Make ml related classes generic

mobile/lib/db/ml/base.dart

@@ -112,7 +112,7 @@ abstract class IMLDataDB<T> {
   Future<List<EmbeddingVector>> getAllClipVectors();
   Future<Map<int, int>> clipIndexedFileWithVersion();
   Future<int> getClipIndexedFileCount({int minimumMlVersion});
-  Future<void> putClip(List<ClipEmbedding> embeddings);
+  Future<void> putClip(List<ClipEmbedding<T>> embeddings);
   Future<void> deleteClipEmbeddings(List<T> fileIDs);
   Future<void> deleteClipIndexes();
 }

mobile/lib/models/ml/clip.dart

@@ -2,8 +2,8 @@ import "dart:convert";
 
 import "package:photos/models/ml/ml_versions.dart";
 
-class ClipEmbedding {
-  final int fileID;
+class ClipEmbedding<T> {
+  final T fileID;
   final List<double> embedding;
   int version;
 
@@ -15,7 +15,7 @@ class ClipEmbedding {
     required this.version,
   });
 
-  factory ClipEmbedding.empty(int fileID) {
+  factory ClipEmbedding.empty(T fileID) {
     return ClipEmbedding(
       fileID: fileID,
       embedding: <double>[],

mobile/lib/models/ml/face/face.dart

@@ -16,7 +16,7 @@ class FileInfo {
   });
 }
 
-class Face {
+class Face<T> {
   final String faceID;
   final List<double> embedding;
   Detection detection;
@@ -26,7 +26,7 @@ class Face {
   ///#region Local DB fields
   // This is not stored on the server, using it for local DB row
   FileInfo? fileInfo;
-  final int fileID;
+  final T fileID;
 
   ///#endregion
 
@@ -48,7 +48,7 @@ class Face {
 
   factory Face.fromFaceResult(
     FaceResult faceResult,
-    int fileID,
+    T fileID,
     Dimensions decodedDimensions,
   ) {
     final detection = Detection(
@@ -81,7 +81,7 @@ class Face {
     );
   }
 
-  factory Face.empty(int fileID, {bool error = false}) {
+  static Face<T> empty<T>(T fileID, {bool error = false}) {
     return Face(
       "${fileID}_0_0_0_0",
       fileID,
@@ -92,9 +92,9 @@ class Face {
     );
   }
 
-  factory Face.fromJson(Map<String, dynamic> json) {
+  static Face<T> fromJson<T>(Map<String, dynamic> json) {
     final String faceID = json['faceID'] as String;
-    final int fileID = getFileIdFromFaceId<int>(faceID);
+    final T fileID = getFileIdFromFaceId(faceID);
     return Face(
       faceID,
       fileID,

mobile/lib/services/machine_learning/face_ml/face_detection/detection.dart

@@ -162,7 +162,8 @@ class FaceDetectionRelative extends Detection {
     );
   }
 
-  String toFaceID({required int fileID}) {
+  // todo: neeraj review later #localgallery
+  String toFaceID({required dynamic fileID}) {
     // Assert that the values are within the expected range
     assert(
       (xMinBox >= 0 && xMinBox <= 1) &&

mobile/lib/services/machine_learning/face_ml/face_recognition_service.dart

@@ -71,8 +71,8 @@ class FaceRecognitionService {
     }
   }
 
-  static Future<List<FaceResult>> runFacesPipeline(
-    int enteFileID,
+  static Future<List<FaceResult>> runFacesPipeline<T>(
+    T enteFileID,
     Image image,
     Uint8List rawRgbaBytes,
     int faceDetectionAddress,
@@ -131,8 +131,8 @@ class FaceRecognitionService {
   }
 
   /// Runs face recognition on the given image data.
-  static Future<List<FaceDetectionRelative>> _detectFacesSync(
-    int fileID,
+  static Future<List<FaceDetectionRelative>> _detectFacesSync<T>(
+    T fileID,
     Image image,
     Uint8List rawRgbaBytes,
     int interpreterAddress,

mobile/lib/services/machine_learning/ml_indexing_isolate.dart

@@ -55,11 +55,11 @@ class MLIndexingIsolate extends SuperIsolate {
   factory MLIndexingIsolate() => instance;
 
   /// Analyzes the given image data by running the full pipeline for faces, using [analyzeImageStatic] in the isolate.
-  Future<MLResult?> analyzeImage(
+  Future<MLResult<T>?> analyzeImage<T>(
     FileMLInstruction instruction,
     String filePath,
   ) async {
-    late MLResult result;
+    late MLResult<T> result;
 
     try {
       final resultJsonString =
@@ -78,7 +78,7 @@ class MLIndexingIsolate extends SuperIsolate {
         }
         return null;
       }
-      result = MLResult.fromJsonString(resultJsonString);
+      result = MLResult.fromJsonString<T>(resultJsonString);
     } catch (e, s) {
       _logger.severe(
         "Could not analyze image with ID ${instruction.file.uploadedFileID} \n",

mobile/lib/services/machine_learning/ml_result.dart

@@ -7,8 +7,8 @@ import 'package:photos/services/machine_learning/face_ml/face_alignment/alignmen
 import 'package:photos/services/machine_learning/face_ml/face_detection/detection.dart';
 import 'package:photos/services/machine_learning/face_ml/face_filtering/face_filtering_constants.dart';
 
-class MLResult {
-  int fileId;
+class MLResult<T> {
+  T fileId;
 
   List<FaceResult>? faces = <FaceResult>[];
   ClipResult? clip;
@@ -20,14 +20,14 @@ class MLResult {
   bool get clipRan => clip != null;
 
   MLResult({
-    this.fileId = -1,
+    required this.fileId,
     this.faces,
     this.clip,
     this.decodedImageSize = const Dimensions(width: -1, height: -1),
   });
 
   MLResult.fromEnteFileID(
-    fileID, {
+    T fileID, {
     this.decodedImageSize = const Dimensions(width: -1, height: -1),
   }) : fileId = fileID;
 
@@ -43,7 +43,9 @@ class MLResult {
 
   String toJsonString() => jsonEncode(_toJson());
 
-  static MLResult _fromJson(Map<String, dynamic> json) {
+  static MLResult<T> _fromJson<T>(
+    Map<String, dynamic> json,
+  ) {
     return MLResult(
       fileId: json['fileId'],
       faces: json['faces'] != null
@@ -70,13 +72,13 @@ class MLResult {
     );
   }
 
-  static MLResult fromJsonString(String jsonString) {
-    return _fromJson(jsonDecode(jsonString));
+  static MLResult<T> fromJsonString<T>(String jsonString) {
+    return _fromJson<T>(jsonDecode(jsonString));
   }
 }
 
-class ClipResult {
-  final int fileID;
+class ClipResult<T> {
+  final T fileID;
   final Embedding embedding;
 
   ClipResult({
@@ -97,12 +99,12 @@ class ClipResult {
   }
 }
 
-class FaceResult {
+class FaceResult<T> {
   late FaceDetectionRelative detection;
   late double blurValue;
   late AlignmentResult alignment;
   late Embedding embedding;
-  late int fileId;
+  late T fileId;
   late String faceId;
 
   bool get isBlurry => blurValue < kLaplacianHardThreshold;
@@ -118,7 +120,7 @@ class FaceResult {
 
   FaceResult.fromFaceDetection(
     FaceDetectionRelative faceDetection,
-    int fileID,
+    T fileID,
   ) {
     fileId = fileID;
     faceId = faceDetection.toFaceID(fileID: fileID);
@@ -146,7 +148,7 @@ class FaceResult {
   }
 }
 
-T getFileIdFromFaceId<T extends Object>(String faceId) {
+T getFileIdFromFaceId<T>(String faceId) {
   final String faceIdSplit = faceId.substring(0, faceId.indexOf('_'));
   if (T == int) {
     return int.parse(faceIdSplit) as T;

mobile/lib/services/machine_learning/ml_service.dart

@@ -10,6 +10,7 @@ import "package:photos/db/files_db.dart";
 import "package:photos/db/ml/db.dart";
 import "package:photos/events/machine_learning_control_event.dart";
 import "package:photos/events/people_changed_event.dart";
+import "package:photos/models/ml/clip.dart";
 import "package:photos/models/ml/face/face.dart";
 import "package:photos/models/ml/ml_versions.dart";
 import "package:photos/service_locator.dart";
@@ -20,7 +21,6 @@ import "package:photos/services/machine_learning/face_ml/face_clustering/face_db
 import "package:photos/services/machine_learning/face_ml/person/person_service.dart";
 import "package:photos/services/machine_learning/ml_indexing_isolate.dart";
 import 'package:photos/services/machine_learning/ml_result.dart';
-import "package:photos/services/machine_learning/semantic_search/semantic_search_service.dart";
 import "package:photos/utils/ml_util.dart";
 import "package:photos/utils/network_util.dart";
 import "package:photos/utils/ram_check_util.dart";
@@ -428,13 +428,14 @@ class MLService {
     }
   }
 
-  Future<bool> processImage(FileMLInstruction instruction) async {
+  Future<bool> processImage<T>(FileMLInstruction instruction) async {
     bool actuallyRanML = false;
 
     try {
       final String filePath = await getImagePathForML(instruction.file);
 
-      final MLResult? result = await MLIndexingIsolate.instance.analyzeImage(
+      final MLResult<T>? result =
+          await MLIndexingIsolate.instance.analyzeImage<T>(
         instruction,
         filePath,
       );
@@ -502,9 +503,13 @@ class MLService {
       // Storing results locally
       if (result.facesRan) await mlDataDB.bulkInsertFaces(faces);
       if (result.clipRan) {
-        await SemanticSearchService.instance.storeClipImageResult(
-          result.clip!,
-        );
+        await mlDataDB.putClip([
+          ClipEmbedding<int>(
+            fileID: result.clip!.fileID,
+            embedding: result.clip!.embedding,
+            version: clipMlVersion,
+          ),
+        ]);
       }
       _logger.info("ML results for fileID ${result.fileId} stored locally");
       return actuallyRanML;
@@ -526,9 +531,9 @@ class MLService {
         await mlDataDB.bulkInsertFaces(
           [Face.empty(instruction.file.uploadedFileID!, error: true)],
         );
-        await SemanticSearchService.instance.storeEmptyClipImageResult(
-          instruction.file,
-        );
+
+        await mlDataDB
+            .putClip([ClipEmbedding.empty(instruction.file.uploadedFileID!)]);
         return true;
       }
       _logger.severe(

mobile/lib/services/machine_learning/semantic_search/semantic_search_service.dart

@@ -12,7 +12,6 @@ import "package:photos/db/ml/db.dart";
 import 'package:photos/events/embedding_updated_event.dart';
 import "package:photos/models/file/file.dart";
 import "package:photos/models/ml/clip.dart";
-import "package:photos/models/ml/ml_versions.dart";
 import "package:photos/service_locator.dart";
 import "package:photos/services/collections_service.dart";
 import "package:photos/services/machine_learning/ml_computer.dart";
@@ -235,20 +234,6 @@ class SemanticSearchService {
     _logger.info("Clip text model loaded");
   }
 
-  Future<void> storeClipImageResult(ClipResult clipResult) async {
-    final embedding = ClipEmbedding(
-      fileID: clipResult.fileID,
-      embedding: clipResult.embedding,
-      version: clipMlVersion,
-    );
-    await mlDataDB.putClip([embedding]);
-  }
-
-  Future<void> storeEmptyClipImageResult(EnteFile entefile) async {
-    final embedding = ClipEmbedding.empty(entefile.uploadedFileID!);
-    await mlDataDB.putClip([embedding]);
-  }
-
   Future<List<double>> _getTextEmbedding(String query) async {
     _logger.info("Searching for ${kDebugMode ? query : ''}");
     final cachedResult = _queryEmbeddingCache.get(query);
@@ -292,8 +277,8 @@ class SemanticSearchService {
     });
   }
 
-  static Future<ClipResult> runClipImage(
-    int enteFileID,
+  static Future<ClipResult<T>> runClipImage<T>(
+    T enteFileID,
     Image image,
     Uint8List rawRgbaBytes,
     int clipImageAddress,

mobile/lib/utils/ml_util.dart

@@ -221,7 +221,7 @@ Stream<List<FileMLInstruction>> fetchEmbeddingsAndInstructions(
           fileMl.getClipEmbeddingIfCompatible(clipMlVersion);
       if (remoteClipEmbedding != null) {
         clipEmbeddings.add(
-          ClipEmbedding(
+          ClipEmbedding<int>(
             fileID: fileMl.fileID,
             embedding: remoteClipEmbedding.embedding,
             version: remoteClipEmbedding.version,
@@ -395,9 +395,9 @@ void normalizeEmbedding(List<double> embedding) {
   }
 }
 
-Future<MLResult> analyzeImageStatic(Map args) async {
+Future<MLResult<T>> analyzeImageStatic<T>(Map args) async {
   try {
-    final int enteFileID = args["enteFileID"] as int;
+    final T enteFileID = args["enteFileID"] as T;
     final String imagePath = args["filePath"] as String;
     final bool runFaces = args["runFaces"] as bool;
     final bool runClip = args["runClip"] as bool;
@@ -422,7 +422,7 @@ Future<MLResult> analyzeImageStatic(Map args) async {
     String faceMsString = "", clipMsString = "";
     final pipelines = await Future.wait([
       runFaces
-          ? FaceRecognitionService.runFacesPipeline(
+          ? FaceRecognitionService.runFacesPipeline<T>(
               enteFileID,
               image,
               rawRgbaBytes,
@@ -435,7 +435,7 @@ Future<MLResult> analyzeImageStatic(Map args) async {
             })
           : Future.value(null),
       runClip
-          ? SemanticSearchService.runClipImage(
+          ? SemanticSearchService.runClipImage<T>(
               enteFileID,
               image,
               rawRgbaBytes,