing && p.Name != nil { // name is mandatory
+		return createMissingPerformer(endpoint, r, p)
+	}
+
+	return nil, nil
+}
+
+func createMissingPerformer(endpoint string, r models.Repository, p *models.ScrapedPerformer) (*int, error) {
+	created, err := r.Performer().Create(scrapedToPerformerInput(p))
+	if err != nil {
+		return nil, fmt.Errorf("error creating performer: %w", err)
+	}
+
+	if endpoint != "" && p.RemoteSiteID != nil {
+		if err := r.Performer().UpdateStashIDs(created.ID, []models.StashID{
+			{
+				Endpoint: endpoint,
+				StashID:  *p.RemoteSiteID,
+			},
+		}); err != nil {
+			return nil, fmt.Errorf("error setting performer stash id: %w", err)
+		}
+	}
+
+	return &created.ID, nil
+}
+
+func scrapedToPerformerInput(performer *models.ScrapedPerformer) models.Performer {
+	currentTime := time.Now()
+	ret := models.Performer{
+		Name:      sql.NullString{String: *performer.Name, Valid: true},
+		Checksum:  utils.MD5FromString(*performer.Name),
+		CreatedAt: models.SQLiteTimestamp{Timestamp: currentTime},
+		UpdatedAt: models.SQLiteTimestamp{Timestamp: currentTime},
+		Favorite:  sql.NullBool{Bool: false, Valid: true},
+	}
+	if performer.Birthdate != nil {
+		ret.Birthdate = models.SQLiteDate{String: *performer.Birthdate, Valid: true}
+	}
+	if performer.DeathDate != nil {
+		ret.DeathDate = models.SQLiteDate{String: *performer.DeathDate, Valid: true}
+	}
+	if performer.Gender != nil {
+		ret.Gender = sql.NullString{String: *performer.Gender, Valid: true}
+	}
+	if performer.Ethnicity != nil {
+		ret.Ethnicity = sql.NullString{String: *performer.Ethnicity, Valid: true}
+	}
+	if performer.Country != nil {
+		ret.Country = sql.NullString{String: *performer.Country, Valid: true}
+	}
+	if performer.EyeColor != nil {
+		ret.EyeColor = sql.NullString{String: *performer.EyeColor, Valid: true}
+	}
+	if performer.HairColor != nil {
+		ret.HairColor = sql.NullString{String: *performer.HairColor, Valid: true}
+	}
+	if performer.Height != nil {
+		ret.Height = sql.NullString{String: *performer.Height, Valid: true}
+	}
+	if performer.Measurements != nil {
+		ret.Measurements = sql.NullString{String: *performer.Measurements, Valid: true}
+	}
+	if performer.FakeTits != nil {
+		ret.FakeTits = sql.NullString{String: *performer.FakeTits, Valid: true}
+	}
+	if performer.CareerLength != nil {
+		ret.CareerLength = sql.NullString{String: *performer.CareerLength, Valid: true}
+	}
+	if performer.Tattoos != nil {
+		ret.Tattoos = sql.NullString{String: *performer.Tattoos, Valid: true}
+	}
+	if performer.Piercings != nil {
+		ret.Piercings = sql.NullString{String: *performer.Piercings, Valid: true}
+	}
+	if performer.Aliases != nil {
+		ret.Aliases = sql.NullString{String: *performer.Aliases, Valid: true}
+	}
+	if performer.Twitter != nil {
+		ret.Twitter = sql.NullString{String: *performer.Twitter, Valid: true}
+	}
+	if performer.Instagram != nil {
+		ret.Instagram = sql.NullString{String: *performer.Instagram, Valid: true}
+	}
+
+	return ret
+}

pkg/identify/performer_test.go

@@ -0,0 +1,329 @@
+package identify
+
+import (
+	"database/sql"
+	"errors"
+	"reflect"
+	"strconv"
+	"testing"
+
+	"github.com/stashapp/stash/pkg/models"
+	"github.com/stashapp/stash/pkg/models/mocks"
+
+	"github.com/stretchr/testify/mock"
+)
+
+func Test_getPerformerID(t *testing.T) {
+	const (
+		emptyEndpoint = ""
+		endpoint      = "endpoint"
+	)
+	invalidStoredID := "invalidStoredID"
+	validStoredIDStr := "1"
+	validStoredID := 1
+	name := "name"
+
+	repo := mocks.NewTransactionManager()
+	repo.PerformerMock().On("Create", mock.Anything).Return(&models.Performer{
+		ID: validStoredID,
+	}, nil)
+
+	type args struct {
+		endpoint      string
+		p             *models.ScrapedPerformer
+		createMissing bool
+	}
+	tests := []struct {
+		name    string
+		args    args
+		want    *int
+		wantErr bool
+	}{
+		{
+			"no performer",
+			args{
+				emptyEndpoint,
+				&models.ScrapedPerformer{},
+				false,
+			},
+			nil,
+			false,
+		},
+		{
+			"invalid stored id",
+			args{
+				emptyEndpoint,
+				&models.ScrapedPerformer{
+					StoredID: &invalidStoredID,
+				},
+				false,
+			},
+			nil,
+			true,
+		},
+		{
+			"valid stored id",
+			args{
+				emptyEndpoint,
+				&models.ScrapedPerformer{
+					StoredID: &validStoredIDStr,
+				},
+				false,
+			},
+			&validStoredID,
+			false,
+		},
+		{
+			"nil stored not creating",
+			args{
+				emptyEndpoint,
+				&models.ScrapedPerformer{
+					Name: &name,
+				},
+				false,
+			},
+			nil,
+			false,
+		},
+		{
+			"nil name creating",
+			args{
+				emptyEndpoint,
+				&models.ScrapedPerformer{},
+				true,
+			},
+			nil,
+			false,
+		},
+		{
+			"valid name creating",
+			args{
+				emptyEndpoint,
+				&models.ScrapedPerformer{
+					Name: &name,
+				},
+				true,
+			},
+			&validStoredID,
+			false,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got, err := getPerformerID(tt.args.endpoint, repo, tt.args.p, tt.args.createMissing)
+			if (err != nil) != tt.wantErr {
+				t.Errorf("getPerformerID() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+			if !reflect.DeepEqual(got, tt.want) {
+				t.Errorf("getPerformerID() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+func Test_createMissingPerformer(t *testing.T) {
+	emptyEndpoint := ""
+	validEndpoint := "validEndpoint"
+	invalidEndpoint := "invalidEndpoint"
+	remoteSiteID := "remoteSiteID"
+	validName := "validName"
+	invalidName := "invalidName"
+	performerID := 1
+
+	repo := mocks.NewTransactionManager()
+	repo.PerformerMock().On("Create", mock.MatchedBy(func(p models.Performer) bool {
+		return p.Name.String == validName
+	})).Return(&models.Performer{
+		ID: performerID,
+	}, nil)
+	repo.PerformerMock().On("Create", mock.MatchedBy(func(p models.Performer) bool {
+		return p.Name.String == invalidName
+	})).Return(nil, errors.New("error creating performer"))
+
+	repo.PerformerMock().On("UpdateStashIDs", performerID, []models.StashID{
+		{
+			Endpoint: invalidEndpoint,
+			StashID:  remoteSiteID,
+		},
+	}).Return(errors.New("error updating stash ids"))
+	repo.PerformerMock().On("UpdateStashIDs", performerID, []models.StashID{
+		{
+			Endpoint: validEndpoint,
+			StashID:  remoteSiteID,
+		},
+	}).Return(nil)
+
+	type args struct {
+		endpoint string
+		p        *models.ScrapedPerformer
+	}
+	tests := []struct {
+		name    string
+		args    args
+		want    *int
+		wantErr bool
+	}{
+		{
+			"simple",
+			args{
+				emptyEndpoint,
+				&models.ScrapedPerformer{
+					Name: &validName,
+				},
+			},
+			&performerID,
+			false,
+		},
+		{
+			"error creating",
+			args{
+				emptyEndpoint,
+				&models.ScrapedPerformer{
+					Name: &invalidName,
+				},
+			},
+			nil,
+			true,
+		},
+		{
+			"valid stash id",
+			args{
+				validEndpoint,
+				&models.ScrapedPerformer{
+					Name:         &validName,
+					RemoteSiteID: &remoteSiteID,
+				},
+			},
+			&performerID,
+			false,
+		},
+		{
+			"invalid stash id",
+			args{
+				invalidEndpoint,
+				&models.ScrapedPerformer{
+					Name:         &validName,
+					RemoteSiteID: &remoteSiteID,
+				},
+			},
+			nil,
+			true,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got, err := createMissingPerformer(tt.args.endpoint, repo, tt.args.p)
+			if (err != nil) != tt.wantErr {
+				t.Errorf("createMissingPerformer() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+			if !reflect.DeepEqual(got, tt.want) {
+				t.Errorf("createMissingPerformer() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+func Test_scrapedToPerformerInput(t *testing.T) {
+	name := "name"
+	md5 := "b068931cc450442b63f5b3d276ea4297"
+
+	var stringValues []string
+	for i := 0; i < 16; i++ {
+		stringValues = append(stringValues, strconv.Itoa(i))
+	}
+
+	upTo := 0
+	nextVal := func() *string {
+		ret := stringValues[upTo]
+		upTo = (upTo + 1) % len(stringValues)
+		return &ret
+	}
+
+	tests := []struct {
+		name      string
+		performer *models.ScrapedPerformer
+		want      models.Performer
+	}{
+		{
+			"set all",
+			&models.ScrapedPerformer{
+				Name:         &name,
+				Birthdate:    nextVal(),
+				DeathDate:    nextVal(),
+				Gender:       nextVal(),
+				Ethnicity:    nextVal(),
+				Country:      nextVal(),
+				EyeColor:     nextVal(),
+				HairColor:    nextVal(),
+				Height:       nextVal(),
+				Measurements: nextVal(),
+				FakeTits:     nextVal(),
+				CareerLength: nextVal(),
+				Tattoos:      nextVal(),
+				Piercings:    nextVal(),
+				Aliases:      nextVal(),
+				Twitter:      nextVal(),
+				Instagram:    nextVal(),
+			},
+			models.Performer{
+				Name:     models.NullString(name),
+				Checksum: md5,
+				Favorite: sql.NullBool{
+					Bool:  false,
+					Valid: true,
+				},
+				Birthdate: models.SQLiteDate{
+					String: *nextVal(),
+					Valid:  true,
+				},
+				DeathDate: models.SQLiteDate{
+					String: *nextVal(),
+					Valid:  true,
+				},
+				Gender:       models.NullString(*nextVal()),
+				Ethnicity:    models.NullString(*nextVal()),
+				Country:      models.NullString(*nextVal()),
+				EyeColor:     models.NullString(*nextVal()),
+				HairColor:    models.NullString(*nextVal()),
+				Height:       models.NullString(*nextVal()),
+				Measurements: models.NullString(*nextVal()),
+				FakeTits:     models.NullString(*nextVal()),
+				CareerLength: models.NullString(*nextVal()),
+				Tattoos:      models.NullString(*nextVal()),
+				Piercings:    models.NullString(*nextVal()),
+				Aliases:      models.NullString(*nextVal()),
+				Twitter:      models.NullString(*nextVal()),
+				Instagram:    models.NullString(*nextVal()),
+			},
+		},
+		{
+			"set none",
+			&models.ScrapedPerformer{
+				Name: &name,
+			},
+			models.Performer{
+				Name:     models.NullString(name),
+				Checksum: md5,
+				Favorite: sql.NullBool{
+					Bool:  false,
+					Valid: true,
+				},
+			},
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got := scrapedToPerformerInput(tt.performer)
+
+			// clear created/updated dates
+			got.CreatedAt = models.SQLiteTimestamp{}
+			got.UpdatedAt = got.CreatedAt
+
+			if !reflect.DeepEqual(got, tt.want) {
+				t.Errorf("scrapedToPerformerInput() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}

pkg/identify/scene.go

@@ -0,0 +1,251 @@
+package identify
+
+import (
+	"bytes"
+	"context"
+	"fmt"
+	"strconv"
+	"strings"
+	"time"
+
+	"github.com/stashapp/stash/pkg/models"
+	"github.com/stashapp/stash/pkg/utils"
+)
+
+type sceneRelationships struct {
+	repo         models.Repository
+	scene        *models.Scene
+	result       *scrapeResult
+	fieldOptions map[string]*models.IdentifyFieldOptionsInput
+}
+
+func (g sceneRelationships) studio() (*int64, error) {
+	existingID := g.scene.StudioID
+	fieldStrategy := g.fieldOptions["studio"]
+	createMissing := fieldStrategy != nil && utils.IsTrue(fieldStrategy.CreateMissing)
+
+	scraped := g.result.result.Studio
+	endpoint := g.result.source.RemoteSite
+
+	if scraped == nil || !shouldSetSingleValueField(fieldStrategy, existingID.Valid) {
+		return nil, nil
+	}
+
+	if scraped.StoredID != nil {
+		// existing studio, just set it
+		studioID, err := strconv.ParseInt(*scraped.StoredID, 10, 64)
+		if err != nil {
+			return nil, fmt.Errorf("error converting studio ID %s: %w", *scraped.StoredID, err)
+		}
+
+		// only return value if different to current
+		if existingID.Int64 != studioID {
+			return &studioID, nil
+		}
+	} else if createMissing {
+		return createMissingStudio(endpoint, g.repo, scraped)
+	}
+
+	return nil, nil
+}
+
+func (g sceneRelationships) performers(ignoreMale bool) ([]int, error) {
+	fieldStrategy := g.fieldOptions["performers"]
+	scraped := g.result.result.Performers
+
+	// just check if ignored
+	if len(scraped) == 0 || !shouldSetSingleValueField(fieldStrategy, false) {
+		return nil, nil
+	}
+
+	createMissing := fieldStrategy != nil && utils.IsTrue(fieldStrategy.CreateMissing)
+	strategy := models.IdentifyFieldStrategyMerge
+	if fieldStrategy != nil {
+		strategy = fieldStrategy.Strategy
+	}
+
+	repo := g.repo
+	endpoint := g.result.source.RemoteSite
+
+	var performerIDs []int
+	originalPerformerIDs, err := repo.Scene().GetPerformerIDs(g.scene.ID)
+	if err != nil {
+		return nil, fmt.Errorf("error getting scene performers: %w", err)
+	}
+
+	if strategy == models.IdentifyFieldStrategyMerge {
+		// add to existing
+		performerIDs = originalPerformerIDs
+	}
+
+	for _, p := range scraped {
+		if ignoreMale && p.Gender != nil && strings.EqualFold(*p.Gender, models.GenderEnumMale.String()) {
+			continue
+		}
+
+		performerID, err := getPerformerID(endpoint, repo, p, createMissing)
+		if err != nil {
+			return nil, err
+		}
+
+		if performerID != nil {
+			performerIDs = utils.IntAppendUnique(performerIDs, *performerID)
+		}
+	}
+
+	// don't return if nothing was added
+	if utils.SliceSame(originalPerformerIDs, performerIDs) {
+		return nil, nil
+	}
+
+	return performerIDs, nil
+}
+
+func (g sceneRelationships) tags() ([]int, error) {
+	fieldStrategy := g.fieldOptions["tags"]
+	scraped := g.result.result.Tags
+	target := g.scene
+	r := g.repo
+
+	// just check if ignored
+	if len(scraped) == 0 || !shouldSetSingleValueField(fieldStrategy, false) {
+		return nil, nil
+	}
+
+	createMissing := fieldStrategy != nil && utils.IsTrue(fieldStrategy.CreateMissing)
+	strategy := models.IdentifyFieldStrategyMerge
+	if fieldStrategy != nil {
+		strategy = fieldStrategy.Strategy
+	}
+
+	var tagIDs []int
+	originalTagIDs, err := r.Scene().GetTagIDs(target.ID)
+	if err != nil {
+		return nil, fmt.Errorf("error getting scene tags: %w", err)
+	}
+
+	if strategy == models.IdentifyFieldStrategyMerge {
+		// add to existing
+		tagIDs = originalTagIDs
+	}
+
+	for _, t := range scraped {
+		if t.StoredID != nil {
+			// existing tag, just add it
+			tagID, err := strconv.ParseInt(*t.StoredID, 10, 64)
+			if err != nil {
+				return nil, fmt.Errorf("error converting tag ID %s: %w", *t.StoredID, err)
+			}
+
+			tagIDs = utils.IntAppendUnique(tagIDs, int(tagID))
+		} else if createMissing {
+			now := time.Now()
+			created, err := r.Tag().Create(models.Tag{
+				Name:      t.Name,
+				CreatedAt: models.SQLiteTimestamp{Timestamp: now},
+				UpdatedAt: models.SQLiteTimestamp{Timestamp: now},
+			})
+			if err != nil {
+				return nil, fmt.Errorf("error creating tag: %w", err)
+			}
+
+			tagIDs = append(tagIDs, created.ID)
+		}
+	}
+
+	// don't return if nothing was added
+	if utils.SliceSame(originalTagIDs, tagIDs) {
+		return nil, nil
+	}
+
+	return tagIDs, nil
+}
+
+func (g sceneRelationships) stashIDs() ([]models.StashID, error) {
+	remoteSiteID := g.result.result.RemoteSiteID
+	fieldStrategy := g.fieldOptions["stash_ids"]
+	target := g.scene
+	r := g.repo
+
+	endpoint := g.result.source.RemoteSite
+
+	// just check if ignored
+	if remoteSiteID == nil || endpoint == "" || !shouldSetSingleValueField(fieldStrategy, false) {
+		return nil, nil
+	}
+
+	strategy := models.IdentifyFieldStrategyMerge
+	if fieldStrategy != nil {
+		strategy = fieldStrategy.Strategy
+	}
+
+	var originalStashIDs []models.StashID
+	var stashIDs []models.StashID
+	stashIDPtrs, err := r.Scene().GetStashIDs(target.ID)
+	if err != nil {
+		return nil, fmt.Errorf("error getting scene tag: %w", err)
+	}
+
+	// convert existing to non-pointer types
+	for _, stashID := range stashIDPtrs {
+		originalStashIDs = append(originalStashIDs, *stashID)
+	}
+
+	if strategy == models.IdentifyFieldStrategyMerge {
+		// add to existing
+		stashIDs = originalStashIDs
+	}
+
+	for i, stashID := range stashIDs {
+		if endpoint == stashID.Endpoint {
+			// if stashID is the same, then don't set
+			if stashID.StashID == *remoteSiteID {
+				return nil, nil
+			}
+
+			// replace the stash id and return
+			stashID.StashID = *remoteSiteID
+			stashIDs[i] = stashID
+			return stashIDs, nil
+		}
+	}
+
+	// not found, create new entry
+	stashIDs = append(stashIDs, models.StashID{
+		StashID:  *remoteSiteID,
+		Endpoint: endpoint,
+	})
+
+	if utils.SliceSame(originalStashIDs, stashIDs) {
+		return nil, nil
+	}
+
+	return stashIDs, nil
+}
+
+func (g sceneRelationships) cover(ctx context.Context) ([]byte, error) {
+	scraped := g.result.result.Image
+	r := g.repo
+
+	if scraped == nil {
+		return nil, nil
+	}
+
+	// always overwrite if present
+	existingCover, err := r.Scene().GetCover(g.scene.ID)
+	if err != nil {
+		return nil, fmt.Errorf("error getting scene cover: %w", err)
+	}
+
+	data, err := utils.ProcessImageInput(ctx, *scraped)
+	if err != nil {
+		return nil, fmt.Errorf("error processing image input: %w", err)
+	}
+
+	// only return if different
+	if !bytes.Equal(existingCover, data) {
+		return data, nil
+	}
+
+	return nil, nil
+}

pkg/identify/scene_test.go

@@ -0,0 +1,782 @@
+package identify
+
+import (
+	"context"
+	"errors"
+	"reflect"
+	"strconv"
+	"testing"
+
+	"github.com/stashapp/stash/pkg/models"
+	"github.com/stashapp/stash/pkg/models/mocks"
+	"github.com/stashapp/stash/pkg/utils"
+	"github.com/stretchr/testify/mock"
+)
+
+func Test_sceneRelationships_studio(t *testing.T) {
+	validStoredID := "1"
+	var validStoredIDInt int64 = 1
+	invalidStoredID := "invalidStoredID"
+	createMissing := true
+
+	defaultOptions := &models.IdentifyFieldOptionsInput{
+		Strategy: models.IdentifyFieldStrategyMerge,
+	}
+
+	repo := mocks.NewTransactionManager()
+	repo.StudioMock().On("Create", mock.Anything).Return(&models.Studio{
+		ID: int(validStoredIDInt),
+	}, nil)
+
+	tr := sceneRelationships{
+		repo:         repo,
+		fieldOptions: make(map[string]*models.IdentifyFieldOptionsInput),
+	}
+
+	tests := []struct {
+		name         string
+		scene        *models.Scene
+		fieldOptions *models.IdentifyFieldOptionsInput
+		result       *models.ScrapedStudio
+		want         *int64
+		wantErr      bool
+	}{
+		{
+			"nil studio",
+			&models.Scene{},
+			defaultOptions,
+			nil,
+			nil,
+			false,
+		},
+		{
+			"ignore",
+			&models.Scene{},
+			&models.IdentifyFieldOptionsInput{
+				Strategy: models.IdentifyFieldStrategyIgnore,
+			},
+			&models.ScrapedStudio{
+				StoredID: &validStoredID,
+			},
+			nil,
+			false,
+		},
+		{
+			"invalid stored id",
+			&models.Scene{},
+			defaultOptions,
+			&models.ScrapedStudio{
+				StoredID: &invalidStoredID,
+			},
+			nil,
+			true,
+		},
+		{
+			"same stored id",
+			&models.Scene{
+				StudioID: models.NullInt64(validStoredIDInt),
+			},
+			defaultOptions,
+			&models.ScrapedStudio{
+				StoredID: &validStoredID,
+			},
+			nil,
+			false,
+		},
+		{
+			"different stored id",
+			&models.Scene{},
+			defaultOptions,
+			&models.ScrapedStudio{
+				StoredID: &validStoredID,
+			},
+			&validStoredIDInt,
+			false,
+		},
+		{
+			"no create missing",
+			&models.Scene{},
+			defaultOptions,
+			&models.ScrapedStudio{},
+			nil,
+			false,
+		},
+		{
+			"create missing",
+			&models.Scene{},
+			&models.IdentifyFieldOptionsInput{
+				Strategy:      models.IdentifyFieldStrategyMerge,
+				CreateMissing: &createMissing,
+			},
+			&models.ScrapedStudio{},
+			&validStoredIDInt,
+			false,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			tr.scene = tt.scene
+			tr.fieldOptions["studio"] = tt.fieldOptions
+			tr.result = &scrapeResult{
+				result: &models.ScrapedScene{
+					Studio: tt.result,
+				},
+			}
+
+			got, err := tr.studio()
+			if (err != nil) != tt.wantErr {
+				t.Errorf("sceneRelationships.studio() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+			if !reflect.DeepEqual(got, tt.want) {
+				t.Errorf("sceneRelationships.studio() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+func Test_sceneRelationships_performers(t *testing.T) {
+	const (
+		sceneID = iota
+		sceneWithPerformerID
+		errSceneID
+		existingPerformerID
+		validStoredIDInt
+	)
+	validStoredID := strconv.Itoa(validStoredIDInt)
+	invalidStoredID := "invalidStoredID"
+	createMissing := true
+	existingPerformerStr := strconv.Itoa(existingPerformerID)
+	validName := "validName"
+	female := models.GenderEnumFemale.String()
+	male := models.GenderEnumMale.String()
+
+	defaultOptions := &models.IdentifyFieldOptionsInput{
+		Strategy: models.IdentifyFieldStrategyMerge,
+	}
+
+	repo := mocks.NewTransactionManager()
+	repo.SceneMock().On("GetPerformerIDs", sceneID).Return(nil, nil)
+	repo.SceneMock().On("GetPerformerIDs", sceneWithPerformerID).Return([]int{existingPerformerID}, nil)
+	repo.SceneMock().On("GetPerformerIDs", errSceneID).Return(nil, errors.New("error getting IDs"))
+
+	tr := sceneRelationships{
+		repo:         repo,
+		fieldOptions: make(map[string]*models.IdentifyFieldOptionsInput),
+	}
+
+	tests := []struct {
+		name         string
+		sceneID      int
+		fieldOptions *models.IdentifyFieldOptionsInput
+		scraped      []*models.ScrapedPerformer
+		ignoreMale   bool
+		want         []int
+		wantErr      bool
+	}{
+		{
+			"ignore",
+			sceneID,
+			&models.IdentifyFieldOptionsInput{
+				Strategy: models.IdentifyFieldStrategyIgnore,
+			},
+			[]*models.ScrapedPerformer{
+				{
+					StoredID: &validStoredID,
+				},
+			},
+			false,
+			nil,
+			false,
+		},
+		{
+			"none",
+			sceneID,
+			defaultOptions,
+			[]*models.ScrapedPerformer{},
+			false,
+			nil,
+			false,
+		},
+		{
+			"error getting ids",
+			errSceneID,
+			defaultOptions,
+			[]*models.ScrapedPerformer{
+				{},
+			},
+			false,
+			nil,
+			true,
+		},
+		{
+			"merge existing",
+			sceneWithPerformerID,
+			defaultOptions,
+			[]*models.ScrapedPerformer{
+				{
+					Name:     &validName,
+					StoredID: &existingPerformerStr,
+				},
+			},
+			false,
+			nil,
+			false,
+		},
+		{
+			"merge add",
+			sceneWithPerformerID,
+			defaultOptions,
+			[]*models.ScrapedPerformer{
+				{
+					Name:     &validName,
+					StoredID: &validStoredID,
+				},
+			},
+			false,
+			[]int{existingPerformerID, validStoredIDInt},
+			false,
+		},
+		{
+			"ignore male",
+			sceneID,
+			defaultOptions,
+			[]*models.ScrapedPerformer{
+				{
+					Name:     &validName,
+					StoredID: &validStoredID,
+					Gender:   &male,
+				},
+			},
+			true,
+			nil,
+			false,
+		},
+		{
+			"overwrite",
+			sceneWithPerformerID,
+			&models.IdentifyFieldOptionsInput{
+				Strategy: models.IdentifyFieldStrategyOverwrite,
+			},
+			[]*models.ScrapedPerformer{
+				{
+					Name:     &validName,
+					StoredID: &validStoredID,
+				},
+			},
+			false,
+			[]int{validStoredIDInt},
+			false,
+		},
+		{
+			"ignore male (not male)",
+			sceneWithPerformerID,
+			&models.IdentifyFieldOptionsInput{
+				Strategy: models.IdentifyFieldStrategyOverwrite,
+			},
+			[]*models.ScrapedPerformer{
+				{
+					Name:     &validName,
+					StoredID: &validStoredID,
+					Gender:   &female,
+				},
+			},
+			true,
+			[]int{validStoredIDInt},
+			false,
+		},
+		{
+			"error getting tag ID",
+			sceneID,
+			&models.IdentifyFieldOptionsInput{
+				Strategy:      models.IdentifyFieldStrategyOverwrite,
+				CreateMissing: &createMissing,
+			},
+			[]*models.ScrapedPerformer{
+				{
+					Name:     &validName,
+					StoredID: &invalidStoredID,
+				},
+			},
+			false,
+			nil,
+			true,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			tr.scene = &models.Scene{
+				ID: tt.sceneID,
+			}
+			tr.fieldOptions["performers"] = tt.fieldOptions
+			tr.result = &scrapeResult{
+				result: &models.ScrapedScene{
+					Performers: tt.scraped,
+				},
+			}
+
+			got, err := tr.performers(tt.ignoreMale)
+			if (err != nil) != tt.wantErr {
+				t.Errorf("sceneRelationships.performers() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+			if !reflect.DeepEqual(got, tt.want) {
+				t.Errorf("sceneRelationships.performers() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+func Test_sceneRelationships_tags(t *testing.T) {
+	const (
+		sceneID = iota
+		sceneWithTagID
+		errSceneID
+		existingID
+		validStoredIDInt
+	)
+	validStoredID := strconv.Itoa(validStoredIDInt)
+	invalidStoredID := "invalidStoredID"
+	createMissing := true
+	existingIDStr := strconv.Itoa(existingID)
+	validName := "validName"
+	invalidName := "invalidName"
+
+	defaultOptions := &models.IdentifyFieldOptionsInput{
+		Strategy: models.IdentifyFieldStrategyMerge,
+	}
+
+	repo := mocks.NewTransactionManager()
+	repo.SceneMock().On("GetTagIDs", sceneID).Return(nil, nil)
+	repo.SceneMock().On("GetTagIDs", sceneWithTagID).Return([]int{existingID}, nil)
+	repo.SceneMock().On("GetTagIDs", errSceneID).Return(nil, errors.New("error getting IDs"))
+
+	repo.TagMock().On("Create", mock.MatchedBy(func(p models.Tag) bool {
+		return p.Name == validName
+	})).Return(&models.Tag{
+		ID: validStoredIDInt,
+	}, nil)
+	repo.TagMock().On("Create", mock.MatchedBy(func(p models.Tag) bool {
+		return p.Name == invalidName
+	})).Return(nil, errors.New("error creating tag"))
+
+	tr := sceneRelationships{
+		repo:         repo,
+		fieldOptions: make(map[string]*models.IdentifyFieldOptionsInput),
+	}
+
+	tests := []struct {
+		name         string
+		sceneID      int
+		fieldOptions *models.IdentifyFieldOptionsInput
+		scraped      []*models.ScrapedTag
+		want         []int
+		wantErr      bool
+	}{
+		{
+			"ignore",
+			sceneID,
+			&models.IdentifyFieldOptionsInput{
+				Strategy: models.IdentifyFieldStrategyIgnore,
+			},
+			[]*models.ScrapedTag{
+				{
+					StoredID: &validStoredID,
+				},
+			},
+			nil,
+			false,
+		},
+		{
+			"none",
+			sceneID,
+			defaultOptions,
+			[]*models.ScrapedTag{},
+			nil,
+			false,
+		},
+		{
+			"error getting ids",
+			errSceneID,
+			defaultOptions,
+			[]*models.ScrapedTag{
+				{},
+			},
+			nil,
+			true,
+		},
+		{
+			"merge existing",
+			sceneWithTagID,
+			defaultOptions,
+			[]*models.ScrapedTag{
+				{
+					Name:     validName,
+					StoredID: &existingIDStr,
+				},
+			},
+			nil,
+			false,
+		},
+		{
+			"merge add",
+			sceneWithTagID,
+			defaultOptions,
+			[]*models.ScrapedTag{
+				{
+					Name:     validName,
+					StoredID: &validStoredID,
+				},
+			},
+			[]int{existingID, validStoredIDInt},
+			false,
+		},
+		{
+			"overwrite",
+			sceneWithTagID,
+			&models.IdentifyFieldOptionsInput{
+				Strategy: models.IdentifyFieldStrategyOverwrite,
+			},
+			[]*models.ScrapedTag{
+				{
+					Name:     validName,
+					StoredID: &validStoredID,
+				},
+			},
+			[]int{validStoredIDInt},
+			false,
+		},
+		{
+			"error getting tag ID",
+			sceneID,
+			&models.IdentifyFieldOptionsInput{
+				Strategy: models.IdentifyFieldStrategyOverwrite,
+			},
+			[]*models.ScrapedTag{
+				{
+					Name:     validName,
+					StoredID: &invalidStoredID,
+				},
+			},
+			nil,
+			true,
+		},
+		{
+			"create missing",
+			sceneID,
+			&models.IdentifyFieldOptionsInput{
+				Strategy:      models.IdentifyFieldStrategyOverwrite,
+				CreateMissing: &createMissing,
+			},
+			[]*models.ScrapedTag{
+				{
+					Name: validName,
+				},
+			},
+			[]int{validStoredIDInt},
+			false,
+		},
+		{
+			"error creating",
+			sceneID,
+			&models.IdentifyFieldOptionsInput{
+				Strategy:      models.IdentifyFieldStrategyOverwrite,
+				CreateMissing: &createMissing,
+			},
+			[]*models.ScrapedTag{
+				{
+					Name: invalidName,
+				},
+			},
+			nil,
+			true,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			tr.scene = &models.Scene{
+				ID: tt.sceneID,
+			}
+			tr.fieldOptions["tags"] = tt.fieldOptions
+			tr.result = &scrapeResult{
+				result: &models.ScrapedScene{
+					Tags: tt.scraped,
+				},
+			}
+
+			got, err := tr.tags()
+			if (err != nil) != tt.wantErr {
+				t.Errorf("sceneRelationships.tags() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+			if !reflect.DeepEqual(got, tt.want) {
+				t.Errorf("sceneRelationships.tags() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+func Test_sceneRelationships_stashIDs(t *testing.T) {
+	const (
+		sceneID = iota
+		sceneWithStashID
+		errSceneID
+		existingID
+		validStoredIDInt
+	)
+	existingEndpoint := "existingEndpoint"
+	newEndpoint := "newEndpoint"
+	remoteSiteID := "remoteSiteID"
+	newRemoteSiteID := "newRemoteSiteID"
+
+	defaultOptions := &models.IdentifyFieldOptionsInput{
+		Strategy: models.IdentifyFieldStrategyMerge,
+	}
+
+	repo := mocks.NewTransactionManager()
+	repo.SceneMock().On("GetStashIDs", sceneID).Return(nil, nil)
+	repo.SceneMock().On("GetStashIDs", sceneWithStashID).Return([]*models.StashID{
+		{
+			StashID:  remoteSiteID,
+			Endpoint: existingEndpoint,
+		},
+	}, nil)
+	repo.SceneMock().On("GetStashIDs", errSceneID).Return(nil, errors.New("error getting IDs"))
+
+	tr := sceneRelationships{
+		repo:         repo,
+		fieldOptions: make(map[string]*models.IdentifyFieldOptionsInput),
+	}
+
+	tests := []struct {
+		name         string
+		sceneID      int
+		fieldOptions *models.IdentifyFieldOptionsInput
+		endpoint     string
+		remoteSiteID *string
+		want         []models.StashID
+		wantErr      bool
+	}{
+		{
+			"ignore",
+			sceneID,
+			&models.IdentifyFieldOptionsInput{
+				Strategy: models.IdentifyFieldStrategyIgnore,
+			},
+			newEndpoint,
+			&remoteSiteID,
+			nil,
+			false,
+		},
+		{
+			"no endpoint",
+			sceneID,
+			defaultOptions,
+			"",
+			&remoteSiteID,
+			nil,
+			false,
+		},
+		{
+			"no site id",
+			sceneID,
+			defaultOptions,
+			newEndpoint,
+			nil,
+			nil,
+			false,
+		},
+		{
+			"error getting ids",
+			errSceneID,
+			defaultOptions,
+			newEndpoint,
+			&remoteSiteID,
+			nil,
+			true,
+		},
+		{
+			"merge existing",
+			sceneWithStashID,
+			defaultOptions,
+			existingEndpoint,
+			&remoteSiteID,
+			nil,
+			false,
+		},
+		{
+			"merge existing new value",
+			sceneWithStashID,
+			defaultOptions,
+			existingEndpoint,
+			&newRemoteSiteID,
+			[]models.StashID{
+				{
+					StashID:  newRemoteSiteID,
+					Endpoint: existingEndpoint,
+				},
+			},
+			false,
+		},
+		{
+			"merge add",
+			sceneWithStashID,
+			defaultOptions,
+			newEndpoint,
+			&newRemoteSiteID,
+			[]models.StashID{
+				{
+					StashID:  remoteSiteID,
+					Endpoint: existingEndpoint,
+				},
+				{
+					StashID:  newRemoteSiteID,
+					Endpoint: newEndpoint,
+				},
+			},
+			false,
+		},
+		{
+			"overwrite",
+			sceneWithStashID,
+			&models.IdentifyFieldOptionsInput{
+				Strategy: models.IdentifyFieldStrategyOverwrite,
+			},
+			newEndpoint,
+			&newRemoteSiteID,
+			[]models.StashID{
+				{
+					StashID:  newRemoteSiteID,
+					Endpoint: newEndpoint,
+				},
+			},
+			false,
+		},
+		{
+			"overwrite same",
+			sceneWithStashID,
+			&models.IdentifyFieldOptionsInput{
+				Strategy: models.IdentifyFieldStrategyOverwrite,
+			},
+			existingEndpoint,
+			&remoteSiteID,
+			nil,
+			false,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			tr.scene = &models.Scene{
+				ID: tt.sceneID,
+			}
+			tr.fieldOptions["stash_ids"] = tt.fieldOptions
+			tr.result = &scrapeResult{
+				source: ScraperSource{
+					RemoteSite: tt.endpoint,
+				},
+				result: &models.ScrapedScene{
+					RemoteSiteID: tt.remoteSiteID,
+				},
+			}
+
+			got, err := tr.stashIDs()
+			if (err != nil) != tt.wantErr {
+				t.Errorf("sceneRelationships.stashIDs() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+			if !reflect.DeepEqual(got, tt.want) {
+				t.Errorf("sceneRelationships.stashIDs() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+func Test_sceneRelationships_cover(t *testing.T) {
+	const (
+		sceneID = iota
+		sceneWithStashID
+		errSceneID
+		existingID
+		validStoredIDInt
+	)
+	existingData := []byte("existingData")
+	newData := []byte("newData")
+	const base64Prefix = "data:image/png;base64,"
+	existingDataEncoded := base64Prefix + utils.GetBase64StringFromData(existingData)
+	newDataEncoded := base64Prefix + utils.GetBase64StringFromData(newData)
+	invalidData := newDataEncoded + "!!!"
+
+	repo := mocks.NewTransactionManager()
+	repo.SceneMock().On("GetCover", sceneID).Return(existingData, nil)
+	repo.SceneMock().On("GetCover", errSceneID).Return(nil, errors.New("error getting cover"))
+
+	tr := sceneRelationships{
+		repo:         repo,
+		fieldOptions: make(map[string]*models.IdentifyFieldOptionsInput),
+	}
+
+	tests := []struct {
+		name    string
+		sceneID int
+		image   *string
+		want    []byte
+		wantErr bool
+	}{
+		{
+			"nil image",
+			sceneID,
+			nil,
+			nil,
+			false,
+		},
+		{
+			"different image",
+			sceneID,
+			&newDataEncoded,
+			newData,
+			false,
+		},
+		{
+			"same image",
+			sceneID,
+			&existingDataEncoded,
+			nil,
+			false,
+		},
+		{
+			"error getting scene cover",
+			errSceneID,
+			&newDataEncoded,
+			nil,
+			true,
+		},
+		{
+			"invalid data",
+			sceneID,
+			&invalidData,
+			nil,
+			true,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			tr.scene = &models.Scene{
+				ID: tt.sceneID,
+			}
+			tr.result = &scrapeResult{
+				result: &models.ScrapedScene{
+					Image: tt.image,
+				},
+			}
+
+			got, err := tr.cover(context.TODO())
+			if (err != nil) != tt.wantErr {
+				t.Errorf("sceneRelationships.cover() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+			if !reflect.DeepEqual(got, tt.want) {
+				t.Errorf("sceneRelationships.cover() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}

pkg/identify/studio.go

@@ -0,0 +1,47 @@
+package identify
+
+import (
+	"database/sql"
+	"fmt"
+	"time"
+
+	"github.com/stashapp/stash/pkg/models"
+	"github.com/stashapp/stash/pkg/utils"
+)
+
+func createMissingStudio(endpoint string, repo models.Repository, studio *models.ScrapedStudio) (*int64, error) {
+	created, err := repo.Studio().Create(scrapedToStudioInput(studio))
+	if err != nil {
+		return nil, fmt.Errorf("error creating studio: %w", err)
+	}
+
+	if endpoint != "" && studio.RemoteSiteID != nil {
+		if err := repo.Studio().UpdateStashIDs(created.ID, []models.StashID{
+			{
+				Endpoint: endpoint,
+				StashID:  *studio.RemoteSiteID,
+			},
+		}); err != nil {
+			return nil, fmt.Errorf("error setting studio stash id: %w", err)
+		}
+	}
+
+	createdID := int64(created.ID)
+	return &createdID, nil
+}
+
+func scrapedToStudioInput(studio *models.ScrapedStudio) models.Studio {
+	currentTime := time.Now()
+	ret := models.Studio{
+		Name:      sql.NullString{String: studio.Name, Valid: true},
+		Checksum:  utils.MD5FromString(studio.Name),
+		CreatedAt: models.SQLiteTimestamp{Timestamp: currentTime},
+		UpdatedAt: models.SQLiteTimestamp{Timestamp: currentTime},
+	}
+
+	if studio.URL != nil {
+		ret.URL = sql.NullString{String: *studio.URL, Valid: true}
+	}
+
+	return ret
+}

pkg/identify/studio_test.go

@@ -0,0 +1,163 @@
+package identify
+
+import (
+	"errors"
+	"reflect"
+	"testing"
+
+	"github.com/stashapp/stash/pkg/models"
+	"github.com/stashapp/stash/pkg/models/mocks"
+	"github.com/stretchr/testify/mock"
+)
+
+func Test_createMissingStudio(t *testing.T) {
+	emptyEndpoint := ""
+	validEndpoint := "validEndpoint"
+	invalidEndpoint := "invalidEndpoint"
+	remoteSiteID := "remoteSiteID"
+	validName := "validName"
+	invalidName := "invalidName"
+	createdID := 1
+	createdID64 := int64(createdID)
+
+	repo := mocks.NewTransactionManager()
+	repo.StudioMock().On("Create", mock.MatchedBy(func(p models.Studio) bool {
+		return p.Name.String == validName
+	})).Return(&models.Studio{
+		ID: createdID,
+	}, nil)
+	repo.StudioMock().On("Create", mock.MatchedBy(func(p models.Studio) bool {
+		return p.Name.String == invalidName
+	})).Return(nil, errors.New("error creating performer"))
+
+	repo.StudioMock().On("UpdateStashIDs", createdID, []models.StashID{
+		{
+			Endpoint: invalidEndpoint,
+			StashID:  remoteSiteID,
+		},
+	}).Return(errors.New("error updating stash ids"))
+	repo.StudioMock().On("UpdateStashIDs", createdID, []models.StashID{
+		{
+			Endpoint: validEndpoint,
+			StashID:  remoteSiteID,
+		},
+	}).Return(nil)
+
+	type args struct {
+		endpoint string
+		studio   *models.ScrapedStudio
+	}
+	tests := []struct {
+		name    string
+		args    args
+		want    *int64
+		wantErr bool
+	}{
+		{
+			"simple",
+			args{
+				emptyEndpoint,
+				&models.ScrapedStudio{
+					Name: validName,
+				},
+			},
+			&createdID64,
+			false,
+		},
+		{
+			"error creating",
+			args{
+				emptyEndpoint,
+				&models.ScrapedStudio{
+					Name: invalidName,
+				},
+			},
+			nil,
+			true,
+		},
+		{
+			"valid stash id",
+			args{
+				validEndpoint,
+				&models.ScrapedStudio{
+					Name:         validName,
+					RemoteSiteID: &remoteSiteID,
+				},
+			},
+			&createdID64,
+			false,
+		},
+		{
+			"invalid stash id",
+			args{
+				invalidEndpoint,
+				&models.ScrapedStudio{
+					Name:         validName,
+					RemoteSiteID: &remoteSiteID,
+				},
+			},
+			nil,
+			true,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got, err := createMissingStudio(tt.args.endpoint, repo, tt.args.studio)
+			if (err != nil) != tt.wantErr {
+				t.Errorf("createMissingStudio() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+			if !reflect.DeepEqual(got, tt.want) {
+				t.Errorf("createMissingStudio() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+func Test_scrapedToStudioInput(t *testing.T) {
+	const name = "name"
+	const md5 = "b068931cc450442b63f5b3d276ea4297"
+	url := "url"
+
+	tests := []struct {
+		name   string
+		studio *models.ScrapedStudio
+		want   models.Studio
+	}{
+		{
+			"set all",
+			&models.ScrapedStudio{
+				Name: name,
+				URL:  &url,
+			},
+			models.Studio{
+				Name:     models.NullString(name),
+				Checksum: md5,
+				URL:      models.NullString(url),
+			},
+		},
+		{
+			"set none",
+			&models.ScrapedStudio{
+				Name: name,
+			},
+			models.Studio{
+				Name:     models.NullString(name),
+				Checksum: md5,
+			},
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got := scrapedToStudioInput(tt.studio)
+
+			// clear created/updated dates
+			got.CreatedAt = models.SQLiteTimestamp{}
+			got.UpdatedAt = got.CreatedAt
+
+			if !reflect.DeepEqual(got, tt.want) {
+				t.Errorf("scrapedToStudioInput() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}

pkg/manager/config/config.go

@@ -144,6 +144,11 @@ const (
 const HandyKey = "handy_key"
 const FunscriptOffset = "funscript_offset"
 
+// Default settings
+const (
+	DefaultIdentifySettings = "defaults.identify_task"
+)
+
 // Security
 const TrustedProxies = "trusted_proxies"
 const dangerousAllowPublicWithoutAuth = "dangerous_allow_public_without_auth"
@@ -476,10 +481,10 @@ func (i *Instance) GetScraperExcludeTagPatterns() []string {
 	return ret
 }
 
-func (i *Instance) GetStashBoxes() []*models.StashBox {
+func (i *Instance) GetStashBoxes() models.StashBoxes {
 	i.RLock()
 	defer i.RUnlock()
-	var boxes []*models.StashBox
+	var boxes models.StashBoxes
 	if err := viper.UnmarshalKey(StashBoxes, &boxes); err != nil {
 		logger.Warnf("error in unmarshalkey: %v", err)
 	}
@@ -869,10 +874,30 @@ func (i *Instance) GetHandyKey() string {
 }
 
 func (i *Instance) GetFunscriptOffset() int {
+	i.Lock()
+	defer i.Unlock()
 	viper.SetDefault(FunscriptOffset, 0)
 	return viper.GetInt(FunscriptOffset)
 }
 
+// GetDefaultIdentifySettings returns the default Identify task settings.
+// Returns nil if the settings could not be unmarshalled, or if it
+// has not been set.
+func (i *Instance) GetDefaultIdentifySettings() *models.IdentifyMetadataTaskOptions {
+	i.RLock()
+	defer i.RUnlock()
+
+	if viper.IsSet(DefaultIdentifySettings) {
+		var ret models.IdentifyMetadataTaskOptions
+		if err := viper.UnmarshalKey(DefaultIdentifySettings, &ret); err != nil {
+			return nil
+		}
+		return &ret
+	}
+
+	return nil
+}
+
 // GetTrustedProxies returns a comma separated list of ip addresses that should allow proxying.
 // When empty, allow from any private network
 func (i *Instance) GetTrustedProxies() []string {

pkg/manager/config/config_concurrency_test.go

@@ -92,6 +92,8 @@ func TestConcurrentConfigAccess(t *testing.T) {
 				i.Set(LogLevel, i.GetLogLevel())
 				i.Set(LogAccess, i.GetLogAccess())
 				i.Set(MaxUploadSize, i.GetMaxUploadSize())
+				i.Set(FunscriptOffset, i.GetFunscriptOffset())
+				i.Set(DefaultIdentifySettings, i.GetDefaultIdentifySettings())
 			}
 			wg.Done()
 		}(k)

pkg/manager/scene_screenshot.go

@@ -1,61 +0,0 @@
-package manager
-
-import (
-	"bytes"
-	"image"
-	"image/jpeg"
-	"os"
-
-	"github.com/disintegration/imaging"
-
-	// needed to decode other image formats
-	_ "image/gif"
-	_ "image/png"
-)
-
-func writeImage(path string, imageData []byte) error {
-	f, err := os.Create(path)
-	if err != nil {
-		return err
-	}
-	defer f.Close()
-
-	_, err = f.Write(imageData)
-	return err
-}
-
-func writeThumbnail(path string, thumbnail image.Image) error {
-	f, err := os.Create(path)
-	if err != nil {
-		return err
-	}
-	defer f.Close()
-
-	return jpeg.Encode(f, thumbnail, nil)
-}
-
-func SetSceneScreenshot(checksum string, imageData []byte) error {
-	thumbPath := instance.Paths.Scene.GetThumbnailScreenshotPath(checksum)
-	normalPath := instance.Paths.Scene.GetScreenshotPath(checksum)
-
-	img, _, err := image.Decode(bytes.NewReader(imageData))
-	if err != nil {
-		return err
-	}
-
-	// resize to 320 width maintaining aspect ratio, for the thumbnail
-	const width = 320
-	origWidth := img.Bounds().Max.X
-	origHeight := img.Bounds().Max.Y
-	height := width / origWidth * origHeight
-
-	thumbnail := imaging.Resize(img, width, height, imaging.Lanczos)
-	err = writeThumbnail(thumbPath, thumbnail)
-	if err != nil {
-		return err
-	}
-
-	err = writeImage(normalPath, imageData)
-
-	return err
-}

pkg/manager/task_autotag.go

@@ -326,28 +326,7 @@ type autoTagFilesTask struct {
 }
 
 func (t *autoTagFilesTask) makeSceneFilter() *models.SceneFilterType {
-	ret := &models.SceneFilterType{}
-	or := ret
-	sep := string(filepath.Separator)
-
-	for _, p := range t.paths {
-		if !strings.HasSuffix(p, sep) {
-			p += sep
-		}
-
-		if ret.Path == nil {
-			or = ret
-		} else {
-			newOr := &models.SceneFilterType{}
-			or.Or = newOr
-			or = newOr
-		}
-
-		or.Path = &models.StringCriterionInput{
-			Modifier: models.CriterionModifierEquals,
-			Value:    p + "%",
-		}
-	}
+	ret := scene.FilterFromPaths(t.paths)
 
 	organized := false
 	ret.Organized = &organized

pkg/manager/task_generate_screenshot.go

@@ -9,6 +9,7 @@ import (
 
 	"github.com/stashapp/stash/pkg/logger"
 	"github.com/stashapp/stash/pkg/models"
+	"github.com/stashapp/stash/pkg/scene"
 )
 
 type GenerateScreenshotTask struct {
@@ -66,7 +67,7 @@ func (t *GenerateScreenshotTask) Start(ctx context.Context) {
 			UpdatedAt: &models.SQLiteTimestamp{Timestamp: updatedTime},
 		}
 
-		if err := SetSceneScreenshot(checksum, coverImageData); err != nil {
+		if err := scene.SetScreenshot(instance.Paths, checksum, coverImageData); err != nil {
 			return fmt.Errorf("error writing screenshot: %v", err)
 		}
 

pkg/manager/task_identify.go

@@ -0,0 +1,244 @@
+package manager
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"strconv"
+
+	"github.com/stashapp/stash/pkg/identify"
+	"github.com/stashapp/stash/pkg/job"
+	"github.com/stashapp/stash/pkg/logger"
+	"github.com/stashapp/stash/pkg/models"
+	"github.com/stashapp/stash/pkg/scene"
+	"github.com/stashapp/stash/pkg/scraper"
+	"github.com/stashapp/stash/pkg/scraper/stashbox"
+	"github.com/stashapp/stash/pkg/utils"
+)
+
+var ErrInput = errors.New("invalid request input")
+
+type IdentifyJob struct {
+	txnManager       models.TransactionManager
+	postHookExecutor identify.SceneUpdatePostHookExecutor
+	input            models.IdentifyMetadataInput
+
+	stashBoxes models.StashBoxes
+	progress   *job.Progress
+}
+
+func CreateIdentifyJob(input models.IdentifyMetadataInput) *IdentifyJob {
+	return &IdentifyJob{
+		txnManager:       instance.TxnManager,
+		postHookExecutor: instance.PluginCache,
+		input:            input,
+		stashBoxes:       instance.Config.GetStashBoxes(),
+	}
+}
+
+func (j *IdentifyJob) Execute(ctx context.Context, progress *job.Progress) {
+	j.progress = progress
+
+	// if no sources provided - just return
+	if len(j.input.Sources) == 0 {
+		return
+	}
+
+	sources, err := j.getSources()
+	if err != nil {
+		logger.Error(err)
+		return
+	}
+
+	// if scene ids provided, use those
+	// otherwise, batch query for all scenes - ordering by path
+	if err := j.txnManager.WithReadTxn(ctx, func(r models.ReaderRepository) error {
+		if len(j.input.SceneIDs) == 0 {
+			return j.identifyAllScenes(ctx, r, sources)
+		}
+
+		sceneIDs, err := utils.StringSliceToIntSlice(j.input.SceneIDs)
+		if err != nil {
+			return fmt.Errorf("invalid scene IDs: %w", err)
+		}
+
+		progress.SetTotal(len(sceneIDs))
+		for _, id := range sceneIDs {
+			if job.IsCancelled(ctx) {
+				break
+			}
+
+			// find the scene
+			var err error
+			scene, err := r.Scene().Find(id)
+			if err != nil {
+				return fmt.Errorf("error finding scene with id %d: %w", id, err)
+			}
+
+			if scene == nil {
+				return fmt.Errorf("%w: scene with id %d", models.ErrNotFound, id)
+			}
+
+			j.identifyScene(ctx, scene, sources)
+		}
+
+		return nil
+	}); err != nil {
+		logger.Errorf("Error encountered while identifying scenes: %v", err)
+	}
+}
+
+func (j *IdentifyJob) identifyAllScenes(ctx context.Context, r models.ReaderRepository, sources []identify.ScraperSource) error {
+	// exclude organised
+	organised := false
+	sceneFilter := scene.FilterFromPaths(j.input.Paths)
+	sceneFilter.Organized = &organised
+
+	sort := "path"
+	findFilter := &models.FindFilterType{
+		Sort: &sort,
+	}
+
+	// get the count
+	pp := 0
+	findFilter.PerPage = &pp
+	countResult, err := r.Scene().Query(models.SceneQueryOptions{
+		QueryOptions: models.QueryOptions{
+			FindFilter: findFilter,
+			Count:      true,
+		},
+		SceneFilter: sceneFilter,
+	})
+	if err != nil {
+		return fmt.Errorf("error getting scene count: %w", err)
+	}
+
+	j.progress.SetTotal(countResult.Count)
+
+	return scene.BatchProcess(ctx, r.Scene(), sceneFilter, findFilter, func(scene *models.Scene) error {
+		if job.IsCancelled(ctx) {
+			return nil
+		}
+
+		j.identifyScene(ctx, scene, sources)
+		return nil
+	})
+}
+
+func (j *IdentifyJob) identifyScene(ctx context.Context, s *models.Scene, sources []identify.ScraperSource) {
+	if job.IsCancelled(ctx) {
+		return
+	}
+
+	if err := j.txnManager.WithTxn(context.TODO(), func(r models.Repository) error {
+		var taskError error
+		j.progress.ExecuteTask("Identifying "+s.Path, func() {
+			task := identify.SceneIdentifier{
+				DefaultOptions: j.input.Options,
+				Sources:        sources,
+				ScreenshotSetter: &scene.PathsScreenshotSetter{
+					Paths:               instance.Paths,
+					FileNamingAlgorithm: instance.Config.GetVideoFileNamingAlgorithm(),
+				},
+				SceneUpdatePostHookExecutor: j.postHookExecutor,
+			}
+
+			taskError = task.Identify(ctx, r, s)
+		})
+
+		return taskError
+	}); err != nil {
+		logger.Errorf("Error encountered identifying %s: %v", s.Path, err)
+	}
+
+	j.progress.Increment()
+}
+
+func (j *IdentifyJob) getSources() ([]identify.ScraperSource, error) {
+	var ret []identify.ScraperSource
+	for _, source := range j.input.Sources {
+		// get scraper source
+		stashBox, err := j.getStashBox(source.Source)
+		if err != nil {
+			return nil, err
+		}
+
+		var src identify.ScraperSource
+		if stashBox != nil {
+			src = identify.ScraperSource{
+				Name: "stash-box: " + stashBox.Endpoint,
+				Scraper: stashboxSource{
+					stashbox.NewClient(*stashBox, j.txnManager),
+					stashBox.Endpoint,
+				},
+				RemoteSite: stashBox.Endpoint,
+			}
+		} else {
+			scraperID := *source.Source.ScraperID
+			s := instance.ScraperCache.GetScraper(scraperID)
+			if s == nil {
+				return nil, fmt.Errorf("%w: scraper with id %q", models.ErrNotFound, scraperID)
+			}
+			src = identify.ScraperSource{
+				Name: s.Name,
+				Scraper: scraperSource{
+					cache:     instance.ScraperCache,
+					scraperID: scraperID,
+				},
+			}
+		}
+
+		src.Options = source.Options
+		ret = append(ret, src)
+	}
+
+	return ret, nil
+}
+
+func (j *IdentifyJob) getStashBox(src *models.ScraperSourceInput) (*models.StashBox, error) {
+	if src.ScraperID != nil {
+		return nil, nil
+	}
+
+	// must be stash-box
+	if src.StashBoxIndex == nil && src.StashBoxEndpoint == nil {
+		return nil, fmt.Errorf("%w: stash_box_index or stash_box_endpoint or scraper_id must be set", ErrInput)
+	}
+
+	return j.stashBoxes.ResolveStashBox(*src)
+}
+
+type stashboxSource struct {
+	*stashbox.Client
+	endpoint string
+}
+
+func (s stashboxSource) ScrapeScene(sceneID int) (*models.ScrapedScene, error) {
+	results, err := s.FindStashBoxScenesByFingerprintsFlat([]string{strconv.Itoa(sceneID)})
+	if err != nil {
+		return nil, fmt.Errorf("error querying stash-box using scene ID %d: %w", sceneID, err)
+	}
+
+	if len(results) > 0 {
+		return results[0], nil
+	}
+
+	return nil, nil
+}
+
+func (s stashboxSource) String() string {
+	return fmt.Sprintf("stash-box %s", s.endpoint)
+}
+
+type scraperSource struct {
+	cache     *scraper.Cache
+	scraperID string
+}
+
+func (s scraperSource) ScrapeScene(sceneID int) (*models.ScrapedScene, error) {
+	return s.cache.ScrapeScene(s.scraperID, sceneID)
+}
+
+func (s scraperSource) String() string {
+	return fmt.Sprintf("scraper %s", s.scraperID)
+}

pkg/models/errors.go

@@ -0,0 +1,5 @@
+package models
+
+import "errors"
+
+var ErrNotFound = errors.New("not found")

pkg/models/mocks/transaction.go

@@ -7,16 +7,16 @@ import (
 )
 
 type TransactionManager struct {
-	gallery     models.GalleryReaderWriter
-	image       models.ImageReaderWriter
-	movie       models.MovieReaderWriter
-	performer   models.PerformerReaderWriter
-	scene       models.SceneReaderWriter
-	sceneMarker models.SceneMarkerReaderWriter
-	scrapedItem models.ScrapedItemReaderWriter
-	studio      models.StudioReaderWriter
-	tag         models.TagReaderWriter
-	savedFilter models.SavedFilterReaderWriter
+	gallery     *GalleryReaderWriter
+	image       *ImageReaderWriter
+	movie       *MovieReaderWriter
+	performer   *PerformerReaderWriter
+	scene       *SceneReaderWriter
+	sceneMarker *SceneMarkerReaderWriter
+	scrapedItem *ScrapedItemReaderWriter
+	studio      *StudioReaderWriter
+	tag         *TagReaderWriter
+	savedFilter *SavedFilterReaderWriter
 }
 
 func NewTransactionManager() *TransactionManager {
@@ -38,90 +38,130 @@ func (t *TransactionManager) WithTxn(ctx context.Context, fn func(r models.Repos
 	return fn(t)
 }
 
-func (t *TransactionManager) Gallery() models.GalleryReaderWriter {
+func (t *TransactionManager) GalleryMock() *GalleryReaderWriter {
 	return t.gallery
 }
 
-func (t *TransactionManager) Image() models.ImageReaderWriter {
+func (t *TransactionManager) ImageMock() *ImageReaderWriter {
 	return t.image
 }
 
-func (t *TransactionManager) Movie() models.MovieReaderWriter {
+func (t *TransactionManager) MovieMock() *MovieReaderWriter {
 	return t.movie
 }
 
-func (t *TransactionManager) Performer() models.PerformerReaderWriter {
+func (t *TransactionManager) PerformerMock() *PerformerReaderWriter {
 	return t.performer
 }
 
-func (t *TransactionManager) SceneMarker() models.SceneMarkerReaderWriter {
+func (t *TransactionManager) SceneMarkerMock() *SceneMarkerReaderWriter {
 	return t.sceneMarker
 }
 
-func (t *TransactionManager) Scene() models.SceneReaderWriter {
+func (t *TransactionManager) SceneMock() *SceneReaderWriter {
 	return t.scene
 }
 
-func (t *TransactionManager) ScrapedItem() models.ScrapedItemReaderWriter {
+func (t *TransactionManager) ScrapedItemMock() *ScrapedItemReaderWriter {
 	return t.scrapedItem
 }
 
-func (t *TransactionManager) Studio() models.StudioReaderWriter {
+func (t *TransactionManager) StudioMock() *StudioReaderWriter {
 	return t.studio
 }
 
-func (t *TransactionManager) Tag() models.TagReaderWriter {
+func (t *TransactionManager) TagMock() *TagReaderWriter {
 	return t.tag
 }
 
-func (t *TransactionManager) SavedFilter() models.SavedFilterReaderWriter {
+func (t *TransactionManager) SavedFilterMock() *SavedFilterReaderWriter {
 	return t.savedFilter
 }
 
+func (t *TransactionManager) Gallery() models.GalleryReaderWriter {
+	return t.GalleryMock()
+}
+
+func (t *TransactionManager) Image() models.ImageReaderWriter {
+	return t.ImageMock()
+}
+
+func (t *TransactionManager) Movie() models.MovieReaderWriter {
+	return t.MovieMock()
+}
+
+func (t *TransactionManager) Performer() models.PerformerReaderWriter {
+	return t.PerformerMock()
+}
+
+func (t *TransactionManager) SceneMarker() models.SceneMarkerReaderWriter {
+	return t.SceneMarkerMock()
+}
+
+func (t *TransactionManager) Scene() models.SceneReaderWriter {
+	return t.SceneMock()
+}
+
+func (t *TransactionManager) ScrapedItem() models.ScrapedItemReaderWriter {
+	return t.ScrapedItemMock()
+}
+
+func (t *TransactionManager) Studio() models.StudioReaderWriter {
+	return t.StudioMock()
+}
+
+func (t *TransactionManager) Tag() models.TagReaderWriter {
+	return t.TagMock()
+}
+
+func (t *TransactionManager) SavedFilter() models.SavedFilterReaderWriter {
+	return t.SavedFilterMock()
+}
+
 type ReadTransaction struct {
-	t *TransactionManager
+	*TransactionManager
 }
 
 func (t *TransactionManager) WithReadTxn(ctx context.Context, fn func(r models.ReaderRepository) error) error {
-	return fn(&ReadTransaction{t: t})
+	return fn(&ReadTransaction{t})
 }
 
 func (r *ReadTransaction) Gallery() models.GalleryReader {
-	return r.t.gallery
+	return r.GalleryMock()
 }
 
 func (r *ReadTransaction) Image() models.ImageReader {
-	return r.t.image
+	return r.ImageMock()
 }
 
 func (r *ReadTransaction) Movie() models.MovieReader {
-	return r.t.movie
+	return r.MovieMock()
 }
 
 func (r *ReadTransaction) Performer() models.PerformerReader {
-	return r.t.performer
+	return r.PerformerMock()
 }
 
 func (r *ReadTransaction) SceneMarker() models.SceneMarkerReader {
-	return r.t.sceneMarker
+	return r.SceneMarkerMock()
 }
 
 func (r *ReadTransaction) Scene() models.SceneReader {
-	return r.t.scene
+	return r.SceneMock()
 }
 
 func (r *ReadTransaction) ScrapedItem() models.ScrapedItemReader {
-	return r.t.scrapedItem
+	return r.ScrapedItemMock()
 }
 
 func (r *ReadTransaction) Studio() models.StudioReader {
-	return r.t.studio
+	return r.StudioMock()
 }
 
 func (r *ReadTransaction) Tag() models.TagReader {
-	return r.t.tag
+	return r.TagMock()
 }
 
 func (r *ReadTransaction) SavedFilter() models.SavedFilterReader {
-	return r.t.savedFilter
+	return r.SavedFilterMock()
 }

pkg/models/model_joins.go

@@ -12,3 +12,10 @@ type StashID struct {
 	StashID  string `db:"stash_id" json:"stash_id"`
 	Endpoint string `db:"endpoint" json:"endpoint"`
 }
+
+func (s StashID) StashIDInput() StashIDInput {
+	return StashIDInput{
+		Endpoint: s.Endpoint,
+		StashID:  s.StashID,
+	}
+}

pkg/models/model_scene.go

@@ -3,6 +3,7 @@ package models
 import (
 	"database/sql"
 	"path/filepath"
+	"strconv"
 	"time"
 )
 
@@ -119,6 +120,29 @@ type ScenePartial struct {
 	Interactive *bool                `db:"interactive" json:"interactive"`
 }
 
+// UpdateInput constructs a SceneUpdateInput using the populated fields in the ScenePartial object.
+func (s ScenePartial) UpdateInput() SceneUpdateInput {
+	boolPtrCopy := func(v *bool) *bool {
+		if v == nil {
+			return nil
+		}
+
+		vv := *v
+		return &vv
+	}
+
+	return SceneUpdateInput{
+		ID:        strconv.Itoa(s.ID),
+		Title:     nullStringPtrToStringPtr(s.Title),
+		Details:   nullStringPtrToStringPtr(s.Details),
+		URL:       nullStringPtrToStringPtr(s.URL),
+		Date:      s.Date.StringPtr(),
+		Rating:    nullInt64PtrToIntPtr(s.Rating),
+		Organized: boolPtrCopy(s.Organized),
+		StudioID:  nullInt64PtrToStringPtr(s.StudioID),
+	}
+}
+
 func (s *ScenePartial) SetFile(f File) {
 	path := f.Path
 	s.Path = &path

pkg/models/model_scene_test.go

@@ -0,0 +1,80 @@
+package models
+
+import (
+	"database/sql"
+	"reflect"
+	"testing"
+)
+
+func TestScenePartial_UpdateInput(t *testing.T) {
+	const (
+		id    = 1
+		idStr = "1"
+	)
+
+	var (
+		title       = "title"
+		details     = "details"
+		url         = "url"
+		date        = "2001-02-03"
+		rating      = 4
+		organized   = true
+		studioID    = 2
+		studioIDStr = "2"
+	)
+
+	tests := []struct {
+		name string
+		s    ScenePartial
+		want SceneUpdateInput
+	}{
+		{
+			"full",
+			ScenePartial{
+				ID:      id,
+				Title:   NullStringPtr(title),
+				Details: NullStringPtr(details),
+				URL:     NullStringPtr(url),
+				Date: &SQLiteDate{
+					String: date,
+					Valid:  true,
+				},
+				Rating: &sql.NullInt64{
+					Int64: int64(rating),
+					Valid: true,
+				},
+				Organized: &organized,
+				StudioID: &sql.NullInt64{
+					Int64: int64(studioID),
+					Valid: true,
+				},
+			},
+			SceneUpdateInput{
+				ID:        idStr,
+				Title:     &title,
+				Details:   &details,
+				URL:       &url,
+				Date:      &date,
+				Rating:    &rating,
+				Organized: &organized,
+				StudioID:  &studioIDStr,
+			},
+		},
+		{
+			"empty",
+			ScenePartial{
+				ID: id,
+			},
+			SceneUpdateInput{
+				ID: idStr,
+			},
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if got := tt.s.UpdateInput(); !reflect.DeepEqual(got, tt.want) {
+				t.Errorf("ScenePartial.UpdateInput() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}

pkg/models/scraped.go

@@ -1,5 +1,9 @@
 package models
 
+import "errors"
+
+var ErrScraperSource = errors.New("invalid ScraperSource")
+
 type ScrapedItemReader interface {
 	All() ([]*ScrapedItem, error)
 }

pkg/models/sql.go

@@ -1,6 +1,9 @@
 package models
 
-import "database/sql"
+import (
+	"database/sql"
+	"strconv"
+)
 
 func NullString(v string) sql.NullString {
 	return sql.NullString{
@@ -9,9 +12,43 @@ func NullString(v string) sql.NullString {
 	}
 }
 
+func NullStringPtr(v string) *sql.NullString {
+	return &sql.NullString{
+		String: v,
+		Valid:  true,
+	}
+}
+
 func NullInt64(v int64) sql.NullInt64 {
 	return sql.NullInt64{
 		Int64: v,
 		Valid: true,
 	}
 }
+
+func nullStringPtrToStringPtr(v *sql.NullString) *string {
+	if v == nil || !v.Valid {
+		return nil
+	}
+
+	vv := v.String
+	return &vv
+}
+
+func nullInt64PtrToIntPtr(v *sql.NullInt64) *int {
+	if v == nil || !v.Valid {
+		return nil
+	}
+
+	vv := int(v.Int64)
+	return &vv
+}
+
+func nullInt64PtrToStringPtr(v *sql.NullInt64) *string {
+	if v == nil || !v.Valid {
+		return nil
+	}
+
+	vv := strconv.FormatInt(v.Int64, 10)
+	return &vv
+}

pkg/models/sqlite_date.go

@@ -44,3 +44,12 @@ func (t SQLiteDate) Value() (driver.Value, error) {
 	}
 	return result, nil
 }
+
+func (t *SQLiteDate) StringPtr() *string {
+	if t == nil || !t.Valid {
+		return nil
+	}
+
+	vv := t.String
+	return &vv
+}

pkg/models/stash_box.go

@@ -0,0 +1,39 @@
+package models
+
+import (
+	"fmt"
+	"strings"
+)
+
+type StashBoxes []*StashBox
+
+func (sb StashBoxes) ResolveStashBox(source ScraperSourceInput) (*StashBox, error) {
+	if source.StashBoxIndex != nil {
+		index := source.StashBoxIndex
+		if *index < 0 || *index >= len(sb) {
+			return nil, fmt.Errorf("%w: invalid stash_box_index: %d", ErrScraperSource, index)
+		}
+
+		return sb[*index], nil
+	}
+
+	if source.StashBoxEndpoint != nil {
+		var ret *StashBox
+		endpoint := *source.StashBoxEndpoint
+		for _, b := range sb {
+			if strings.EqualFold(endpoint, b.Endpoint) {
+				ret = b
+			}
+		}
+
+		if ret == nil {
+			return nil, fmt.Errorf(`%w: stash-box with endpoint "%s"`, ErrNotFound, endpoint)
+		}
+
+		return ret, nil
+	}
+
+	// neither stash-box inputs were provided, so assume it is a scraper
+
+	return nil, nil
+}

pkg/plugin/plugins.go

@@ -13,6 +13,7 @@ import (
 	"net/http"
 	"os"
 	"path/filepath"
+	"strconv"
 
 	"github.com/stashapp/stash/pkg/logger"
 	"github.com/stashapp/stash/pkg/manager/config"
@@ -179,6 +180,15 @@ func (c Cache) ExecutePostHooks(ctx context.Context, id int, hookType HookTrigge
 	}
 }
 
+func (c Cache) ExecuteSceneUpdatePostHooks(ctx context.Context, input models.SceneUpdateInput, inputFields []string) {
+	id, err := strconv.Atoi(input.ID)
+	if err != nil {
+		logger.Errorf("error converting id in SceneUpdatePostHooks: %v", err)
+		return
+	}
+	c.ExecutePostHooks(ctx, id, SceneUpdatePost, input, inputFields)
+}
+
 func (c Cache) executePostHooks(ctx context.Context, hookType HookTriggerEnum, hookContext common.HookContext) error {
 	visitedPlugins := session.GetVisitedPlugins(ctx)
 

pkg/scene/export_test.go

@@ -85,7 +85,7 @@ var names = []string{
 
 var imageBytes = []byte("imageBytes")
 
-const image = "aW1hZ2VCeXRlcw=="
+const imageBase64 = "aW1hZ2VCeXRlcw=="
 
 var (
 	createTime = time.Date(2001, 01, 01, 0, 0, 0, 0, time.UTC)
@@ -198,7 +198,7 @@ type basicTestScenario struct {
 var scenarios = []basicTestScenario{
 	{
 		createFullScene(sceneID),
-		createFullJSONScene(image),
+		createFullJSONScene(imageBase64),
 		false,
 	},
 	{

pkg/scene/import_test.go

@@ -75,7 +75,7 @@ func TestImporterPreImport(t *testing.T) {
 	err := i.PreImport()
 	assert.NotNil(t, err)
 
-	i.Input.Cover = image
+	i.Input.Cover = imageBase64
 
 	err = i.PreImport()
 	assert.Nil(t, err)

pkg/scene/query.go

@@ -1,6 +1,14 @@
 package scene
 
-import "github.com/stashapp/stash/pkg/models"
+import (
+	"context"
+	"fmt"
+	"path/filepath"
+	"strings"
+
+	"github.com/stashapp/stash/pkg/job"
+	"github.com/stashapp/stash/pkg/models"
+)
 
 type Queryer interface {
 	Query(options models.SceneQueryOptions) (*models.SceneQueryResult, error)
@@ -48,3 +56,70 @@ func Query(qb Queryer, sceneFilter *models.SceneFilterType, findFilter *models.F
 
 	return scenes, nil
 }
+
+func BatchProcess(ctx context.Context, reader models.SceneReader, sceneFilter *models.SceneFilterType, findFilter *models.FindFilterType, fn func(scene *models.Scene) error) error {
+	const batchSize = 1000
+
+	if findFilter == nil {
+		findFilter = &models.FindFilterType{}
+	}
+
+	page := 1
+	perPage := batchSize
+	findFilter.Page = &page
+	findFilter.PerPage = &perPage
+
+	for more := true; more; {
+		if job.IsCancelled(ctx) {
+			return nil
+		}
+
+		scenes, err := Query(reader, sceneFilter, findFilter)
+		if err != nil {
+			return fmt.Errorf("error querying for scenes: %w", err)
+		}
+
+		for _, scene := range scenes {
+			if err := fn(scene); err != nil {
+				return err
+			}
+		}
+
+		if len(scenes) != batchSize {
+			more = false
+		} else {
+			*findFilter.Page++
+		}
+	}
+
+	return nil
+}
+
+// FilterFromPaths creates a SceneFilterType that filters using the provided
+// paths.
+func FilterFromPaths(paths []string) *models.SceneFilterType {
+	ret := &models.SceneFilterType{}
+	or := ret
+	sep := string(filepath.Separator)
+
+	for _, p := range paths {
+		if !strings.HasSuffix(p, sep) {
+			p += sep
+		}
+
+		if ret.Path == nil {
+			or = ret
+		} else {
+			newOr := &models.SceneFilterType{}
+			or.Or = newOr
+			or = newOr
+		}
+
+		or.Path = &models.StringCriterionInput{
+			Modifier: models.CriterionModifierEquals,
+			Value:    p + "%",
+		}
+	}
+
+	return ret
+}

pkg/scene/screenshot.go

@@ -1,8 +1,21 @@
 package scene
 
 import (
+	"bytes"
+	"image"
+	"image/jpeg"
+	"os"
+
 	"github.com/stashapp/stash/pkg/ffmpeg"
 	"github.com/stashapp/stash/pkg/logger"
+	"github.com/stashapp/stash/pkg/manager/paths"
+	"github.com/stashapp/stash/pkg/models"
+
+	"github.com/disintegration/imaging"
+
+	// needed to decode other image formats
+	_ "image/gif"
+	_ "image/png"
 )
 
 type screenshotter interface {
@@ -21,3 +34,64 @@ func makeScreenshot(encoder screenshotter, probeResult ffmpeg.VideoFile, outputP
 		logger.Warnf("[encoder] failure to generate screenshot: %v", err)
 	}
 }
+
+type ScreenshotSetter interface {
+	SetScreenshot(scene *models.Scene, imageData []byte) error
+}
+
+type PathsScreenshotSetter struct {
+	Paths               *paths.Paths
+	FileNamingAlgorithm models.HashAlgorithm
+}
+
+func (ss *PathsScreenshotSetter) SetScreenshot(scene *models.Scene, imageData []byte) error {
+	checksum := scene.GetHash(ss.FileNamingAlgorithm)
+	return SetScreenshot(ss.Paths, checksum, imageData)
+}
+
+func writeImage(path string, imageData []byte) error {
+	f, err := os.Create(path)
+	if err != nil {
+		return err
+	}
+	defer f.Close()
+
+	_, err = f.Write(imageData)
+	return err
+}
+
+func writeThumbnail(path string, thumbnail image.Image) error {
+	f, err := os.Create(path)
+	if err != nil {
+		return err
+	}
+	defer f.Close()
+
+	return jpeg.Encode(f, thumbnail, nil)
+}
+
+func SetScreenshot(paths *paths.Paths, checksum string, imageData []byte) error {
+	thumbPath := paths.Scene.GetThumbnailScreenshotPath(checksum)
+	normalPath := paths.Scene.GetScreenshotPath(checksum)
+
+	img, _, err := image.Decode(bytes.NewReader(imageData))
+	if err != nil {
+		return err
+	}
+
+	// resize to 320 width maintaining aspect ratio, for the thumbnail
+	const width = 320
+	origWidth := img.Bounds().Max.X
+	origHeight := img.Bounds().Max.Y
+	height := width / origWidth * origHeight
+
+	thumbnail := imaging.Resize(img, width, height, imaging.Lanczos)
+	err = writeThumbnail(thumbPath, thumbnail)
+	if err != nil {
+		return err
+	}
+
+	err = writeImage(normalPath, imageData)
+
+	return err
+}

pkg/scene/update.go

@@ -2,11 +2,127 @@ package scene
 
 import (
 	"database/sql"
+	"errors"
+	"fmt"
+	"time"
 
 	"github.com/stashapp/stash/pkg/models"
 	"github.com/stashapp/stash/pkg/utils"
 )
 
+var ErrEmptyUpdater = errors.New("no fields have been set")
+
+// UpdateSet is used to update a scene and its relationships.
+type UpdateSet struct {
+	ID int
+
+	Partial models.ScenePartial
+
+	// in future these could be moved into a separate struct and reused
+	// for a Creator struct
+
+	// Not set if nil. Set to []int{} to clear existing
+	PerformerIDs []int
+	// Not set if nil. Set to []int{} to clear existing
+	TagIDs []int
+	// Not set if nil. Set to []int{} to clear existing
+	StashIDs []models.StashID
+	// Not set if nil. Set to []byte{} to clear existing
+	CoverImage []byte
+}
+
+// IsEmpty returns true if there is nothing to update.
+func (u *UpdateSet) IsEmpty() bool {
+	withoutID := u.Partial
+	withoutID.ID = 0
+
+	return withoutID == models.ScenePartial{} &&
+		u.PerformerIDs == nil &&
+		u.TagIDs == nil &&
+		u.StashIDs == nil &&
+		u.CoverImage == nil
+}
+
+// Update updates a scene by updating the fields in the Partial field, then
+// updates non-nil relationships. Returns an error if there is no work to
+// be done.
+func (u *UpdateSet) Update(qb models.SceneWriter, screenshotSetter ScreenshotSetter) (*models.Scene, error) {
+	if u.IsEmpty() {
+		return nil, ErrEmptyUpdater
+	}
+
+	partial := u.Partial
+	partial.ID = u.ID
+	partial.UpdatedAt = &models.SQLiteTimestamp{
+		Timestamp: time.Now(),
+	}
+
+	ret, err := qb.Update(partial)
+	if err != nil {
+		return nil, fmt.Errorf("error updating scene: %w", err)
+	}
+
+	if u.PerformerIDs != nil {
+		if err := qb.UpdatePerformers(u.ID, u.PerformerIDs); err != nil {
+			return nil, fmt.Errorf("error updating scene performers: %w", err)
+		}
+	}
+
+	if u.TagIDs != nil {
+		if err := qb.UpdateTags(u.ID, u.TagIDs); err != nil {
+			return nil, fmt.Errorf("error updating scene tags: %w", err)
+		}
+	}
+
+	if u.StashIDs != nil {
+		if err := qb.UpdateStashIDs(u.ID, u.StashIDs); err != nil {
+			return nil, fmt.Errorf("error updating scene stash_ids: %w", err)
+		}
+	}
+
+	if u.CoverImage != nil {
+		if err := qb.UpdateCover(u.ID, u.CoverImage); err != nil {
+			return nil, fmt.Errorf("error updating scene cover: %w", err)
+		}
+
+		if err := screenshotSetter.SetScreenshot(ret, u.CoverImage); err != nil {
+			return nil, fmt.Errorf("error setting scene screenshot: %w", err)
+		}
+	}
+
+	return ret, nil
+}
+
+// UpdateInput converts the UpdateSet into SceneUpdateInput for hook firing purposes.
+func (u UpdateSet) UpdateInput() models.SceneUpdateInput {
+	// ensure the partial ID is set
+	u.Partial.ID = u.ID
+	ret := u.Partial.UpdateInput()
+
+	if u.PerformerIDs != nil {
+		ret.PerformerIds = utils.IntSliceToStringSlice(u.PerformerIDs)
+	}
+
+	if u.TagIDs != nil {
+		ret.TagIds = utils.IntSliceToStringSlice(u.TagIDs)
+	}
+
+	if u.StashIDs != nil {
+		for _, s := range u.StashIDs {
+			ss := s.StashIDInput()
+			ret.StashIds = append(ret.StashIds, &ss)
+		}
+	}
+
+	if u.CoverImage != nil {
+		// convert back to base64
+		data := utils.GetBase64StringFromData(u.CoverImage)
+		ret.CoverImage = &data
+	}
+
+	return ret
+}
+
 func UpdateFormat(qb models.SceneWriter, id int, format string) (*models.Scene, error) {
 	return qb.Update(models.ScenePartial{
 		ID: id,

pkg/scene/update_test.go

@@ -0,0 +1,337 @@
+package scene
+
+import (
+	"errors"
+	"strconv"
+	"testing"
+
+	"github.com/stashapp/stash/pkg/models"
+	"github.com/stashapp/stash/pkg/models/mocks"
+	"github.com/stashapp/stash/pkg/utils"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/mock"
+)
+
+func TestUpdater_IsEmpty(t *testing.T) {
+	organized := true
+	ids := []int{1}
+	stashIDs := []models.StashID{
+		{},
+	}
+	cover := []byte{1}
+
+	tests := []struct {
+		name string
+		u    *UpdateSet
+		want bool
+	}{
+		{
+			"empty",
+			&UpdateSet{},
+			true,
+		},
+		{
+			"id only",
+			&UpdateSet{
+				Partial: models.ScenePartial{
+					ID: 1,
+				},
+			},
+			true,
+		},
+		{
+			"partial set",
+			&UpdateSet{
+				Partial: models.ScenePartial{
+					Organized: &organized,
+				},
+			},
+			false,
+		},
+		{
+			"performer set",
+			&UpdateSet{
+				PerformerIDs: ids,
+			},
+			false,
+		},
+		{
+			"tags set",
+			&UpdateSet{
+				TagIDs: ids,
+			},
+			false,
+		},
+		{
+			"performer set",
+			&UpdateSet{
+				StashIDs: stashIDs,
+			},
+			false,
+		},
+		{
+			"cover set",
+			&UpdateSet{
+				CoverImage: cover,
+			},
+			false,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if got := tt.u.IsEmpty(); got != tt.want {
+				t.Errorf("Updater.IsEmpty() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}
+
+type mockScreenshotSetter struct{}
+
+func (s *mockScreenshotSetter) SetScreenshot(scene *models.Scene, imageData []byte) error {
+	return nil
+}
+
+func TestUpdater_Update(t *testing.T) {
+	const (
+		sceneID = iota + 1
+		badUpdateID
+		badPerformersID
+		badTagsID
+		badStashIDsID
+		badCoverID
+		performerID
+		tagID
+	)
+
+	performerIDs := []int{performerID}
+	tagIDs := []int{tagID}
+	stashID := "stashID"
+	endpoint := "endpoint"
+	stashIDs := []models.StashID{
+		{
+			StashID:  stashID,
+			Endpoint: endpoint,
+		},
+	}
+
+	title := "title"
+	cover := []byte("cover")
+
+	validScene := &models.Scene{}
+
+	updateErr := errors.New("error updating")
+
+	qb := mocks.SceneReaderWriter{}
+	qb.On("Update", mock.MatchedBy(func(s models.ScenePartial) bool {
+		return s.ID != badUpdateID
+	})).Return(validScene, nil)
+	qb.On("Update", mock.MatchedBy(func(s models.ScenePartial) bool {
+		return s.ID == badUpdateID
+	})).Return(nil, updateErr)
+
+	qb.On("UpdatePerformers", sceneID, performerIDs).Return(nil).Once()
+	qb.On("UpdateTags", sceneID, tagIDs).Return(nil).Once()
+	qb.On("UpdateStashIDs", sceneID, stashIDs).Return(nil).Once()
+	qb.On("UpdateCover", sceneID, cover).Return(nil).Once()
+
+	qb.On("UpdatePerformers", badPerformersID, performerIDs).Return(updateErr).Once()
+	qb.On("UpdateTags", badTagsID, tagIDs).Return(updateErr).Once()
+	qb.On("UpdateStashIDs", badStashIDsID, stashIDs).Return(updateErr).Once()
+	qb.On("UpdateCover", badCoverID, cover).Return(updateErr).Once()
+
+	tests := []struct {
+		name    string
+		u       *UpdateSet
+		wantNil bool
+		wantErr bool
+	}{
+		{
+			"empty",
+			&UpdateSet{
+				ID: sceneID,
+			},
+			true,
+			true,
+		},
+		{
+			"update all",
+			&UpdateSet{
+				ID:           sceneID,
+				PerformerIDs: performerIDs,
+				TagIDs:       tagIDs,
+				StashIDs: []models.StashID{
+					{
+						StashID:  stashID,
+						Endpoint: endpoint,
+					},
+				},
+				CoverImage: cover,
+			},
+			false,
+			false,
+		},
+		{
+			"update fields only",
+			&UpdateSet{
+				ID: sceneID,
+				Partial: models.ScenePartial{
+					Title: models.NullStringPtr(title),
+				},
+			},
+			false,
+			false,
+		},
+		{
+			"error updating scene",
+			&UpdateSet{
+				ID: badUpdateID,
+				Partial: models.ScenePartial{
+					Title: models.NullStringPtr(title),
+				},
+			},
+			true,
+			true,
+		},
+		{
+			"error updating performers",
+			&UpdateSet{
+				ID:           badPerformersID,
+				PerformerIDs: performerIDs,
+			},
+			true,
+			true,
+		},
+		{
+			"error updating tags",
+			&UpdateSet{
+				ID:     badTagsID,
+				TagIDs: tagIDs,
+			},
+			true,
+			true,
+		},
+		{
+			"error updating stash IDs",
+			&UpdateSet{
+				ID:       badStashIDsID,
+				StashIDs: stashIDs,
+			},
+			true,
+			true,
+		},
+		{
+			"error updating cover",
+			&UpdateSet{
+				ID:         badCoverID,
+				CoverImage: cover,
+			},
+			true,
+			true,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got, err := tt.u.Update(&qb, &mockScreenshotSetter{})
+			if (err != nil) != tt.wantErr {
+				t.Errorf("Updater.Update() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+			if (got == nil) != tt.wantNil {
+				t.Errorf("Updater.Update() = %v, want %v", got, tt.wantNil)
+			}
+		})
+	}
+
+	qb.AssertExpectations(t)
+}
+
+func TestUpdateSet_UpdateInput(t *testing.T) {
+	const (
+		sceneID = iota + 1
+		badUpdateID
+		badPerformersID
+		badTagsID
+		badStashIDsID
+		badCoverID
+		performerID
+		tagID
+	)
+
+	sceneIDStr := strconv.Itoa(sceneID)
+
+	performerIDs := []int{performerID}
+	performerIDStrs := utils.IntSliceToStringSlice(performerIDs)
+	tagIDs := []int{tagID}
+	tagIDStrs := utils.IntSliceToStringSlice(tagIDs)
+	stashID := "stashID"
+	endpoint := "endpoint"
+	stashIDs := []models.StashID{
+		{
+			StashID:  stashID,
+			Endpoint: endpoint,
+		},
+	}
+	stashIDInputs := []*models.StashIDInput{
+		{
+			StashID:  stashID,
+			Endpoint: endpoint,
+		},
+	}
+
+	title := "title"
+	cover := []byte("cover")
+	coverB64 := "Y292ZXI="
+
+	tests := []struct {
+		name string
+		u    UpdateSet
+		want models.SceneUpdateInput
+	}{
+		{
+			"empty",
+			UpdateSet{
+				ID: sceneID,
+			},
+			models.SceneUpdateInput{
+				ID: sceneIDStr,
+			},
+		},
+		{
+			"update all",
+			UpdateSet{
+				ID:           sceneID,
+				PerformerIDs: performerIDs,
+				TagIDs:       tagIDs,
+				StashIDs:     stashIDs,
+				CoverImage:   cover,
+			},
+			models.SceneUpdateInput{
+				ID:           sceneIDStr,
+				PerformerIds: performerIDStrs,
+				TagIds:       tagIDStrs,
+				StashIds:     stashIDInputs,
+				CoverImage:   &coverB64,
+			},
+		},
+		{
+			"update fields only",
+			UpdateSet{
+				ID: sceneID,
+				Partial: models.ScenePartial{
+					Title: models.NullStringPtr(title),
+				},
+			},
+			models.SceneUpdateInput{
+				ID:    sceneIDStr,
+				Title: &title,
+			},
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got := tt.u.UpdateInput()
+			assert.Equal(t, tt.want, got)
+		})
+	}
+}

pkg/scraper/scrapers.go

@@ -212,6 +212,16 @@ func (c Cache) ListMovieScrapers() []*models.Scraper {
 	return ret
 }
 
+// GetScraper returns the scraper matching the provided id.
+func (c Cache) GetScraper(scraperID string) *models.Scraper {
+	ret := c.findScraper(scraperID)
+	if ret != nil {
+		return ret.Spec
+	}
+
+	return nil
+}
+
 func (c Cache) findScraper(scraperID string) *scraper {
 	for _, s := range c.scrapers {
 		if s.ID == scraperID {

pkg/utils/collections.go

@@ -0,0 +1,60 @@
+package utils
+
+import "reflect"
+
+// SliceSame returns true if the two provided lists have the same elements,
+// regardless of order. Panics if either parameter is not a slice.
+func SliceSame(a, b interface{}) bool {
+	v1 := reflect.ValueOf(a)
+	v2 := reflect.ValueOf(b)
+
+	if (v1.IsValid() && v1.Kind() != reflect.Slice) || (v2.IsValid() && v2.Kind() != reflect.Slice) {
+		panic("not a slice")
+	}
+
+	v1Len := 0
+	v2Len := 0
+
+	v1Valid := v1.IsValid()
+	v2Valid := v2.IsValid()
+
+	if v1Valid {
+		v1Len = v1.Len()
+	}
+	if v2Valid {
+		v2Len = v2.Len()
+	}
+
+	if !v1Valid || !v2Valid {
+		return v1Len == v2Len
+	}
+
+	if v1Len != v2Len {
+		return false
+	}
+
+	if v1.Type() != v2.Type() {
+		return false
+	}
+
+	visited := make(map[int]bool)
+	for i := 0; i < v1.Len(); i++ {
+		found := false
+		for j := 0; j < v2.Len(); j++ {
+			if visited[j] {
+				continue
+			}
+			if reflect.DeepEqual(v1.Index(i).Interface(), v2.Index(j).Interface()) {
+				found = true
+				visited[j] = true
+				break
+			}
+		}
+
+		if !found {
+			return false
+		}
+	}
+
+	return true
+}

pkg/utils/collections_test.go

@@ -0,0 +1,92 @@
+package utils
+
+import "testing"
+
+func TestSliceSame(t *testing.T) {
+	objs := []struct {
+		a string
+		b int
+	}{
+		{"1", 2},
+		{"1", 2},
+		{"2", 1},
+	}
+
+	tests := []struct {
+		name string
+		a    interface{}
+		b    interface{}
+		want bool
+	}{
+		{"nil values", nil, nil, true},
+		{"empty", []int{}, []int{}, true},
+		{"nil and empty", nil, []int{}, true},
+		{
+			"different type",
+			[]string{"1"},
+			[]int{1},
+			false,
+		},
+		{
+			"different length",
+			[]int{1, 2, 3},
+			[]int{1, 2},
+			false,
+		},
+		{
+			"equal",
+			[]int{1, 2, 3, 4, 5},
+			[]int{1, 2, 3, 4, 5},
+			true,
+		},
+		{
+			"different order",
+			[]int{5, 4, 3, 2, 1},
+			[]int{1, 2, 3, 4, 5},
+			true,
+		},
+		{
+			"different",
+			[]int{5, 4, 3, 2, 6},
+			[]int{1, 2, 3, 4, 5},
+			false,
+		},
+		{
+			"same with duplicates",
+			[]int{1, 1, 2, 3, 4},
+			[]int{1, 2, 3, 4, 1},
+			true,
+		},
+		{
+			"subset",
+			[]int{1, 1, 2, 2, 3},
+			[]int{1, 2, 3, 4, 5},
+			false,
+		},
+		{
+			"superset",
+			[]int{1, 2, 3, 4, 5},
+			[]int{1, 1, 2, 2, 3},
+			false,
+		},
+		{
+			"structs equal",
+			objs[0:1],
+			objs[0:1],
+			true,
+		},
+		{
+			"structs not equal",
+			objs[0:2],
+			objs[1:3],
+			false,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if got := SliceSame(tt.a, tt.b); got != tt.want {
+				t.Errorf("SliceSame() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}

pkg/utils/int_collections.go

@@ -1,5 +1,7 @@
 package utils
 
+import "strconv"
+
 // IntIndex returns the first index of the provided int value in the provided
 // int slice. It returns -1 if it is not found.
 func IntIndex(vs []int, t int) int {
@@ -50,3 +52,13 @@ func IntExclude(vs []int, toExclude []int) []int {
 
 	return ret
 }
+
+// IntSliceToStringSlice converts a slice of ints to a slice of strings.
+func IntSliceToStringSlice(ss []int) []string {
+	ret := make([]string, len(ss))
+	for i, v := range ss {
+		ret[i] = strconv.Itoa(v)
+	}
+
+	return ret
+}

pkg/utils/reflect.go

@@ -0,0 +1,30 @@
+package utils
+
+import "reflect"
+
+// NotNilFields returns the matching tag values of fields from an object that are not nil.
+// Panics if the provided object is not a struct.
+func NotNilFields(subject interface{}, tag string) []string {
+	value := reflect.ValueOf(subject)
+	structType := value.Type()
+
+	if structType.Kind() != reflect.Struct {
+		panic("subject must be struct")
+	}
+
+	var ret []string
+
+	for i := 0; i < value.NumField(); i++ {
+		field := value.Field(i)
+
+		kind := field.Type().Kind()
+		if (kind == reflect.Ptr || kind == reflect.Slice) && !field.IsNil() {
+			tagValue := structType.Field(i).Tag.Get(tag)
+			if tagValue != "" {
+				ret = append(ret, tagValue)
+			}
+		}
+	}
+
+	return ret
+}

pkg/utils/reflect_test.go

@@ -0,0 +1,83 @@
+package utils
+
+import (
+	"reflect"
+	"testing"
+)
+
+func TestNotNilFields(t *testing.T) {
+	v := "value"
+	var zeroStr string
+
+	type testObject struct {
+		ptrField      *string `tag:"ptrField"`
+		noTagField    *string
+		otherTagField *string  `otherTag:"otherTagField"`
+		sliceField    []string `tag:"sliceField"`
+	}
+
+	type args struct {
+		subject interface{}
+		tag     string
+	}
+	tests := []struct {
+		name string
+		args args
+		want []string
+	}{
+		{
+			"basic",
+			args{
+				testObject{
+					ptrField:      &v,
+					noTagField:    &v,
+					otherTagField: &v,
+					sliceField:    []string{v},
+				},
+				"tag",
+			},
+			[]string{"ptrField", "sliceField"},
+		},
+		{
+			"empty",
+			args{
+				testObject{},
+				"tag",
+			},
+			nil,
+		},
+		{
+			"zero values",
+			args{
+				testObject{
+					ptrField:      &zeroStr,
+					noTagField:    &zeroStr,
+					otherTagField: &zeroStr,
+					sliceField:    []string{},
+				},
+				"tag",
+			},
+			[]string{"ptrField", "sliceField"},
+		},
+		{
+			"other tag",
+			args{
+				testObject{
+					ptrField:      &v,
+					noTagField:    &v,
+					otherTagField: &v,
+					sliceField:    []string{v},
+				},
+				"otherTag",
+			},
+			[]string{"otherTagField"},
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if got := NotNilFields(tt.args.subject, tt.args.tag); !reflect.DeepEqual(got, tt.want) {
+				t.Errorf("NotNilFields() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}

ui/v2.5/src/components/Changelog/versions/v0110.md

@@ -1,4 +1,5 @@
 ### ✨ New Features
+* Added Identify task to automatically identify scenes from stash-box/scraper sources. See manual entry for details. ([#1839](https://github.com/stashapp/stash/pull/1839))
 * Added support for matching scenes using perceptual hashes when querying stash-box. ([#1858](https://github.com/stashapp/stash/pull/1858))
 * Generalised Tagger view to support tagging using supported scene scrapers. ([#1812](https://github.com/stashapp/stash/pull/1812))
 * Added built-in `Auto Tag` scene scraper to match performers, studio and tags from filename - using AutoTag logic. ([#1817](https://github.com/stashapp/stash/pull/1817))

ui/v2.5/src/components/Dialogs/IdentifyDialog/FieldOptions.tsx

@@ -0,0 +1,338 @@
+import React, { useState, useEffect, useCallback } from "react";
+import { Form, Button, Table } from "react-bootstrap";
+import { Icon } from "src/components/Shared";
+import * as GQL from "src/core/generated-graphql";
+import { FormattedMessage, useIntl } from "react-intl";
+import { multiValueSceneFields, SceneField, sceneFields } from "./constants";
+import { ThreeStateBoolean } from "./ThreeStateBoolean";
+
+interface IFieldOptionsEditor {
+  options: GQL.IdentifyFieldOptions | undefined;
+  field: string;
+  editField: () => void;
+  editOptions: (o?: GQL.IdentifyFieldOptions | null) => void;
+  editing: boolean;
+  allowSetDefault: boolean;
+  defaultOptions?: GQL.IdentifyMetadataOptionsInput;
+}
+
+interface IFieldOptions {
+  field: string;
+  strategy: GQL.IdentifyFieldStrategy | undefined;
+  createMissing?: GQL.Maybe<boolean> | undefined;
+}
+
+const FieldOptionsEditor: React.FC<IFieldOptionsEditor> = ({
+  options,
+  field,
+  editField,
+  editOptions,
+  editing,
+  allowSetDefault,
+  defaultOptions,
+}) => {
+  const intl = useIntl();
+
+  const [localOptions, setLocalOptions] = useState<IFieldOptions>();
+
+  const resetOptions = useCallback(() => {
+    let toSet: IFieldOptions;
+    if (!options) {
+      // unset - use default values
+      toSet = {
+        field,
+        strategy: undefined,
+        createMissing: undefined,
+      };
+    } else {
+      toSet = {
+        field,
+        strategy: options.strategy,
+        createMissing: options.createMissing,
+      };
+    }
+    setLocalOptions(toSet);
+  }, [options, field]);
+
+  useEffect(() => {
+    resetOptions();
+  }, [resetOptions]);
+
+  function renderField() {
+    return intl.formatMessage({ id: field });
+  }
+
+  function renderStrategy() {
+    if (!localOptions) {
+      return;
+    }
+
+    const strategies = Object.entries(GQL.IdentifyFieldStrategy);
+    let { strategy } = localOptions;
+    if (strategy === undefined) {
+      if (!allowSetDefault) {
+        strategy = GQL.IdentifyFieldStrategy.Merge;
+      }
+    }
+
+    if (!editing) {
+      if (strategy === undefined) {
+        return intl.formatMessage({ id: "use_default" });
+      }
+
+      const f = strategies.find((s) => s[1] === strategy);
+      return intl.formatMessage({
+        id: `config.tasks.identify.field_strategies.${f![0].toLowerCase()}`,
+      });
+    }
+
+    if (!localOptions) {
+      return <></>;
+    }
+
+    return (
+      <Form.Group>
+        {allowSetDefault ? (
+          <Form.Check
+            type="radio"
+            id={`${field}-strategy-default`}
+            checked={localOptions.strategy === undefined}
+            onChange={() =>
+              setLocalOptions({
+                ...localOptions,
+                strategy: undefined,
+              })
+            }
+            disabled={!editing}
+            label={intl.formatMessage({ id: "use_default" })}
+          />
+        ) : undefined}
+        {strategies.map((f) => (
+          <Form.Check
+            type="radio"
+            key={f[0]}
+            id={`${field}-strategy-${f[0]}`}
+            checked={localOptions.strategy === f[1]}
+            onChange={() =>
+              setLocalOptions({
+                ...localOptions,
+                strategy: f[1],
+              })
+            }
+            disabled={!editing}
+            label={intl.formatMessage({
+              id: `config.tasks.identify.field_strategies.${f[0].toLowerCase()}`,
+            })}
+          />
+        ))}
+      </Form.Group>
+    );
+  }
+
+  function maybeRenderCreateMissing() {
+    if (!localOptions) {
+      return;
+    }
+
+    if (
+      multiValueSceneFields.includes(localOptions.field as SceneField) &&
+      localOptions.strategy !== GQL.IdentifyFieldStrategy.Ignore
+    ) {
+      const value =
+        localOptions.createMissing === null
+          ? undefined
+          : localOptions.createMissing;
+
+      if (!editing) {
+        if (value === undefined && allowSetDefault) {
+          return intl.formatMessage({ id: "use_default" });
+        }
+        if (value) {
+          return <Icon icon="check" className="text-success" />;
+        }
+
+        return <Icon icon="times" className="text-danger" />;
+      }
+
+      const defaultVal = defaultOptions?.fieldOptions?.find(
+        (f) => f.field === localOptions.field
+      )?.createMissing;
+
+      if (localOptions.strategy === undefined) {
+        return;
+      }
+
+      return (
+        <ThreeStateBoolean
+          id="create-missing"
+          disabled={!editing}
+          allowUndefined={allowSetDefault}
+          value={value}
+          setValue={(v) =>
+            setLocalOptions({ ...localOptions, createMissing: v })
+          }
+          defaultValue={defaultVal ?? undefined}
+        />
+      );
+    }
+  }
+
+  function onEditOptions() {
+    if (!localOptions) {
+      return;
+    }
+
+    // send null if strategy is undefined
+    if (localOptions.strategy === undefined) {
+      editOptions(null);
+      resetOptions();
+    } else {
+      let { createMissing } = localOptions;
+      if (createMissing === undefined && !allowSetDefault) {
+        createMissing = false;
+      }
+
+      editOptions({
+        ...localOptions,
+        strategy: localOptions.strategy,
+        createMissing,
+      });
+    }
+  }
+
+  return (
+    <tr>
+      <td>{renderField()}</td>
+      <td>{renderStrategy()}</td>
+      <td>{maybeRenderCreateMissing()}</td>
+      <td className="text-right">
+        {editing ? (
+          <>
+            <Button
+              className="minimal text-success"
+              onClick={() => onEditOptions()}
+            >
+              <Icon icon="check" />
+            </Button>
+            <Button
+              className="minimal text-danger"
+              onClick={() => {
+                editOptions();
+                resetOptions();
+              }}
+            >
+              <Icon icon="times" />
+            </Button>
+          </>
+        ) : (
+          <>
+            <Button className="minimal" onClick={() => editField()}>
+              <Icon icon="pencil-alt" />
+            </Button>
+          </>
+        )}
+      </td>
+    </tr>
+  );
+};
+
+interface IFieldOptionsList {
+  fieldOptions?: GQL.IdentifyFieldOptions[];
+  setFieldOptions: (o: GQL.IdentifyFieldOptions[]) => void;
+  setEditingField: (v: boolean) => void;
+  allowSetDefault?: boolean;
+  defaultOptions?: GQL.IdentifyMetadataOptionsInput;
+}
+
+export const FieldOptionsList: React.FC<IFieldOptionsList> = ({
+  fieldOptions,
+  setFieldOptions,
+  setEditingField,
+  allowSetDefault = true,
+  defaultOptions,
+}) => {
+  const [localFieldOptions, setLocalFieldOptions] = useState<
+    GQL.IdentifyFieldOptions[]
+  >();
+  const [editField, setEditField] = useState<string | undefined>();
+
+  useEffect(() => {
+    if (fieldOptions) {
+      setLocalFieldOptions([...fieldOptions]);
+    } else {
+      setLocalFieldOptions([]);
+    }
+  }, [fieldOptions]);
+
+  function handleEditOptions(o?: GQL.IdentifyFieldOptions | null) {
+    if (!localFieldOptions) {
+      return;
+    }
+
+    if (o !== undefined) {
+      const newOptions = [...localFieldOptions];
+      const index = newOptions.findIndex(
+        (option) => option.field === editField
+      );
+      if (index !== -1) {
+        // if null, then we're removing
+        if (o === null) {
+          newOptions.splice(index, 1);
+        } else {
+          // replace in list
+          newOptions.splice(index, 1, o);
+        }
+      } else if (o !== null) {
+        // don't add if null
+        newOptions.push(o);
+      }
+
+      setFieldOptions(newOptions);
+    }
+
+    setEditField(undefined);
+    setEditingField(false);
+  }
+
+  function onEditField(field: string) {
+    setEditField(field);
+    setEditingField(true);
+  }
+
+  if (!localFieldOptions) {
+    return <></>;
+  }
+
+  return (
+    <Form.Group className="scraper-sources">
+      <h5>
+        <FormattedMessage id="config.tasks.identify.field_options" />
+      </h5>
+      <Table responsive className="field-options-table">
+        <thead>
+          <tr>
+            <th className="w-25">Field</th>
+            <th className="w-25">Strategy</th>
+            <th className="w-25">Create missing</th>
+            {/* eslint-disable-next-line jsx-a11y/control-has-associated-label */}
+            <th className="w-25" />
+          </tr>
+        </thead>
+        <tbody>
+          {sceneFields.map((f) => (
+            <FieldOptionsEditor
+              key={f}
+              field={f}
+              allowSetDefault={allowSetDefault}
+              options={localFieldOptions.find((o) => o.field === f)}
+              editField={() => onEditField(f)}
+              editOptions={handleEditOptions}
+              editing={f === editField}
+              defaultOptions={defaultOptions}
+            />
+          ))}
+        </tbody>
+      </Table>
+    </Form.Group>
+  );
+};

ui/v2.5/src/components/Dialogs/IdentifyDialog/IdentifyDialog.tsx

@@ -0,0 +1,451 @@
+import React, { useState, useEffect, useMemo } from "react";
+import { Button, Form, Spinner } from "react-bootstrap";
+import {
+  mutateMetadataIdentify,
+  useConfiguration,
+  useConfigureDefaults,
+  useListSceneScrapers,
+} from "src/core/StashService";
+import { Icon, Modal } from "src/components/Shared";
+import { useToast } from "src/hooks";
+import * as GQL from "src/core/generated-graphql";
+import { FormattedMessage, useIntl } from "react-intl";
+import { withoutTypename } from "src/utils";
+import {
+  SCRAPER_PREFIX,
+  STASH_BOX_PREFIX,
+} from "src/components/Tagger/constants";
+import { DirectorySelectionDialog } from "src/components/Settings/SettingsTasksPanel/DirectorySelectionDialog";
+import { Manual } from "src/components/Help/Manual";
+import { IScraperSource } from "./constants";
+import { OptionsEditor } from "./Options";
+import { SourcesEditor, SourcesList } from "./Sources";
+
+const autoTagScraperID = "builtin_autotag";
+
+interface IIdentifyDialogProps {
+  selectedIds?: string[];
+  onClose: () => void;
+}
+
+export const IdentifyDialog: React.FC<IIdentifyDialogProps> = ({
+  selectedIds,
+  onClose,
+}) => {
+  function getDefaultOptions(): GQL.IdentifyMetadataOptionsInput {
+    return {
+      fieldOptions: [
+        {
+          field: "title",
+          strategy: GQL.IdentifyFieldStrategy.Overwrite,
+        },
+      ],
+      includeMalePerformers: true,
+      setCoverImage: true,
+      setOrganized: false,
+    };
+  }
+
+  const [configureDefaults] = useConfigureDefaults();
+
+  const [options, setOptions] = useState<GQL.IdentifyMetadataOptionsInput>(
+    getDefaultOptions()
+  );
+  const [sources, setSources] = useState<IScraperSource[]>([]);
+  const [editingSource, setEditingSource] = useState<
+    IScraperSource | undefined
+  >();
+  const [paths, setPaths] = useState<string[]>([]);
+  const [showManual, setShowManual] = useState(false);
+  const [settingPaths, setSettingPaths] = useState(false);
+  const [animation, setAnimation] = useState(true);
+  const [editingField, setEditingField] = useState(false);
+  const [savingDefaults, setSavingDefaults] = useState(false);
+
+  const intl = useIntl();
+  const Toast = useToast();
+
+  const { data: configData, error: configError } = useConfiguration();
+  const { data: scraperData, error: scraperError } = useListSceneScrapers();
+
+  const allSources = useMemo(() => {
+    if (!configData || !scraperData) return;
+
+    const ret: IScraperSource[] = [];
+
+    ret.push(
+      ...configData.configuration.general.stashBoxes.map((b, i) => {
+        return {
+          id: `${STASH_BOX_PREFIX}${i}`,
+          displayName: `stash-box: ${b.name}`,
+          stash_box_endpoint: b.endpoint,
+        };
+      })
+    );
+
+    const scrapers = scraperData.listSceneScrapers;
+
+    const fragmentScrapers = scrapers.filter((s) =>
+      s.scene?.supported_scrapes.includes(GQL.ScrapeType.Fragment)
+    );
+
+    ret.push(
+      ...fragmentScrapers.map((s) => {
+        return {
+          id: `${SCRAPER_PREFIX}${s.id}`,
+          displayName: s.name,
+          scraper_id: s.id,
+        };
+      })
+    );
+
+    return ret;
+  }, [configData, scraperData]);
+
+  const selectionStatus = useMemo(() => {
+    if (selectedIds) {
+      return (
+        <Form.Group id="selected-identify-ids">
+          <FormattedMessage
+            id="config.tasks.identify.identifying_scenes"
+            values={{
+              num: selectedIds.length,
+              scene: intl.formatMessage(
+                {
+                  id: "countables.scenes",
+                },
+                {
+                  count: selectedIds.length,
+                }
+              ),
+            }}
+          />
+          .
+        </Form.Group>
+      );
+    }
+    const message = paths.length ? (
+      <div>
+        <FormattedMessage id="config.tasks.identify.identifying_from_paths" />:
+        <ul>
+          {paths.map((p) => (
+            <li key={p}>{p}</li>
+          ))}
+        </ul>
+      </div>
+    ) : (
+      <span>
+        <FormattedMessage
+          id="config.tasks.identify.identifying_scenes"
+          values={{
+            num: intl.formatMessage({ id: "all" }),
+            scene: intl.formatMessage(
+              {
+                id: "countables.scenes",
+              },
+              {
+                count: 0,
+              }
+            ),
+          }}
+        />
+        .
+      </span>
+    );
+
+    function onClick() {
+      setAnimation(false);
+      setSettingPaths(true);
+    }
+
+    return (
+      <Form.Group id="selected-identify-folders">
+        <div>
+          {message}
+          <div>
+            <Button
+              title={intl.formatMessage({ id: "actions.select_folders" })}
+              onClick={() => onClick()}
+            >
+              <Icon icon="folder-open" />
+            </Button>
+          </div>
+        </div>
+      </Form.Group>
+    );
+  }, [selectedIds, intl, paths]);
+
+  useEffect(() => {
+    if (!configData || !allSources) return;
+
+    const { identify: identifyDefaults } = configData.configuration.defaults;
+
+    if (identifyDefaults) {
+      const mappedSources = identifyDefaults.sources
+        .map((s) => {
+          const found = allSources.find(
+            (ss) =>
+              ss.scraper_id === s.source.scraper_id ||
+              ss.stash_box_endpoint === s.source.stash_box_endpoint
+          );
+
+          if (!found) return;
+
+          const ret: IScraperSource = {
+            ...found,
+          };
+
+          if (s.options) {
+            const sourceOptions = withoutTypename(s.options);
+            sourceOptions.fieldOptions = sourceOptions.fieldOptions?.map(
+              withoutTypename
+            );
+            ret.options = sourceOptions;
+          }
+
+          return ret;
+        })
+        .filter((s) => s) as IScraperSource[];
+
+      setSources(mappedSources);
+      if (identifyDefaults.options) {
+        const defaultOptions = withoutTypename(identifyDefaults.options);
+        defaultOptions.fieldOptions = defaultOptions.fieldOptions?.map(
+          withoutTypename
+        );
+        setOptions(defaultOptions);
+      }
+    } else {
+      // default to first stash-box instance only
+      const stashBox = allSources.find((s) => s.stash_box_endpoint);
+
+      // add auto-tag as well
+      const autoTag = allSources.find(
+        (s) => s.id === `${SCRAPER_PREFIX}${autoTagScraperID}`
+      );
+
+      const newSources: IScraperSource[] = [];
+      if (stashBox) {
+        newSources.push(stashBox);
+      }
+
+      // sanity check - this should always be true
+      if (autoTag) {
+        // don't set organised by default
+        const autoTagCopy = { ...autoTag };
+        autoTagCopy.options = {
+          setOrganized: false,
+        };
+        newSources.push(autoTagCopy);
+      }
+
+      setSources(newSources);
+    }
+  }, [allSources, configData]);
+
+  if (configError || scraperError)
+    return <div>{configError ?? scraperError}</div>;
+  if (!allSources || !configData) return <div />;
+
+  function makeIdentifyInput(): GQL.IdentifyMetadataInput {
+    return {
+      sources: sources.map((s) => {
+        return {
+          source: {
+            scraper_id: s.scraper_id,
+            stash_box_endpoint: s.stash_box_endpoint,
+          },
+          options: s.options,
+        };
+      }),
+      options,
+      sceneIDs: selectedIds,
+      paths,
+    };
+  }
+
+  function makeDefaultIdentifyInput() {
+    const ret = makeIdentifyInput();
+    const { sceneIDs, paths: _paths, ...withoutSpecifics } = ret;
+    return withoutSpecifics;
+  }
+
+  async function onIdentify() {
+    try {
+      await mutateMetadataIdentify(makeIdentifyInput());
+
+      Toast.success({
+        content: intl.formatMessage(
+          { id: "config.tasks.added_job_to_queue" },
+          { operation_name: intl.formatMessage({ id: "actions.identify" }) }
+        ),
+      });
+    } catch (e) {
+      Toast.error(e);
+    } finally {
+      onClose();
+    }
+  }
+
+  function getAvailableSources() {
+    // only include scrapers not already present
+    return !editingSource?.id === undefined
+      ? []
+      : allSources?.filter((s) => {
+          return !sources.some((ss) => ss.id === s.id);
+        }) ?? [];
+  }
+
+  function onEditSource(s?: IScraperSource) {
+    setAnimation(false);
+
+    // if undefined, then set a dummy source to create a new one
+    if (!s) {
+      setEditingSource(getAvailableSources()[0]);
+    } else {
+      setEditingSource(s);
+    }
+  }
+
+  function onShowManual() {
+    setAnimation(false);
+    setShowManual(true);
+  }
+
+  function isNewSource() {
+    return !!editingSource && !sources.includes(editingSource);
+  }
+
+  function onSaveSource(s?: IScraperSource) {
+    if (s) {
+      let found = false;
+      const newSources = sources.map((ss) => {
+        if (ss.id === s.id) {
+          found = true;
+          return s;
+        }
+        return ss;
+      });
+
+      if (!found) {
+        newSources.push(s);
+      }
+
+      setSources(newSources);
+    }
+    setEditingSource(undefined);
+  }
+
+  async function setAsDefault() {
+    try {
+      setSavingDefaults(true);
+      await configureDefaults({
+        variables: {
+          input: {
+            identify: makeDefaultIdentifyInput(),
+          },
+        },
+      });
+    } catch (e) {
+      Toast.error(e);
+    } finally {
+      setSavingDefaults(false);
+    }
+  }
+
+  if (editingSource) {
+    return (
+      <SourcesEditor
+        availableSources={getAvailableSources()}
+        source={editingSource}
+        saveSource={onSaveSource}
+        isNew={isNewSource()}
+        defaultOptions={options}
+      />
+    );
+  }
+
+  if (settingPaths) {
+    return (
+      <DirectorySelectionDialog
+        animation={false}
+        allowEmpty
+        initialPaths={paths}
+        onClose={(p) => {
+          if (p) {
+            setPaths(p);
+          }
+          setSettingPaths(false);
+        }}
+      />
+    );
+  }
+
+  if (showManual) {
+    return (
+      <Manual
+        animation={false}
+        show
+        onClose={() => setShowManual(false)}
+        defaultActiveTab="Identify.md"
+      />
+    );
+  }
+
+  return (
+    <Modal
+      modalProps={{ animation, size: "lg" }}
+      show
+      icon="cogs"
+      header={intl.formatMessage({ id: "actions.identify" })}
+      accept={{
+        onClick: onIdentify,
+        text: intl.formatMessage({ id: "actions.identify" }),
+      }}
+      cancel={{
+        onClick: () => onClose(),
+        text: intl.formatMessage({ id: "actions.cancel" }),
+        variant: "secondary",
+      }}
+      disabled={editingField || savingDefaults || sources.length === 0}
+      footerButtons={
+        <Button
+          variant="secondary"
+          disabled={editingField || savingDefaults}
+          onClick={() => setAsDefault()}
+        >
+          {savingDefaults && (
+            <Spinner animation="border" role="status" size="sm" />
+          )}
+          <FormattedMessage id="actions.set_as_default" />
+        </Button>
+      }
+      leftFooterButtons={
+        <Button
+          title="Help"
+          className="minimal help-button"
+          onClick={() => onShowManual()}
+        >
+          <Icon icon="question-circle" />
+        </Button>
+      }
+    >
+      <Form>
+        {selectionStatus}
+        <SourcesList
+          sources={sources}
+          setSources={(s) => setSources(s)}
+          editSource={onEditSource}
+          canAdd={sources.length < allSources.length}
+        />
+        <OptionsEditor
+          options={options}
+          setOptions={(o) => setOptions(o)}
+          setEditingField={(v) => setEditingField(v)}
+        />
+      </Form>
+    </Modal>
+  );
+};
+
+export default IdentifyDialog;

ui/v2.5/src/components/Dialogs/IdentifyDialog/Options.tsx

@@ -0,0 +1,117 @@
+import React from "react";
+import { Form } from "react-bootstrap";
+import * as GQL from "src/core/generated-graphql";
+import { FormattedMessage, useIntl } from "react-intl";
+import { IScraperSource } from "./constants";
+import { FieldOptionsList } from "./FieldOptions";
+import { ThreeStateBoolean } from "./ThreeStateBoolean";
+
+interface IOptionsEditor {
+  options: GQL.IdentifyMetadataOptionsInput;
+  setOptions: (s: GQL.IdentifyMetadataOptionsInput) => void;
+  source?: IScraperSource;
+  defaultOptions?: GQL.IdentifyMetadataOptionsInput;
+  setEditingField: (v: boolean) => void;
+}
+
+export const OptionsEditor: React.FC<IOptionsEditor> = ({
+  options,
+  setOptions: setOptionsState,
+  source,
+  setEditingField,
+  defaultOptions,
+}) => {
+  const intl = useIntl();
+
+  function setOptions(v: Partial<GQL.IdentifyMetadataOptionsInput>) {
+    setOptionsState({ ...options, ...v });
+  }
+
+  const headingID = !source
+    ? "config.tasks.identify.default_options"
+    : "config.tasks.identify.source_options";
+  const checkboxProps = {
+    allowUndefined: !!source,
+    indeterminateClassname: "text-muted",
+  };
+
+  return (
+    <Form.Group>
+      <Form.Group>
+        <h5>
+          <FormattedMessage
+            id={headingID}
+            values={{ source: source?.displayName }}
+          />
+        </h5>
+        {!source && (
+          <Form.Text className="text-muted">
+            {intl.formatMessage({
+              id: "config.tasks.identify.explicit_set_description",
+            })}
+          </Form.Text>
+        )}
+      </Form.Group>
+      <Form.Group>
+        <ThreeStateBoolean
+          id="include-male-performers"
+          value={
+            options.includeMalePerformers === null
+              ? undefined
+              : options.includeMalePerformers
+          }
+          setValue={(v) =>
+            setOptions({
+              includeMalePerformers: v,
+            })
+          }
+          label={intl.formatMessage({
+            id: "config.tasks.identify.include_male_performers",
+          })}
+          defaultValue={defaultOptions?.includeMalePerformers ?? undefined}
+          {...checkboxProps}
+        />
+        <ThreeStateBoolean
+          id="set-cover-image"
+          value={
+            options.setCoverImage === null ? undefined : options.setCoverImage
+          }
+          setValue={(v) =>
+            setOptions({
+              setCoverImage: v,
+            })
+          }
+          label={intl.formatMessage({
+            id: "config.tasks.identify.set_cover_images",
+          })}
+          defaultValue={defaultOptions?.setCoverImage ?? undefined}
+          {...checkboxProps}
+        />
+        <ThreeStateBoolean
+          id="set-organized"
+          value={
+            options.setOrganized === null ? undefined : options.setOrganized
+          }
+          setValue={(v) =>
+            setOptions({
+              setOrganized: v,
+            })
+          }
+          label={intl.formatMessage({
+            id: "config.tasks.identify.set_organized",
+          })}
+          defaultValue={defaultOptions?.setOrganized ?? undefined}
+          {...checkboxProps}
+        />
+      </Form.Group>
+
+      <FieldOptionsList
+        fieldOptions={options.fieldOptions ?? undefined}
+        setFieldOptions={(o) => setOptions({ fieldOptions: o })}
+        setEditingField={setEditingField}
+        allowSetDefault={!!source}
+        defaultOptions={defaultOptions}
+      />
+    </Form.Group>
+  );
+};

ui/v2.5/src/components/Dialogs/IdentifyDialog/Sources.tsx

@@ -0,0 +1,216 @@
+import React, { useState, useEffect } from "react";
+import { Form, Button, ListGroup } from "react-bootstrap";
+import { Modal, Icon } from "src/components/Shared";
+import { FormattedMessage, useIntl } from "react-intl";
+import * as GQL from "src/core/generated-graphql";
+import { IScraperSource } from "./constants";
+import { OptionsEditor } from "./Options";
+
+interface ISourceEditor {
+  isNew: boolean;
+  availableSources: IScraperSource[];
+  source: IScraperSource;
+  saveSource: (s?: IScraperSource) => void;
+  defaultOptions: GQL.IdentifyMetadataOptionsInput;
+}
+
+export const SourcesEditor: React.FC<ISourceEditor> = ({
+  isNew,
+  availableSources,
+  source: initialSource,
+  saveSource,
+  defaultOptions,
+}) => {
+  const [source, setSource] = useState<IScraperSource>(initialSource);
+  const [editingField, setEditingField] = useState(false);
+
+  const intl = useIntl();
+
+  // if id is empty, then we are adding a new source
+  const headerMsgId = isNew ? "actions.add" : "dialogs.edit_entity_title";
+  const acceptMsgId = isNew ? "actions.add" : "actions.confirm";
+
+  function handleSourceSelect(e: React.ChangeEvent<HTMLSelectElement>) {
+    const selectedSource = availableSources.find(
+      (s) => s.id === e.currentTarget.value
+    );
+    if (!selectedSource) return;
+
+    setSource({
+      ...source,
+      id: selectedSource.id,
+      displayName: selectedSource.displayName,
+      scraper_id: selectedSource.scraper_id,
+      stash_box_endpoint: selectedSource.stash_box_endpoint,
+    });
+  }
+
+  return (
+    <Modal
+      dialogClassName="identify-source-editor"
+      modalProps={{ animation: false, size: "lg" }}
+      show
+      icon={isNew ? "plus" : "pencil-alt"}
+      header={intl.formatMessage(
+        { id: headerMsgId },
+        {
+          count: 1,
+          singularEntity: source?.displayName,
+        }
+      )}
+      accept={{
+        onClick: () => saveSource(source),
+        text: intl.formatMessage({ id: acceptMsgId }),
+      }}
+      cancel={{
+        onClick: () => saveSource(),
+        text: intl.formatMessage({ id: "actions.cancel" }),
+        variant: "secondary",
+      }}
+      disabled={
+        (!source.scraper_id && !source.stash_box_endpoint) || editingField
+      }
+    >
+      <Form>
+        {isNew && (
+          <Form.Group>
+            <h5>
+              <FormattedMessage id="config.tasks.identify.source" />
+            </h5>
+            <Form.Control
+              as="select"
+              value={source.id}
+              className="input-control"
+              onChange={handleSourceSelect}
+            >
+              {availableSources.map((i) => (
+                <option value={i.id} key={i.id}>
+                  {i.displayName}
+                </option>
+              ))}
+            </Form.Control>
+          </Form.Group>
+        )}
+        <OptionsEditor
+          options={source.options ?? {}}
+          setOptions={(o) => setSource({ ...source, options: o })}
+          source={source}
+          setEditingField={(v) => setEditingField(v)}
+          defaultOptions={defaultOptions}
+        />
+      </Form>
+    </Modal>
+  );
+};
+
+interface ISourcesList {
+  sources: IScraperSource[];
+  setSources: (s: IScraperSource[]) => void;
+  editSource: (s?: IScraperSource) => void;
+  canAdd: boolean;
+}
+
+export const SourcesList: React.FC<ISourcesList> = ({
+  sources,
+  setSources,
+  editSource,
+  canAdd,
+}) => {
+  const [tempSources, setTempSources] = useState(sources);
+  const [dragIndex, setDragIndex] = useState<number | undefined>();
+  const [mouseOverIndex, setMouseOverIndex] = useState<number | undefined>();
+
+  useEffect(() => {
+    setTempSources([...sources]);
+  }, [sources]);
+
+  function removeSource(index: number) {
+    const newSources = [...sources];
+    newSources.splice(index, 1);
+    setSources(newSources);
+  }
+
+  function onDragStart(event: React.DragEvent<HTMLElement>, index: number) {
+    event.dataTransfer.effectAllowed = "move";
+    setDragIndex(index);
+  }
+
+  function onDragOver(event: React.DragEvent<HTMLElement>, index?: number) {
+    if (dragIndex !== undefined && index !== undefined && index !== dragIndex) {
+      const newSources = [...tempSources];
+      const moved = newSources.splice(dragIndex, 1);
+      newSources.splice(index, 0, moved[0]);
+      setTempSources(newSources);
+      setDragIndex(index);
+    }
+
+    event.dataTransfer.dropEffect = "move";
+    event.preventDefault();
+  }
+
+  function onDragOverDefault(event: React.DragEvent<HTMLDivElement>) {
+    event.dataTransfer.dropEffect = "move";
+    event.preventDefault();
+  }
+
+  function onDrop() {
+    // assume we've already set the temp source list
+    // feed it up
+    setSources(tempSources);
+    setDragIndex(undefined);
+    setMouseOverIndex(undefined);
+  }
+
+  return (
+    <Form.Group className="scraper-sources" onDragOver={onDragOverDefault}>
+      <h5>
+        <FormattedMessage id="config.tasks.identify.sources" />
+      </h5>
+      <ListGroup as="ul" className="scraper-source-list">
+        {tempSources.map((s, index) => (
+          <ListGroup.Item
+            as="li"
+            key={s.id}
+            className="d-flex justify-content-between align-items-center"
+            draggable={mouseOverIndex === index}
+            onDragStart={(e) => onDragStart(e, index)}
+            onDragEnter={(e) => onDragOver(e, index)}
+            onDrop={() => onDrop()}
+          >
+            <div>
+              <div
+                className="minimal text-muted drag-handle"
+                onMouseEnter={() => setMouseOverIndex(index)}
+                onMouseLeave={() => setMouseOverIndex(undefined)}
+              >
+                <Icon icon="grip-vertical" />
+              </div>
+              {s.displayName}
+            </div>
+            <div>
+              <Button className="minimal" onClick={() => editSource(s)}>
+                <Icon icon="cog" />
+              </Button>
+              <Button
+                className="minimal text-danger"
+                onClick={() => removeSource(index)}
+              >
+                <Icon icon="minus" />
+              </Button>
+            </div>
+          </ListGroup.Item>
+        ))}
+      </ListGroup>
+      {canAdd && (
+        <div className="text-right">
+          <Button
+            className="minimal add-scraper-source-button"
+            onClick={() => editSource()}
+          >
+            <Icon icon="plus" />
+          </Button>
+        </div>
+      )}
+    </Form.Group>
+  );
+};

ui/v2.5/src/components/Dialogs/IdentifyDialog/ThreeStateBoolean.tsx

@@ -0,0 +1,90 @@
+import React from "react";
+import { Form } from "react-bootstrap";
+import { useIntl } from "react-intl";
+
+interface IThreeStateBoolean {
+  id: string;
+  value: boolean | undefined;
+  setValue: (v: boolean | undefined) => void;
+  allowUndefined?: boolean;
+  label?: React.ReactNode;
+  disabled?: boolean;
+  defaultValue?: boolean;
+}
+
+export const ThreeStateBoolean: React.FC<IThreeStateBoolean> = ({
+  id,
+  value,
+  setValue,
+  allowUndefined = true,
+  label,
+  disabled,
+  defaultValue,
+}) => {
+  const intl = useIntl();
+
+  if (!allowUndefined) {
+    return (
+      <Form.Check
+        id={id}
+        disabled={disabled}
+        checked={value}
+        label={label}
+        onChange={() => setValue(!value)}
+      />
+    );
+  }
+
+  function getBooleanText(v: boolean) {
+    if (v) {
+      return intl.formatMessage({ id: "true" });
+    }
+    return intl.formatMessage({ id: "false" });
+  }
+
+  function getButtonText(v: boolean | undefined) {
+    if (v === undefined) {
+      const defaultVal =
+        defaultValue !== undefined ? (
+          <span className="default-value">
+            {" "}
+            ({getBooleanText(defaultValue)})
+          </span>
+        ) : (
+          ""
+        );
+      return (
+        <span>
+          {intl.formatMessage({ id: "use_default" })}
+          {defaultVal}
+        </span>
+      );
+    }
+
+    return getBooleanText(v);
+  }
+
+  function renderModeButton(v: boolean | undefined) {
+    return (
+      <Form.Check
+        type="radio"
+        id={`${id}-value-${v ?? "undefined"}`}
+        checked={value === v}
+        onChange={() => setValue(v)}
+        disabled={disabled}
+        label={getButtonText(v)}
+      />
+    );
+  }
+
+  return (
+    <Form.Group>
+      <h6>{label}</h6>
+      <Form.Group>
+        {renderModeButton(undefined)}
+        {renderModeButton(false)}
+        {renderModeButton(true)}
+      </Form.Group>
+    </Form.Group>
+  );
+};

ui/v2.5/src/components/Dialogs/IdentifyDialog/constants.ts

@@ -0,0 +1,27 @@
+import * as GQL from "src/core/generated-graphql";
+
+export interface IScraperSource {
+  id: string;
+  displayName: string;
+  stash_box_endpoint?: string;
+  scraper_id?: string;
+  options?: GQL.IdentifyMetadataOptionsInput;
+}
+
+export const sceneFields = [
+  "title",
+  "date",
+  "details",
+  "url",
+  "studio",
+  "performers",
+  "tags",
+  "stash_ids",
+] as const;
+export type SceneField = typeof sceneFields[number];
+
+export const multiValueSceneFields: SceneField[] = [
+  "studio",
+  "performers",
+  "tags",
+];

ui/v2.5/src/components/Dialogs/IdentifyDialog/styles.scss

@@ -0,0 +1,45 @@
+.identify-source-editor {
+  .default-value {
+    color: #bfccd6;
+  }
+}
+
+.scraper-source-list {
+  .list-group-item {
+    background-color: $textfield-bg;
+    padding: 0.25em;
+
+    .drag-handle {
+      cursor: move;
+      display: inline-block;
+      margin: -0.25em 0.25em -0.25em -0.25em;
+      padding: 0.25em 0.5em 0.25em;
+    }
+
+    .drag-handle:hover,
+    .drag-handle:active,
+    .drag-handle:focus,
+    .drag-handle:focus:active {
+      background-color: initial;
+      border-color: initial;
+      box-shadow: initial;
+    }
+  }
+}
+
+.scraper-sources {
+  .add-scraper-source-button {
+    margin-right: 0.25em;
+  }
+}
+
+.field-options-table td:first-child {
+  padding-left: 0.75rem;
+}
+
+#selected-identify-folders {
+  & > div {
+    display: flex;
+    justify-content: space-between;
+  }
+}

ui/v2.5/src/components/Help/Manual.tsx

@@ -18,15 +18,18 @@ import KeyboardShortcuts from "src/docs/en/KeyboardShortcuts.md";
 import Help from "src/docs/en/Help.md";
 import Deduplication from "src/docs/en/Deduplication.md";
 import Interactive from "src/docs/en/Interactive.md";
+import Identify from "src/docs/en/Identify.md";
 import { MarkdownPage } from "../Shared/MarkdownPage";
 
 interface IManualProps {
+  animation?: boolean;
   show: boolean;
   onClose: () => void;
   defaultActiveTab?: string;
 }
 
 export const Manual: React.FC<IManualProps> = ({
+  animation,
   show,
   onClose,
   defaultActiveTab,
@@ -52,6 +55,12 @@ export const Manual: React.FC<IManualProps> = ({
       title: "Tasks",
       content: Tasks,
     },
+    {
+      key: "Identify.md",
+      title: "Identify",
+      content: Identify,
+      className: "indent-1",
+    },
     {
       key: "AutoTagging.md",
       title: "Auto Tagging",
@@ -152,6 +161,7 @@ export const Manual: React.FC<IManualProps> = ({
 
   return (
     <Modal
+      animation={animation}
       show={show}
       onHide={onClose}
       dialogClassName="modal-dialog-scrollable manual modal-xl"

ui/v2.5/src/components/Scenes/SceneList.tsx

@@ -23,6 +23,7 @@ import { SceneGenerateDialog } from "./SceneGenerateDialog";
 import { ExportDialog } from "../Shared/ExportDialog";
 import { SceneCardsGrid } from "./SceneCardsGrid";
 import { TaggerContext } from "../Tagger/context";
+import { IdentifyDialog } from "../Dialogs/IdentifyDialog/IdentifyDialog";
 
 interface ISceneList {
   filterHook?: (filter: ListFilterModel) => ListFilterModel;
@@ -38,6 +39,7 @@ export const SceneList: React.FC<ISceneList> = ({
   const intl = useIntl();
   const history = useHistory();
   const [isGenerateDialogOpen, setIsGenerateDialogOpen] = useState(false);
+  const [isIdentifyDialogOpen, setIsIdentifyDialogOpen] = useState(false);
   const [isExportDialogOpen, setIsExportDialogOpen] = useState(false);
   const [isExportAll, setIsExportAll] = useState(false);
 
@@ -53,10 +55,15 @@ export const SceneList: React.FC<ISceneList> = ({
       onClick: playRandom,
     },
     {
-      text: intl.formatMessage({ id: "actions.generate" }),
+      text: `${intl.formatMessage({ id: "actions.generate" })}…`,
       onClick: generate,
       isDisplayed: showWhenSelected,
     },
+    {
+      text: `${intl.formatMessage({ id: "actions.identify" })}…`,
+      onClick: identify,
+      isDisplayed: showWhenSelected,
+    },
     {
       text: intl.formatMessage({ id: "actions.export" }),
       onClick: onExport,
@@ -138,6 +145,10 @@ export const SceneList: React.FC<ISceneList> = ({
     setIsGenerateDialogOpen(true);
   }
 
+  async function identify() {
+    setIsIdentifyDialogOpen(true);
+  }
+
   async function onExport() {
     setIsExportAll(false);
     setIsExportDialogOpen(true);
@@ -163,6 +174,21 @@ export const SceneList: React.FC<ISceneList> = ({
     }
   }
 
+  function maybeRenderSceneIdentifyDialog(selectedIds: Set<string>) {
+    if (isIdentifyDialogOpen) {
+      return (
+        <>
+          <IdentifyDialog
+            selectedIds={Array.from(selectedIds.values())}
+            onClose={() => {
+              setIsIdentifyDialogOpen(false);
+            }}
+          />
+        </>
+      );
+    }
+  }
+
   function maybeRenderSceneExportDialog(selectedIds: Set<string>) {
     if (isExportDialogOpen) {
       return (
@@ -248,6 +274,7 @@ export const SceneList: React.FC<ISceneList> = ({
     return (
       <>
         {maybeRenderSceneGenerateDialog(selectedIds)}
+        {maybeRenderSceneIdentifyDialog(selectedIds)}
         {maybeRenderSceneExportDialog(selectedIds)}
         {renderScenes(result, filter, selectedIds)}
       </>

ui/v2.5/src/components/Settings/SettingsTasksPanel/DirectorySelectionDialog.tsx

@@ -6,18 +6,24 @@ import { FolderSelect } from "src/components/Shared/FolderSelect/FolderSelect";
 import { ConfigurationContext } from "src/hooks/Config";
 
 interface IDirectorySelectionDialogProps {
+  animation?: boolean;
+  initialPaths?: string[];
+  allowEmpty?: boolean;
   onClose: (paths?: string[]) => void;
 }
 
-export const DirectorySelectionDialog: React.FC<IDirectorySelectionDialogProps> = (
-  props: IDirectorySelectionDialogProps
-) => {
+export const DirectorySelectionDialog: React.FC<IDirectorySelectionDialogProps> = ({
+  animation,
+  allowEmpty = false,
+  initialPaths = [],
+  onClose,
+}) => {
   const intl = useIntl();
   const { configuration } = React.useContext(ConfigurationContext);
 
   const libraryPaths = configuration?.general.stashes.map((s) => s.path);
 
-  const [paths, setPaths] = useState<string[]>([]);
+  const [paths, setPaths] = useState<string[]>(initialPaths);
   const [currentDirectory, setCurrentDirectory] = useState<string>("");
 
   function removePath(p: string) {
@@ -33,17 +39,18 @@ export const DirectorySelectionDialog: React.FC<IDirectorySelectionDialogProps>
   return (
     <Modal
       show
-      disabled={paths.length === 0}
+      modalProps={{ animation }}
+      disabled={!allowEmpty && paths.length === 0}
       icon="pencil-alt"
       header={intl.formatMessage({ id: "actions.select_folders" })}
       accept={{
         onClick: () => {
-          props.onClose(paths);
+          onClose(paths);
         },
         text: intl.formatMessage({ id: "actions.confirm" }),
       }}
       cancel={{
-        onClick: () => props.onClose(),
+        onClick: () => onClose(),
         text: intl.formatMessage({ id: "actions.cancel" }),
         variant: "secondary",
       }}

ui/v2.5/src/components/Settings/SettingsTasksPanel/SettingsTasksPanel.tsx

@@ -16,6 +16,7 @@ import { useToast } from "src/hooks";
 import * as GQL from "src/core/generated-graphql";
 import { LoadingIndicator, Modal } from "src/components/Shared";
 import { downloadFile } from "src/utils";
+import IdentifyDialog from "src/components/Dialogs/IdentifyDialog/IdentifyDialog";
 import { GenerateButton } from "./GenerateButton";
 import { ImportDialog } from "./ImportDialog";
 import { DirectorySelectionDialog } from "./DirectorySelectionDialog";
@@ -27,13 +28,18 @@ type PluginTask = Pick<GQL.PluginTask, "name" | "description">;
 export const SettingsTasksPanel: React.FC = () => {
   const intl = useIntl();
   const Toast = useToast();
-  const [isImportAlertOpen, setIsImportAlertOpen] = useState<boolean>(false);
-  const [isCleanAlertOpen, setIsCleanAlertOpen] = useState<boolean>(false);
-  const [isImportDialogOpen, setIsImportDialogOpen] = useState<boolean>(false);
-  const [isScanDialogOpen, setIsScanDialogOpen] = useState<boolean>(false);
-  const [isAutoTagDialogOpen, setIsAutoTagDialogOpen] = useState<boolean>(
-    false
-  );
+  const [dialogOpen, setDialogOpenState] = useState({
+    importAlert: false,
+    cleanAlert: false,
+    import: false,
+    clean: false,
+    scan: false,
+    autoTag: false,
+    identify: false,
+  });
+
+  type DialogOpenState = typeof dialogOpen;
+
   const [isBackupRunning, setIsBackupRunning] = useState<boolean>(false);
   const [useFileMetadata, setUseFileMetadata] = useState<boolean>(false);
   const [stripFileExtension, setStripFileExtension] = useState<boolean>(false);
@@ -61,8 +67,14 @@ export const SettingsTasksPanel: React.FC = () => {
 
   const plugins = usePlugins();
 
+  function setDialogOpen(s: Partial<DialogOpenState>) {
+    setDialogOpenState((v) => {
+      return { ...v, ...s };
+    });
+  }
+
   async function onImport() {
-    setIsImportAlertOpen(false);
+    setDialogOpen({ importAlert: false });
     try {
       await mutateMetadataImport();
       Toast.success({
@@ -79,14 +91,14 @@ export const SettingsTasksPanel: React.FC = () => {
   function renderImportAlert() {
     return (
       <Modal
-        show={isImportAlertOpen}
+        show={dialogOpen.importAlert}
         icon="trash-alt"
         accept={{
           text: intl.formatMessage({ id: "actions.import" }),
           variant: "danger",
           onClick: onImport,
         }}
-        cancel={{ onClick: () => setIsImportAlertOpen(false) }}
+        cancel={{ onClick: () => setDialogOpen({ importAlert: false }) }}
       >
         <p>{intl.formatMessage({ id: "actions.tasks.import_warning" })}</p>
       </Modal>
@@ -94,7 +106,7 @@ export const SettingsTasksPanel: React.FC = () => {
   }
 
   function onClean() {
-    setIsCleanAlertOpen(false);
+    setDialogOpen({ cleanAlert: false });
     mutateMetadataClean({
       dryRun: cleanDryRun,
     });
@@ -116,14 +128,14 @@ export const SettingsTasksPanel: React.FC = () => {
 
     return (
       <Modal
-        show={isCleanAlertOpen}
+        show={dialogOpen.cleanAlert}
         icon="trash-alt"
         accept={{
           text: intl.formatMessage({ id: "actions.clean" }),
           variant: "danger",
           onClick: onClean,
         }}
-        cancel={{ onClick: () => setIsCleanAlertOpen(false) }}
+        cancel={{ onClick: () => setDialogOpen({ cleanAlert: false }) }}
       >
         {msg}
       </Modal>
@@ -131,15 +143,15 @@ export const SettingsTasksPanel: React.FC = () => {
   }
 
   function renderImportDialog() {
-    if (!isImportDialogOpen) {
+    if (!dialogOpen.import) {
       return;
     }
 
-    return <ImportDialog onClose={() => setIsImportDialogOpen(false)} />;
+    return <ImportDialog onClose={() => setDialogOpen({ import: false })} />;
   }
 
   function renderScanDialog() {
-    if (!isScanDialogOpen) {
+    if (!dialogOpen.scan) {
       return;
     }
 
@@ -151,7 +163,7 @@ export const SettingsTasksPanel: React.FC = () => {
       onScan(paths);
     }
 
-    setIsScanDialogOpen(false);
+    setDialogOpen({ scan: false });
   }
 
   async function onScan(paths?: string[]) {
@@ -178,19 +190,27 @@ export const SettingsTasksPanel: React.FC = () => {
   }
 
   function renderAutoTagDialog() {
-    if (!isAutoTagDialogOpen) {
+    if (!dialogOpen.autoTag) {
       return;
     }
 
     return <DirectorySelectionDialog onClose={onAutoTagDialogClosed} />;
   }
 
+  function maybeRenderIdentifyDialog() {
+    if (!dialogOpen.identify) return;
+
+    return (
+      <IdentifyDialog onClose={() => setDialogOpen({ identify: false })} />
+    );
+  }
+
   function onAutoTagDialogClosed(paths?: string[]) {
     if (paths) {
       onAutoTag(paths);
     }
 
-    setIsAutoTagDialogOpen(false);
+    setDialogOpen({ autoTag: false });
   }
 
   function getAutoTagInput(paths?: string[]) {
@@ -343,6 +363,7 @@ export const SettingsTasksPanel: React.FC = () => {
       {renderImportDialog()}
       {renderScanDialog()}
       {renderAutoTagDialog()}
+      {maybeRenderIdentifyDialog()}
 
       <h4>{intl.formatMessage({ id: "config.tasks.job_queue" })}</h4>
 
@@ -350,138 +371,159 @@ export const SettingsTasksPanel: React.FC = () => {
 
       <hr />
 
-      <h5>{intl.formatMessage({ id: "library" })}</h5>
       <Form.Group>
-        <Form.Check
-          id="use-file-metadata"
-          checked={useFileMetadata}
-          label={intl.formatMessage({
-            id: "config.tasks.set_name_date_details_from_metadata_if_present",
-          })}
-          onChange={() => setUseFileMetadata(!useFileMetadata)}
-        />
-        <Form.Check
-          id="strip-file-extension"
-          checked={stripFileExtension}
-          label={intl.formatMessage({
-            id: "config.tasks.dont_include_file_extension_as_part_of_the_title",
-          })}
-          onChange={() => setStripFileExtension(!stripFileExtension)}
-        />
-        <Form.Check
-          id="scan-generate-previews"
-          checked={scanGeneratePreviews}
-          label={intl.formatMessage({
-            id: "config.tasks.generate_video_previews_during_scan",
-          })}
-          onChange={() => setScanGeneratePreviews(!scanGeneratePreviews)}
-        />
-        <div className="d-flex flex-row">
-          <div>↳</div>
+        <h5>{intl.formatMessage({ id: "library" })}</h5>
+        <Form.Group>
+          <h6>{intl.formatMessage({ id: "actions.scan" })}</h6>
           <Form.Check
-            id="scan-generate-image-previews"
-            checked={scanGenerateImagePreviews}
-            disabled={!scanGeneratePreviews}
+            id="use-file-metadata"
+            checked={useFileMetadata}
             label={intl.formatMessage({
-              id: "config.tasks.generate_previews_during_scan",
+              id: "config.tasks.set_name_date_details_from_metadata_if_present",
             })}
-            onChange={() =>
-              setScanGenerateImagePreviews(!scanGenerateImagePreviews)
-            }
-            className="ml-2 flex-grow"
+            onChange={() => setUseFileMetadata(!useFileMetadata)}
           />
-        </div>
-        <Form.Check
-          id="scan-generate-sprites"
-          checked={scanGenerateSprites}
-          label={intl.formatMessage({
-            id: "config.tasks.generate_sprites_during_scan",
-          })}
-          onChange={() => setScanGenerateSprites(!scanGenerateSprites)}
-        />
-        <Form.Check
-          id="scan-generate-phashes"
-          checked={scanGeneratePhashes}
-          label={intl.formatMessage({
-            id: "config.tasks.generate_phashes_during_scan",
-          })}
-          onChange={() => setScanGeneratePhashes(!scanGeneratePhashes)}
-        />
-        <Form.Check
-          id="scan-generate-thumbnails"
-          checked={scanGenerateThumbnails}
-          label={intl.formatMessage({
-            id: "config.tasks.generate_thumbnails_during_scan",
-          })}
-          onChange={() => setScanGenerateThumbnails(!scanGenerateThumbnails)}
-        />
-      </Form.Group>
-      <Form.Group>
-        <Button
-          className="mr-2"
-          variant="secondary"
-          type="submit"
-          onClick={() => onScan()}
-        >
-          <FormattedMessage id="actions.scan" />
-        </Button>
-        <Button
-          variant="secondary"
-          type="submit"
-          onClick={() => setIsScanDialogOpen(true)}
-        >
-          <FormattedMessage id="actions.selective_scan" />
-        </Button>
-        <Form.Text className="text-muted">
-          {intl.formatMessage({ id: "config.tasks.scan_for_content_desc" })}
-        </Form.Text>
-      </Form.Group>
+          <Form.Check
+            id="strip-file-extension"
+            checked={stripFileExtension}
+            label={intl.formatMessage({
+              id:
+                "config.tasks.dont_include_file_extension_as_part_of_the_title",
+            })}
+            onChange={() => setStripFileExtension(!stripFileExtension)}
+          />
+          <Form.Check
+            id="scan-generate-previews"
+            checked={scanGeneratePreviews}
+            label={intl.formatMessage({
+              id: "config.tasks.generate_video_previews_during_scan",
+            })}
+            onChange={() => setScanGeneratePreviews(!scanGeneratePreviews)}
+          />
+          <div className="d-flex flex-row">
+            <div>↳</div>
+            <Form.Check
+              id="scan-generate-image-previews"
+              checked={scanGenerateImagePreviews}
+              disabled={!scanGeneratePreviews}
+              label={intl.formatMessage({
+                id: "config.tasks.generate_previews_during_scan",
+              })}
+              onChange={() =>
+                setScanGenerateImagePreviews(!scanGenerateImagePreviews)
+              }
+              className="ml-2 flex-grow"
+            />
+          </div>
+          <Form.Check
+            id="scan-generate-sprites"
+            checked={scanGenerateSprites}
+            label={intl.formatMessage({
+              id: "config.tasks.generate_sprites_during_scan",
+            })}
+            onChange={() => setScanGenerateSprites(!scanGenerateSprites)}
+          />
+          <Form.Check
+            id="scan-generate-phashes"
+            checked={scanGeneratePhashes}
+            label={intl.formatMessage({
+              id: "config.tasks.generate_phashes_during_scan",
+            })}
+            onChange={() => setScanGeneratePhashes(!scanGeneratePhashes)}
+          />
+          <Form.Check
+            id="scan-generate-thumbnails"
+            checked={scanGenerateThumbnails}
+            label={intl.formatMessage({
+              id: "config.tasks.generate_thumbnails_during_scan",
+            })}
+            onChange={() => setScanGenerateThumbnails(!scanGenerateThumbnails)}
+          />
+        </Form.Group>
+        <Form.Group>
+          <Button
+            className="mr-2"
+            variant="secondary"
+            type="submit"
+            onClick={() => onScan()}
+          >
+            <FormattedMessage id="actions.scan" />
+          </Button>
+          <Button
+            variant="secondary"
+            type="submit"
+            onClick={() => setDialogOpen({ scan: true })}
+          >
+            <FormattedMessage id="actions.selective_scan" />
+          </Button>
+          <Form.Text className="text-muted">
+            {intl.formatMessage({ id: "config.tasks.scan_for_content_desc" })}
+          </Form.Text>
+        </Form.Group>
 
-      <hr />
+        <Form.Group>
+          <h6>
+            <FormattedMessage id="config.tasks.identify.heading" />
+          </h6>
+          <Button
+            className="mr-2"
+            variant="secondary"
+            type="submit"
+            onClick={() => setDialogOpen({ identify: true })}
+          >
+            <FormattedMessage id="actions.identify" />…
+          </Button>
+          <Form.Text className="text-muted">
+            <FormattedMessage id="config.tasks.identify.description" />
+          </Form.Text>
+        </Form.Group>
 
-      <h5>{intl.formatMessage({ id: "config.tasks.auto_tagging" })}</h5>
+        <Form.Group>
+          <h6>{intl.formatMessage({ id: "config.tasks.auto_tagging" })}</h6>
 
-      <Form.Group>
-        <Form.Check
-          id="autotag-performers"
-          checked={autoTagPerformers}
-          label={intl.formatMessage({ id: "performers" })}
-          onChange={() => setAutoTagPerformers(!autoTagPerformers)}
-        />
-        <Form.Check
-          id="autotag-studios"
-          checked={autoTagStudios}
-          label={intl.formatMessage({ id: "studios" })}
-          onChange={() => setAutoTagStudios(!autoTagStudios)}
-        />
-        <Form.Check
-          id="autotag-tags"
-          checked={autoTagTags}
-          label={intl.formatMessage({ id: "tags" })}
-          onChange={() => setAutoTagTags(!autoTagTags)}
-        />
-      </Form.Group>
-      <Form.Group>
-        <Button
-          variant="secondary"
-          type="submit"
-          className="mr-2"
-          onClick={() => onAutoTag()}
-        >
-          <FormattedMessage id="actions.auto_tag" />
-        </Button>
-        <Button
-          variant="secondary"
-          type="submit"
-          onClick={() => setIsAutoTagDialogOpen(true)}
-        >
-          <FormattedMessage id="actions.selective_auto_tag" />
-        </Button>
-        <Form.Text className="text-muted">
-          {intl.formatMessage({
-            id: "config.tasks.auto_tag_based_on_filenames",
-          })}
-        </Form.Text>
+          <Form.Group>
+            <Form.Check
+              id="autotag-performers"
+              checked={autoTagPerformers}
+              label={intl.formatMessage({ id: "performers" })}
+              onChange={() => setAutoTagPerformers(!autoTagPerformers)}
+            />
+            <Form.Check
+              id="autotag-studios"
+              checked={autoTagStudios}
+              label={intl.formatMessage({ id: "studios" })}
+              onChange={() => setAutoTagStudios(!autoTagStudios)}
+            />
+            <Form.Check
+              id="autotag-tags"
+              checked={autoTagTags}
+              label={intl.formatMessage({ id: "tags" })}
+              onChange={() => setAutoTagTags(!autoTagTags)}
+            />
+          </Form.Group>
+          <Form.Group>
+            <Button
+              variant="secondary"
+              type="submit"
+              className="mr-2"
+              onClick={() => onAutoTag()}
+            >
+              <FormattedMessage id="actions.auto_tag" />
+            </Button>
+            <Button
+              variant="secondary"
+              type="submit"
+              onClick={() => setDialogOpen({ autoTag: true })}
+            >
+              <FormattedMessage id="actions.selective_auto_tag" />
+            </Button>
+            <Form.Text className="text-muted">
+              {intl.formatMessage({
+                id: "config.tasks.auto_tag_based_on_filenames",
+              })}
+            </Form.Text>
+          </Form.Group>
+        </Form.Group>
       </Form.Group>
 
       <hr />
@@ -503,7 +545,7 @@ export const SettingsTasksPanel: React.FC = () => {
         <Button
           id="clean"
           variant="danger"
-          onClick={() => setIsCleanAlertOpen(true)}
+          onClick={() => setDialogOpen({ cleanAlert: true })}
         >
           <FormattedMessage id="actions.clean" />
         </Button>
@@ -533,7 +575,7 @@ export const SettingsTasksPanel: React.FC = () => {
         <Button
           id="import"
           variant="danger"
-          onClick={() => setIsImportAlertOpen(true)}
+          onClick={() => setDialogOpen({ importAlert: true })}
         >
           <FormattedMessage id="actions.full_import" />
         </Button>
@@ -546,7 +588,7 @@ export const SettingsTasksPanel: React.FC = () => {
         <Button
           id="partial-import"
           variant="danger"
-          onClick={() => setIsImportDialogOpen(true)}
+          onClick={() => setDialogOpen({ import: true })}
         >
           <FormattedMessage id="actions.import_from_file" />
         </Button>

ui/v2.5/src/components/Shared/FolderSelect/FolderSelect.tsx

@@ -1,6 +1,7 @@
-import React, { useEffect } from "react";
+import React, { useEffect, useState, useMemo } from "react";
 import { FormattedMessage } from "react-intl";
 import { Button, InputGroup, Form } from "react-bootstrap";
+import { debounce } from "lodash";
 import { LoadingIndicator } from "src/components/Shared";
 import { useDirectory } from "src/core/StashService";
 
@@ -17,22 +18,43 @@ export const FolderSelect: React.FC<IProps> = ({
   defaultDirectories,
   appendButton,
 }) => {
-  const { data, error, loading } = useDirectory(currentDirectory);
+  const [debouncedDirectory, setDebouncedDirectory] = useState(
+    currentDirectory
+  );
+  const { data, error, loading } = useDirectory(debouncedDirectory);
 
   const selectableDirectories: string[] = currentDirectory
     ? data?.directory.directories ?? defaultDirectories ?? []
     : defaultDirectories ?? [];
 
+  const debouncedSetDirectory = useMemo(
+    () =>
+      debounce((input: string) => {
+        setDebouncedDirectory(input);
+      }, 250),
+    []
+  );
+
   useEffect(() => {
     if (currentDirectory === "" && !defaultDirectories && data?.directory.path)
       setCurrentDirectory(data.directory.path);
   }, [currentDirectory, setCurrentDirectory, data, defaultDirectories]);
 
+  function setInstant(value: string) {
+    setCurrentDirectory(value);
+    setDebouncedDirectory(value);
+  }
+
+  function setDebounced(value: string) {
+    setCurrentDirectory(value);
+    debouncedSetDirectory(value);
+  }
+
   function goUp() {
     if (defaultDirectories?.includes(currentDirectory)) {
-      setCurrentDirectory("");
+      setInstant("");
     } else if (data?.directory.parent) {
-      setCurrentDirectory(data.directory.parent);
+      setInstant(data.directory.parent);
     }
   }
 
@@ -51,9 +73,9 @@ export const FolderSelect: React.FC<IProps> = ({
       <InputGroup>
         <Form.Control
           placeholder="File path"
-          onChange={(e: React.ChangeEvent<HTMLInputElement>) =>
-            setCurrentDirectory(e.currentTarget.value)
-          }
+          onChange={(e: React.ChangeEvent<HTMLInputElement>) => {
+            setDebounced(e.currentTarget.value);
+          }}
           value={currentDirectory}
           spellCheck={false}
         />
@@ -71,7 +93,7 @@ export const FolderSelect: React.FC<IProps> = ({
         {selectableDirectories.map((path) => {
           return (
             <li key={path} className="folder-list-item">
-              <Button variant="link" onClick={() => setCurrentDirectory(path)}>
+              <Button variant="link" onClick={() => setInstant(path)}>
                 {path}
               </Button>
             </li>

ui/v2.5/src/components/Shared/IndeterminateCheckbox.tsx

@@ -0,0 +1,55 @@
+import React, { useEffect } from "react";
+import { Form, FormCheckProps } from "react-bootstrap";
+
+const useIndeterminate = (
+  ref: React.RefObject<HTMLInputElement>,
+  value: boolean | undefined
+) => {
+  useEffect(() => {
+    if (ref.current) {
+      // eslint-disable-next-line no-param-reassign
+      ref.current.indeterminate = value === undefined;
+    }
+  }, [ref, value]);
+};
+
+interface IIndeterminateCheckbox extends FormCheckProps {
+  setChecked: (v: boolean | undefined) => void;
+  allowIndeterminate?: boolean;
+  indeterminateClassname?: string;
+}
+
+export const IndeterminateCheckbox: React.FC<IIndeterminateCheckbox> = ({
+  checked,
+  setChecked,
+  allowIndeterminate,
+  indeterminateClassname,
+  ...props
+}) => {
+  const ref = React.createRef<HTMLInputElement>();
+
+  useIndeterminate(ref, checked);
+
+  function cycleState() {
+    const undefAllowed = allowIndeterminate ?? true;
+    if (undefAllowed && checked) {
+      return undefined;
+    }
+    if ((!undefAllowed && checked) || checked === undefined) {
+      return false;
+    }
+    return true;
+  }
+
+  return (
+    <Form.Check
+      {...props}
+      className={`${props.className ?? ""} ${
+        checked === undefined ? indeterminateClassname : ""
+      }`}
+      ref={ref}
+      checked={checked}
+      onChange={() => setChecked(cycleState())}
+    />
+  );
+};

ui/v2.5/src/components/Shared/Modal.tsx

@@ -21,6 +21,8 @@ interface IModal {
   disabled?: boolean;
   modalProps?: ModalProps;
   dialogClassName?: string;
+  footerButtons?: React.ReactNode;
+  leftFooterButtons?: React.ReactNode;
 }
 
 const defaultOnHide = () => {};
@@ -37,8 +39,11 @@ const ModalComponent: React.FC<IModal> = ({
   disabled,
   modalProps,
   dialogClassName,
+  footerButtons,
+  leftFooterButtons,
 }) => (
   <Modal
+    className="ModalComponent"
     keyboard={false}
     onHide={onHide ?? defaultOnHide}
     show={show}
@@ -50,14 +55,16 @@ const ModalComponent: React.FC<IModal> = ({
       <span>{header ?? ""}</span>
     </Modal.Header>
     <Modal.Body>{children}</Modal.Body>
-    <Modal.Footer>
+    <Modal.Footer className="ModalFooter">
+      <div>{leftFooterButtons}</div>
       <div>
+        {footerButtons}
         {cancel ? (
           <Button
             disabled={isRunning}
             variant={cancel.variant ?? "primary"}
             onClick={cancel.onClick}
-            className="mr-2"
+            className="ml-2"
           >
             {cancel.text ?? (
               <FormattedMessage

ui/v2.5/src/components/Shared/ThreeStateCheckbox.tsx

@@ -0,0 +1,47 @@
+import React from "react";
+import { Button } from "react-bootstrap";
+import { Icon } from ".";
+
+interface IThreeStateCheckbox {
+  value: boolean | undefined;
+  setValue: (v: boolean | undefined) => void;
+  allowUndefined?: boolean;
+  label?: React.ReactNode;
+  disabled?: boolean;
+}
+
+export const ThreeStateCheckbox: React.FC<IThreeStateCheckbox> = ({
+  value,
+  setValue,
+  allowUndefined,
+  label,
+  disabled = false,
+}) => {
+  function cycleState() {
+    const undefAllowed = allowUndefined ?? true;
+    if (undefAllowed && value) {
+      return undefined;
+    }
+    if ((!undefAllowed && value) || value === undefined) {
+      return false;
+    }
+    return true;
+  }
+
+  const icon = value === undefined ? "minus" : value ? "check" : "times";
+  const labelClassName =
+    value === undefined ? "unset" : value ? "checked" : "not-checked";
+
+  return (
+    <span className={`three-state-checkbox ${labelClassName}`}>
+      <Button
+        onClick={() => setValue(cycleState())}
+        className="minimal"
+        disabled={disabled}
+      >
+        <Icon icon={icon} className="fa-fw" />
+      </Button>
+      <span className="label">{label}</span>
+    </span>
+  );
+};

ui/v2.5/src/components/Shared/index.ts

@@ -17,5 +17,6 @@ export { GridCard } from "./GridCard";
 export { RatingStars } from "./RatingStars";
 export { ExportDialog } from "./ExportDialog";
 export { default as DeleteEntityDialog } from "./DeleteEntityDialog";
+export { IndeterminateCheckbox } from "./IndeterminateCheckbox";
 export { OperationButton } from "./OperationButton";
 export const TITLE_SUFFIX = " | Stash";

ui/v2.5/src/components/Shared/styles.scss

@@ -209,9 +209,48 @@ button.collapse-button.btn-primary:not(:disabled):not(.disabled):active {
   }
 }
 
+.three-state-checkbox {
+  align-items: center;
+  display: flex;
+
+  button.btn {
+    font-size: 12.67px;
+    margin-left: -0.2em;
+    margin-right: 0.25rem;
+    padding: 0;
+
+    &:not(:disabled):active,
+    &:not(:disabled):active:focus,
+    &:not(:disabled):hover,
+    &:not(:disabled):not(:hover) {
+      background-color: initial;
+      box-shadow: none;
+    }
+  }
+
+  &.unset {
+    .label {
+      color: #bfccd6;
+      text-decoration: line-through;
+    }
+  }
+
+  &.checked svg {
+    color: #0f9960;
+  }
+
+  &.not-checked svg {
+    color: #db3737;
+  }
+}
+
 .input-group-prepend {
   .btn {
     border-bottom-right-radius: 0;
     border-top-right-radius: 0;
   }
 }
+
+.ModalComponent .modal-footer {
+  justify-content: space-between;
+}

ui/v2.5/src/core/StashService.ts

@@ -757,6 +757,12 @@ export const useGenerateAPIKey = () =>
     update: deleteCache([GQL.ConfigurationDocument]),
   });
 
+export const useConfigureDefaults = () =>
+  GQL.useConfigureDefaultsMutation({
+    refetchQueries: getQueryNames([GQL.ConfigurationDocument]),
+    update: deleteCache([GQL.ConfigurationDocument]),
+  });
+
 export const useJobsSubscribe = () => GQL.useJobsSubscribeSubscription();
 
 export const useConfigureDLNA = () =>
@@ -1001,6 +1007,12 @@ export const mutateMetadataClean = (input: GQL.CleanMetadataInput) =>
     variables: { input },
   });
 
+export const mutateMetadataIdentify = (input: GQL.IdentifyMetadataInput) =>
+  client.mutate<GQL.MetadataIdentifyMutation>({
+    mutation: GQL.MetadataIdentifyDocument,
+    variables: { input },
+  });
+
 export const mutateMigrateHashNaming = () =>
   client.mutate<GQL.MigrateHashNamingMutation>({
     mutation: GQL.MigrateHashNamingDocument,

ui/v2.5/src/docs/en/Identify.md

@@ -0,0 +1,31 @@
+# Identify
+
+This task iterates through your Scenes and attempts to identify the scene using a selection of scraping sources.
+
+This task accepts one or more scraper sources. Valid scraper sources for the Identify task are stash-box instances, and scene scrapers which support scraping via Scene Fragment. The order of the sources may be rearranged.
+
+For each Scene, the Identify task iterates through the scraper sources, in the order provided, and tries to identify the scene using each source. If a result is found in a source, then the Scene is updated, and no further sources are checked for that scene.
+
+## Options
+
+The following options can be set:
+
+| Option | Description |
+|--------|-------------|
+| Include male performers | If false, then male performers will not be created or set on scenes. |
+| Set cover images | If false, then scene cover images will not be modified. |
+| Set organised flag | If true, the organised flag is set to true when a scene is organised. |
+
+Field specific options may be set as well. Each field may have a Strategy. The behaviour for each strategy value is as follows:
+
+| Strategy | Description |
+|----------|-------------|
+| Ignore | Not set. |
+| Overwrite | Overwrite existing value. |
+| Merge (*default*) | For multi-value fields, adds to existing values. For single-value fields, only sets if not already set. |
+
+For Studio, Performers and Tags, an option is also available to Create Missing objects. This is false by default. When true, if a Studio/Performer/Tag is included during the identification process and does not exist in the system, then it will be created.
+
+Default Options are applied to all sources unless overridden in specific source options. 
+
+The result of the identification process for each scene is output to the log.

ui/v2.5/src/index.scss

@@ -20,6 +20,7 @@
 @import "../node_modules/flag-icon-css/css/flag-icon.min.css";
 @import "src/components/Tagger/styles.scss";
 @import "src/hooks/Lightbox/lightbox.scss";
+@import "src/components/Dialogs/IdentifyDialog/styles.scss";
 
 /* stylelint-disable */
 #root {

ui/v2.5/src/locales/en-GB.json

@@ -41,6 +41,7 @@
     "generate_thumb_from_current": "Generate thumbnail from current",
     "hash_migration": "hash migration",
     "hide": "Hide",
+    "identify": "Identify",
     "import": "Import…",
     "import_from_file": "Import from file",
     "merge": "Merge",
@@ -96,6 +97,7 @@
   "actions_name": "Actions",
   "age": "Age",
   "aliases": "Aliases",
+  "all": "all",
   "also_known_as": "Also known as",
   "ascending": "Ascending",
   "average_resolution": "Average Resolution",
@@ -271,6 +273,7 @@
       "entity_scrapers": "{entityType} scrapers",
       "excluded_tag_patterns_desc": "Regexps of tag names to exclude from scraping results",
       "excluded_tag_patterns_head": "Excluded Tag Patterns",
+      "scraper": "Scraper",
       "scrapers": "Scrapers",
       "search_by_name": "Search by name",
       "supported_types": "Supported types",
@@ -302,6 +305,29 @@
       "generate_video_previews_during_scan": "Generate previews during scan (video previews which play when hovering over a scene)",
       "generate_thumbnails_during_scan": "Generate thumbnails for images during scan.",
       "generated_content": "Generated Content",
+      "identify": {
+        "and_create_missing": "and create missing",
+        "create_missing": "Create missing",
+        "heading": "Identify",
+        "description": "Automatically set scene metadata using stash-box and scraper sources.",
+        "default_options": "Default Options",
+        "explicit_set_description": "The following options will be used where not overridden in the source-specific options.",
+        "field_behaviour": "{strategy} {field}",
+        "field_options": "Field Options",
+        "field_strategies": {
+          "ignore": "Ignore",
+          "merge": "Merge",
+          "overwrite": "Overwrite"
+        },
+        "identifying_scenes": "Identifying {num} {scene}",
+        "identifying_from_paths": "Identifying scenes from the following paths",
+        "include_male_performers": "Include male performers",
+        "set_cover_images": "Set cover images",
+        "set_organized": "Set organised flag",
+        "source_options": "{source} Options",
+        "source": "Source",
+        "sources": "Sources"
+      },
       "import_from_exported_json": "Import from exported JSON in the metadata directory. Wipes the existing database.",
       "incremental_import": "Incremental import from a supplied export zip file.",
       "job_queue": "Job Queue",
@@ -572,6 +598,7 @@
   "ethnicity": "Ethnicity",
   "eye_color": "Eye Colour",
   "fake_tits": "Fake Tits",
+  "false": "False",
   "favourite": "Favourite",
   "file_info": "File Info",
   "file_mod_time": "File Modification Time",
@@ -668,6 +695,7 @@
   "settings": "Settings",
   "sub_tag_of": "Sub-tag of {parent}",
   "stash_id": "Stash ID",
+  "stash_ids": "Stash IDs",
   "status": "Status: {statusText}",
   "studio": "Studio",
   "studio_depth": "Levels (empty for all)",
@@ -693,10 +721,12 @@
     "updated_entity": "Updated {entity}"
   },
   "total": "Total",
+  "true": "True",
   "twitter": "Twitter",
   "up-dir": "Up a directory",
   "updated_at": "Updated At",
   "url": "URL",
+  "use_default": "Use default",
   "weight": "Weight",
   "years_old": "years old",
   "stats": {

ui/v2.5/src/utils/data.ts

@@ -1,2 +1,11 @@
 export const filterData = <T>(data?: (T | null | undefined)[] | null) =>
   data ? (data.filter((item) => item) as T[]) : [];
+
+interface ITypename {
+  __typename?: string;
+}
+
+export function withoutTypename<T extends ITypename>(o: T) {
+  const { __typename, ...ret } = o;
+  return ret;
+}