From 7623e69de394da783418d80764646fc4a2567c31 Mon Sep 17 00:00:00 2001
From: Neeraj Gupta <254676+ua741@users.noreply.github.com>
Date: Fri, 10 May 2024 13:07:24 +0530
Subject: [PATCH] [mob] Return list instead of set

---
 mobile/lib/face/db.dart                       | 85 ++++++++++---------
 .../face_ml/face_ml_service.dart              | 16 ++--
 2 files changed, 52 insertions(+), 49 deletions(-)

diff --git a/mobile/lib/face/db.dart b/mobile/lib/face/db.dart
index 2fd1b6e1d..b2173552d 100644
--- a/mobile/lib/face/db.dart
+++ b/mobile/lib/face/db.dart
@@ -511,56 +511,61 @@ class FaceMLDataDB {
     });
   }
 
-  Future<Set<FaceInfoForClustering>> getFaceInfoForClustering({
+  Future<List<FaceInfoForClustering>> getFaceInfoForClustering({
     double minScore = kMinimumQualityFaceScore,
     int minClarity = kLaplacianHardThreshold,
     int maxFaces = 20000,
     int offset = 0,
     int batchSize = 10000,
   }) async {
-    final EnteWatch w = EnteWatch("getFaceEmbeddingMap")..start();
-    w.logAndReset(
-      'reading as float offset: $offset, maxFaces: $maxFaces, batchSize: $batchSize',
-    );
-    final db = await instance.asyncDB;
-
-    final Set<FaceInfoForClustering> result = {};
-    while (true) {
-      // Query a batch of rows
-      final List<Map<String, dynamic>> maps = await db.getAll(
-        'SELECT $faceIDColumn, $faceEmbeddingBlob, $faceScore, $faceBlur, $isSideways FROM $facesTable'
-        ' WHERE $faceScore > $minScore AND $faceBlur > $minClarity'
-        ' ORDER BY $faceIDColumn'
-        ' DESC LIMIT $batchSize OFFSET $offset',
+    try {
+      final EnteWatch w = EnteWatch("getFaceEmbeddingMap")..start();
+      w.logAndReset(
+        'reading as float offset: $offset, maxFaces: $maxFaces, batchSize: $batchSize',
       );
-      // Break the loop if no more rows
-      if (maps.isEmpty) {
-        break;
-      }
-      final List<String> faceIds = [];
-      for (final map in maps) {
-        faceIds.add(map[faceIDColumn] as String);
-      }
-      final faceIdToClusterId = await getFaceIdsToClusterIds(faceIds);
-      for (final map in maps) {
-        final faceID = map[faceIDColumn] as String;
-        final faceInfo = FaceInfoForClustering(
-          faceID: faceID,
-          clusterId: faceIdToClusterId[faceID],
-          embeddingBytes: map[faceEmbeddingBlob] as Uint8List,
-          faceScore: map[faceScore] as double,
-          blurValue: map[faceBlur] as double,
-          isSideways: (map[isSideways] as int) == 1,
+      final db = await instance.asyncDB;
+
+      final List<FaceInfoForClustering> result = {};
+      while (true) {
+        // Query a batch of rows
+        final List<Map<String, dynamic>> maps = await db.getAll(
+          'SELECT $faceIDColumn, $faceEmbeddingBlob, $faceScore, $faceBlur, $isSideways FROM $facesTable'
+          ' WHERE $faceScore > $minScore AND $faceBlur > $minClarity'
+          ' ORDER BY $faceIDColumn'
+          ' DESC LIMIT $batchSize OFFSET $offset',
         );
-        result.add(faceInfo);
+        // Break the loop if no more rows
+        if (maps.isEmpty) {
+          break;
+        }
+        final List<String> faceIds = [];
+        for (final map in maps) {
+          faceIds.add(map[faceIDColumn] as String);
+        }
+        final faceIdToClusterId = await getFaceIdsToClusterIds(faceIds);
+        for (final map in maps) {
+          final faceID = map[faceIDColumn] as String;
+          final faceInfo = FaceInfoForClustering(
+            faceID: faceID,
+            clusterId: faceIdToClusterId[faceID],
+            embeddingBytes: map[faceEmbeddingBlob] as Uint8List,
+            faceScore: map[faceScore] as double,
+            blurValue: map[faceBlur] as double,
+            isSideways: (map[isSideways] as int) == 1,
+          );
+          result.add(faceInfo);
+        }
+        if (result.length >= maxFaces) {
+          break;
+        }
+        offset += batchSize;
       }
-      if (result.length >= maxFaces) {
-        break;
-      }
-      offset += batchSize;
+      w.stopWithLog('done reading face embeddings ${result.length}');
+      return result;
+    } catch (e) {
+      _logger.severe('Error in getFaceInfoForClustering', e);
+      rethrow;
     }
-    w.stopWithLog('done reading face embeddings ${result.length}');
-    return result;
   }
 
   /// Returns a map of faceID to record of clusterId and faceEmbeddingBlob
diff --git a/mobile/lib/services/machine_learning/face_ml/face_ml_service.dart b/mobile/lib/services/machine_learning/face_ml/face_ml_service.dart
index b39a09af1..d36015818 100644
--- a/mobile/lib/services/machine_learning/face_ml/face_ml_service.dart
+++ b/mobile/lib/services/machine_learning/face_ml/face_ml_service.dart
@@ -71,7 +71,9 @@ class FaceMlService {
 
   // singleton pattern
   FaceMlService._privateConstructor();
+
   static final instance = FaceMlService._privateConstructor();
+
   factory FaceMlService() => instance;
 
   final _initLock = Lock();
@@ -302,19 +304,15 @@ class FaceMlService {
       // Get a sense of the total number of faces in the database
       final int totalFaces = await FaceMLDataDB.instance
           .getTotalFaceCount(minFaceScore: minFaceScore);
-
-      // read the creation times from Files DB, in a map from fileID to creation time
       final fileIDToCreationTime =
           await FilesDB.instance.getFileIDToCreationTime();
-
       final startEmbeddingFetch = DateTime.now();
       // read all embeddings
-      final allFaceInfoForClustering = await FaceMLDataDB.instance
-          .getFaceInfoForClustering(
-            minScore: minFaceScore,
-            maxFaces: totalFaces,
-          )
-          .then((set) => set.toList());
+      final allFaceInfoForClustering =
+          await FaceMLDataDB.instance.getFaceInfoForClustering(
+        minScore: minFaceScore,
+        maxFaces: totalFaces,
+      );
       // sort the embeddings based on file creation time, oldest first
       allFaceInfoForClustering.sort((a, b) {
         final aFileId = getFileIdFromFaceId(a.faceID);