endoreg-db 0.8.8.9__py3-none-any.whl → 0.8.9.10__py3-none-any.whl

endoreg_db/management/commands/validate_video_files.py

@@ -1,36 +1,39 @@
 """
 Django management command to validate video file existence and accessibility.
 """
+
+import logging
+from pathlib import Path
+
 from django.core.management.base import BaseCommand
 
 from endoreg_db.models import VideoFile
-from pathlib import Path
-import logging
 
 logger = logging.getLogger(__name__)
 
 # TODO Review if this is still used. Delete if not.
 
+
 class Command(BaseCommand):
-    help = 'Validate video file existence and accessibility'
+    help = "Validate video file existence and accessibility"
 
     def add_arguments(self, parser):
         parser.add_argument(
-            '--video-id',
+            "--video-id",
             type=int,
-            help='Check specific video ID',
+            help="Check specific video ID",
         )
         parser.add_argument(
-            '--verbose',
-            action='store_true',
-            help='Enable verbose output',
+            "--verbose",
+            action="store_true",
+            help="Enable verbose output",
         )
 
     def handle(self, *args, **options):
         """Validate video files and their accessibility."""
-        verbose = options['verbose']
-        video_id = options.get('video_id')
-        fix_missing = options['fix_missing']
+        verbose = options["verbose"]
+        video_id = options.get("video_id")
+        fix_missing = options["fix_missing", False]
 
         if verbose:
             self.stdout.write(self.style.SUCCESS("Starting video validation..."))
@@ -41,7 +44,9 @@ class Command(BaseCommand):
                 videos = [VideoFile.objects.get(pk=video_id)]
                 self.stdout.write(f"Checking specific video ID: {video_id}")
             except VideoFile.DoesNotExist:
-                self.stdout.write(self.style.ERROR(f"Video with ID {video_id} not found"))
+                self.stdout.write(
+                    self.style.ERROR(f"Video with ID {video_id} not found")
+                )
                 return
         else:
             videos = VideoFile.objects.all()
@@ -53,19 +58,19 @@ class Command(BaseCommand):
 
         for video in videos:
             video_status = self.check_video_file(video, verbose)
-            
-            if video_status['status'] == 'missing':
+
+            if video_status["status"] == "missing":
                 missing_files.append(video_status)
-            elif video_status['status'] == 'corrupted':
+            elif video_status["status"] == "corrupted":
                 corrupted_files.append(video_status)
             else:
                 accessible_files.append(video_status)
 
         # Report results
-        self.stdout.write("\n" + "="*60)
+        self.stdout.write("\n" + "=" * 60)
         self.stdout.write(self.style.SUCCESS("VALIDATION COMPLETE"))
-        self.stdout.write("="*60)
-        
+        self.stdout.write("=" * 60)
+
         self.stdout.write(f"✅ Accessible videos: {len(accessible_files)}")
         self.stdout.write(f"❌ Missing files: {len(missing_files)}")
         self.stdout.write(f"⚠️  Potentially corrupted: {len(corrupted_files)}")
@@ -73,37 +78,43 @@ class Command(BaseCommand):
         if missing_files:
             self.stdout.write(self.style.WARNING("\nMISSING FILES:"))
             for file_info in missing_files:
-                self.stdout.write(f"  - Video ID {file_info['video_id']}: {file_info['error']}")
+                self.stdout.write(
+                    f"  - Video ID {file_info['video_id']}: {file_info['error']}"
+                )
                 if fix_missing:
                     self.stdout.write("    → Marking as inactive (if applicable)")
 
         if corrupted_files:
             self.stdout.write(self.style.WARNING("\nPOTENTIALLY CORRUPTED FILES:"))
             for file_info in corrupted_files:
-                self.stdout.write(f"  - Video ID {file_info['video_id']}: {file_info['error']}")
+                self.stdout.write(
+                    f"  - Video ID {file_info['video_id']}: {file_info['error']}"
+                )
 
         if verbose and accessible_files:
             self.stdout.write(self.style.SUCCESS("\nACCESSIBLE FILES:"))
             for file_info in accessible_files[:10]:  # Show first 10
-                self.stdout.write(f"  ✅ Video ID {file_info['video_id']}: {file_info['path']} ({file_info['size_mb']:.1f} MB)")
-            
+                self.stdout.write(
+                    f"  ✅ Video ID {file_info['video_id']}: {file_info['path']} ({file_info['size_mb']:.1f} MB)"
+                )
+
             if len(accessible_files) > 10:
                 self.stdout.write(f"  ... and {len(accessible_files) - 10} more")
 
     def check_video_file(self, video, verbose=False):
         """
         Check a single video file for existence and basic accessibility.
-        
+
         Returns:
             dict: Status information about the video file
         """
         video_info = {
-            'video_id': video.id,
-            'video_uuid': str(video.uuid) if hasattr(video, 'uuid') else 'N/A',
-            'status': 'unknown',
-            'path': None,
-            'size_mb': 0,
-            'error': None
+            "video_id": video.id,
+            "video_uuid": str(video.video_hash) if hasattr(video, "uuid") else "N/A",
+            "status": "unknown",
+            "path": None,
+            "size_mb": 0,
+            "error": None,
         }
 
         # Helper to check a file attribute
@@ -116,42 +127,63 @@ class Command(BaseCommand):
             try:
                 file_path = Path(path_getter(file_field) if path_getter else file_field)
                 info = video_info.copy()
-                info['path'] = str(file_path)
+                info["path"] = str(file_path)
                 if not file_path.exists():
-                    info['status'] = 'missing'
-                    info['error'] = f"{label or attr.replace('_', ' ').title()} does not exist: {file_path}"
+                    info["status"] = "missing"
+                    info["error"] = (
+                        f"{label or attr.replace('_', ' ').title()} does not exist: {file_path}"
+                    )
                     return info
                 file_size = file_path.stat().st_size
-                info['size_mb'] = file_size / (1024 * 1024)
+                info["size_mb"] = file_size / (1024 * 1024)
                 if file_size == 0:
-                    info['status'] = 'corrupted'
-                    info['error'] = f"{label or attr.replace('_', ' ').title()} exists but has zero size"
+                    info["status"] = "corrupted"
+                    info["error"] = (
+                        f"{label or attr.replace('_', ' ').title()} exists but has zero size"
+                    )
                 else:
-                    info['status'] = 'accessible'
+                    info["status"] = "accessible"
                 return info
             except (ValueError, OSError) as e:
                 info = video_info.copy()
-                info['path'] = str(getattr(file_field, 'path', file_field))
-                info['status'] = 'corrupted'
-                info['error'] = f"Cannot access {label or attr.replace('_', ' ').title()}: {e}"
+                info["path"] = str(getattr(file_field, "path", file_field))
+                info["status"] = "corrupted"
+                info["error"] = (
+                    f"Cannot access {label or attr.replace('_', ' ').title()}: {e}"
+                )
                 return info
 
         # Try each file attribute in order of preference
         result = None
         # active_file_path: direct path string
-        result = _check_file_attr(video, 'active_file_path', label='Active file path')
-        if result: return result
+        result = _check_file_attr(video, "active_file_path", label="Active file path")
+        if result:
+            return result
         # active_file: Django FileField
-        result = _check_file_attr(video, 'active_file', path_getter=lambda f: f.path, label='Active file')
-        if result: return result
+        result = _check_file_attr(
+            video, "active_file", path_getter=lambda f: f.path, label="Active file"
+        )
+        if result:
+            return result
         # raw_file: Django FileField
-        result = _check_file_attr(video, 'raw_file', path_getter=lambda f: f.path, label='Raw file')
-        if result: return result
+        result = _check_file_attr(
+            video, "raw_file", path_getter=lambda f: f.path, label="Raw file"
+        )
+        if result:
+            return result
         # processed_file: Django FileField
-        result = _check_file_attr(video, 'processed_file', path_getter=lambda f: f.path, label='Processed file')
-        if result: return result
+        result = _check_file_attr(
+            video,
+            "processed_file",
+            path_getter=lambda f: f.path,
+            label="Processed file",
+        )
+        if result:
+            return result
 
         # If none found
-        video_info['status'] = 'missing'
-        video_info['error'] = "No video file paths found (no active_file, raw_file, or processed_file)"
-        return video_info
+        video_info["status"] = "missing"
+        video_info["error"] = (
+            "No video file paths found (no active_file, raw_file, or processed_file)"
+        )
+        return video_info

endoreg_db/management/commands/video_validation.py

@@ -1,4 +1,3 @@
-from endoreg_db.models import VideoPredictionMeta
 from typing import TYPE_CHECKING
 from django.core.management.base import BaseCommand
 
@@ -7,16 +6,15 @@ This command handles the video validation as seen in tests _pipe_1
 """
 
 if TYPE_CHECKING:
-    from endoreg_db.models import VideoFile, VideoState
-    
+    pass
+
+
 class Command(BaseCommand):
     help = "Data extraction and validation of video files in the database and updating their states accordingly."
 
     def handle(self, *args, **options):
         """
         Validates video files stored in the database and updates their states based on validation results.
-        
+
         This method is intended to be executed as a Django management command to ensure the integrity and correct status of video file records.
         """
-        
-    

endoreg_db/migrations/0001_initial.py

@@ -1,4 +1,4 @@
-# Generated by Django 5.2.7 on 2025-11-26 11:26
+# Generated by Django 6.0 on 2025-12-20 13:34
 
 import django.core.validators
 import django.db.models.deletion
@@ -185,11 +185,7 @@ class Migration(migrations.Migration):
             fields=[
                 ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
                 ('name', models.CharField(max_length=100, unique=True)),
-                ('name_de', models.CharField(max_length=100, null=True, unique=True)),
-                ('name_en', models.CharField(max_length=100, null=True, unique=True)),
                 ('description', models.TextField(blank=True, null=True)),
-                ('description_de', models.TextField(blank=True, null=True)),
-                ('description_en', models.TextField(blank=True, null=True)),
             ],
         ),
         migrations.CreateModel(
@@ -197,11 +193,7 @@ class Migration(migrations.Migration):
             fields=[
                 ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
                 ('name', models.CharField(max_length=255, unique=True)),
-                ('name_de', models.CharField(max_length=255, null=True, unique=True)),
-                ('name_en', models.CharField(max_length=255, null=True, unique=True)),
                 ('description', models.TextField(blank=True)),
-                ('description_de', models.TextField(blank=True, null=True)),
-                ('description_en', models.TextField(blank=True, null=True)),
             ],
         ),
         migrations.CreateModel(
@@ -209,11 +201,7 @@ class Migration(migrations.Migration):
             fields=[
                 ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
                 ('name', models.CharField(max_length=255, unique=True)),
-                ('name_de', models.CharField(max_length=255, null=True, unique=True)),
-                ('name_en', models.CharField(max_length=255, null=True, unique=True)),
                 ('description', models.TextField(blank=True)),
-                ('description_de', models.TextField(blank=True, null=True)),
-                ('description_en', models.TextField(blank=True, null=True)),
                 ('subcategories', models.JSONField(default=dict)),
                 ('numerical_descriptors', models.JSONField(default=dict)),
             ],
@@ -231,11 +219,7 @@ class Migration(migrations.Migration):
             fields=[
                 ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
                 ('name', models.CharField(max_length=100, unique=True)),
-                ('name_de', models.CharField(max_length=100, null=True, unique=True)),
-                ('name_en', models.CharField(max_length=100, null=True, unique=True)),
                 ('description', models.TextField(blank=True, null=True)),
-                ('description_de', models.TextField(blank=True, null=True)),
-                ('description_en', models.TextField(blank=True, null=True)),
             ],
         ),
         migrations.CreateModel(
@@ -261,6 +245,19 @@ class Migration(migrations.Migration):
                 ('name', models.CharField(max_length=255, unique=True)),
             ],
         ),
+        migrations.CreateModel(
+            name='Frame',
+            fields=[
+                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
+                ('frame_number', models.PositiveIntegerField()),
+                ('relative_path', models.CharField(max_length=512)),
+                ('timestamp', models.FloatField(blank=True, null=True)),
+                ('is_extracted', models.BooleanField(default=False)),
+            ],
+            options={
+                'ordering': ['video', 'frame_number'],
+            },
+        ),
         migrations.CreateModel(
             name='Gender',
             fields=[
@@ -387,6 +384,18 @@ class Migration(migrations.Migration):
                 ('name', models.CharField(max_length=255)),
             ],
         ),
+        migrations.CreateModel(
+            name='ProcessingHistory',
+            fields=[
+                ('file_hash', models.CharField(blank=True, help_text='Content hash of the original file (e.g. video_hash/pdf_hash).', max_length=64, primary_key=True, serialize=False)),
+                ('created_at', models.DateTimeField(auto_now_add=True)),
+                ('success', models.BooleanField(blank=True, default=False)),
+                ('object_id', models.PositiveBigIntegerField(blank=True, null=True)),
+            ],
+            options={
+                'ordering': ['-created_at'],
+            },
+        ),
         migrations.CreateModel(
             name='Product',
             fields=[
@@ -441,11 +450,11 @@ class Migration(migrations.Migration):
                 ('date_created', models.DateTimeField(auto_now_add=True)),
                 ('date_modified', models.DateTimeField(auto_now=True)),
                 ('was_created', models.BooleanField(default=True, help_text='True if this state was created for the first time.')),
-                ('pdf_meta_extracted', models.BooleanField(default=False, help_text='True if PDF metadata has been extracted.')),
+                ('pdf_meta_extracted', models.BooleanField(default=False, help_text='True if report metadata has been extracted.')),
             ],
             options={
-                'verbose_name': 'Raw PDF Processing State',
-                'verbose_name_plural': 'Raw PDF Processing States',
+                'verbose_name': 'Raw report Processing State',
+                'verbose_name_plural': 'Raw report Processing States',
             },
         ),
         migrations.CreateModel(
@@ -700,11 +709,7 @@ class Migration(migrations.Migration):
             fields=[
                 ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
                 ('name', models.CharField(max_length=100, unique=True)),
-                ('name_de', models.CharField(max_length=100, null=True, unique=True)),
-                ('name_en', models.CharField(max_length=100, null=True, unique=True)),
                 ('description', models.TextField(blank=True, null=True)),
-                ('description_de', models.TextField(blank=True, null=True)),
-                ('description_en', models.TextField(blank=True, null=True)),
                 ('indications', models.ManyToManyField(blank=True, related_name='examinations', to='endoreg_db.examinationindication')),
                 ('examination_times', models.ManyToManyField(blank=True, related_name='examinations', to='endoreg_db.examinationtime')),
                 ('examination_types', models.ManyToManyField(blank=True, to='endoreg_db.examinationtype')),
@@ -844,6 +849,32 @@ class Migration(migrations.Migration):
                 'abstract': False,
             },
         ),
+        migrations.CreateModel(
+            name='ImageClassificationAnnotation',
+            fields=[
+                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
+                ('value', models.BooleanField()),
+                ('float_value', models.FloatField(blank=True, null=True)),
+                ('annotator', models.CharField(blank=True, max_length=255, null=True)),
+                ('date_created', models.DateTimeField(auto_now_add=True)),
+                ('date_modified', models.DateTimeField(auto_now=True)),
+                ('frame', models.ForeignKey(on_delete=django.db.models.deletion.CASCADE, related_name='image_classification_annotations', to='endoreg_db.frame')),
+            ],
+        ),
+        migrations.CreateModel(
+            name='AIDataSet',
+            fields=[
+                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
+                ('name', models.CharField(blank=True, help_text='Human-readable identifier, e.g. "Legacy multilabel dataset v1".', max_length=255, null=True)),
+                ('description', models.TextField(blank=True, help_text='Optional notes / explanation about this dataset.', null=True)),
+                ('ai_model_type', models.CharField(default='image_multilabel_classification', help_text='AI model family this dataset is for, e.g. "image_multilabel_classification". Used to pick the correct architecture and output dimension logic.', max_length=255)),
+                ('dataset_type', models.CharField(choices=[('image', 'Image')], default='image', help_text='Controls which annotation table will be used. Currently only "image" is implemented; later "video", "text", etc.', max_length=32)),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='When this AIDataSet was created.')),
+                ('updated_at', models.DateTimeField(auto_now=True, help_text='When this AIDataSet was last modified.')),
+                ('is_active', models.BooleanField(default=True, help_text='Soft toggle to enable/disable this dataset for training.')),
+                ('image_annotations', models.ManyToManyField(blank=True, help_text="For dataset_type='image', this is the set of ImageClassificationAnnotation rows that define this AIDataSet. Each annotation has frame_id and label_id.", related_name='image_ai_datasets', to='endoreg_db.imageclassificationannotation')),
+            ],
+        ),
         migrations.CreateModel(
             name='InformationSource',
             fields=[
@@ -862,6 +893,11 @@ class Migration(migrations.Migration):
                 'indexes': [models.Index(fields=['name'], name='endoreg_db__name_febac5_idx'), models.Index(fields=['abbreviation'], name='endoreg_db__abbrevi_5a07c4_idx')],
             },
         ),
+        migrations.AddField(
+            model_name='imageclassificationannotation',
+            name='information_source',
+            field=models.ForeignKey(blank=True, default=None, null=True, on_delete=django.db.models.deletion.SET_NULL, related_name='image_classification_annotations', to='endoreg_db.informationsource'),
+        ),
         migrations.AddField(
             model_name='findingintervention',
             name='information_sources',
@@ -905,6 +941,11 @@ class Migration(migrations.Migration):
                 'verbose_name_plural': 'Information Source Types',
             },
         ),
+        migrations.AddField(
+            model_name='imageclassificationannotation',
+            name='label',
+            field=models.ForeignKey(on_delete=django.db.models.deletion.CASCADE, related_name='image_classification_annotations', to='endoreg_db.label'),
+        ),
         migrations.CreateModel(
             name='LabelSet',
             fields=[
@@ -1013,6 +1054,11 @@ class Migration(migrations.Migration):
                 'unique_together': {('name', 'version', 'model')},
             },
         ),
+        migrations.AddField(
+            model_name='imageclassificationannotation',
+            name='model_meta',
+            field=models.ForeignKey(blank=True, default=None, null=True, on_delete=django.db.models.deletion.SET_NULL, related_name='image_classification_annotations', to='endoreg_db.modelmeta'),
+        ),
         migrations.AddField(
             model_name='aimodel',
             name='active_meta',
@@ -1050,6 +1096,30 @@ class Migration(migrations.Migration):
                 ('default_unit', models.ForeignKey(blank=True, null=True, on_delete=django.db.models.deletion.CASCADE, to='endoreg_db.unit')),
             ],
         ),
+        migrations.CreateModel(
+            name='OperationLog',
+            fields=[
+                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
+                ('actor_username', models.CharField(blank=True, max_length=150)),
+                ('actor_email', models.EmailField(blank=True, max_length=254)),
+                ('actor_keycloak_id', models.CharField(blank=True, help_text='Keycloak subject/ID if you later want to store it.', max_length=255)),
+                ('action', models.CharField(help_text="e.g. 'anonymization.start', 'anonymization.validate'", max_length=100)),
+                ('http_method', models.CharField(blank=True, max_length=10)),
+                ('path', models.CharField(blank=True, max_length=512)),
+                ('resource_type', models.CharField(blank=True, help_text="e.g. 'video', 'pdf'", max_length=50)),
+                ('resource_id', models.IntegerField(blank=True, help_text='ID of VideoFile / RawPdfFile etc.', null=True)),
+                ('status_before', models.CharField(blank=True, max_length=50)),
+                ('status_after', models.CharField(blank=True, max_length=50)),
+                ('meta', models.JSONField(blank=True, null=True)),
+                ('created_at', models.DateTimeField(auto_now_add=True)),
+                ('actor_user', models.ForeignKey(blank=True, null=True, on_delete=django.db.models.deletion.SET_NULL, related_name='operation_logs', to=settings.AUTH_USER_MODEL)),
+            ],
+            options={
+                'verbose_name': 'Operation Log',
+                'verbose_name_plural': 'Operation Logs',
+                'ordering': ['-created_at'],
+            },
+        ),
         migrations.CreateModel(
             name='Patient',
             fields=[
@@ -1149,7 +1219,7 @@ class Migration(migrations.Migration):
                 ('text', models.TextField(blank=True, null=True)),
                 ('date', models.DateField(blank=True, null=True)),
                 ('time', models.TimeField(blank=True, null=True)),
-                ('file', models.FileField(blank=True, null=True, upload_to='pdfs')),
+                ('file', models.FileField(blank=True, null=True, upload_to='documents')),
                 ('center', models.ForeignKey(blank=True, null=True, on_delete=django.db.models.deletion.SET_NULL, to='endoreg_db.center')),
                 ('type', models.ForeignKey(blank=True, null=True, on_delete=django.db.models.deletion.SET_NULL, to='endoreg_db.documenttype')),
                 ('examiners', models.ManyToManyField(blank=True, to='endoreg_db.examiner')),
@@ -1457,9 +1527,8 @@ class Migration(migrations.Migration):
                 ('text', models.TextField(blank=True, null=True)),
                 ('date_created', models.DateTimeField(auto_now_add=True)),
                 ('date_modified', models.DateTimeField(auto_now=True)),
-                ('anonymized', models.BooleanField(default=False, help_text='True if the PDF has been anonymized.')),
-                ('file', models.FileField(upload_to='pdfs', validators=[django.core.validators.FileExtensionValidator(allowed_extensions=['pdf'])])),
-                ('anonymized_file', models.FileField(blank=True, null=True, upload_to='pdfs', validators=[django.core.validators.FileExtensionValidator(allowed_extensions=['pdf'])])),
+                ('file', models.FileField(upload_to='sensitive_reports', validators=[django.core.validators.FileExtensionValidator(allowed_extensions=['pdf'])])),
+                ('processed_file', models.FileField(blank=True, null=True, upload_to='processed_reports_final', validators=[django.core.validators.FileExtensionValidator(allowed_extensions=['pdf'])])),
                 ('state_report_processing_required', models.BooleanField(default=True)),
                 ('state_report_processed', models.BooleanField(default=False)),
                 ('raw_meta', models.JSONField(blank=True, null=True)),
@@ -1482,7 +1551,7 @@ class Migration(migrations.Migration):
                 ('text', models.TextField(blank=True, null=True)),
                 ('date', models.DateField(blank=True, null=True)),
                 ('time', models.TimeField(blank=True, null=True)),
-                ('file', models.FileField(blank=True, null=True, upload_to='pdfs')),
+                ('file', models.FileField(blank=True, null=True, upload_to='documents')),
                 ('center', models.ForeignKey(blank=True, null=True, on_delete=django.db.models.deletion.SET_NULL, to='endoreg_db.center')),
                 ('type', models.ForeignKey(blank=True, null=True, on_delete=django.db.models.deletion.SET_NULL, to='endoreg_db.documenttype')),
                 ('examiners', models.ManyToManyField(blank=True, to='endoreg_db.examiner')),
@@ -1634,7 +1703,7 @@ class Migration(migrations.Migration):
             name='UploadJob',
             fields=[
                 ('id', models.UUIDField(default=uuid.uuid4, editable=False, help_text='Unique identifier for the upload job', primary_key=True, serialize=False)),
-                ('file', models.FileField(help_text='Uploaded file (PDF or video)', upload_to='uploads/%Y/%m/%d/')),
+                ('file', models.FileField(help_text='Uploaded file (report or video)', upload_to='uploads/%Y/%m/%d/')),
                 ('status', models.CharField(choices=[('pending', 'Pending'), ('processing', 'Processing'), ('anonymized', 'Anonymized'), ('error', 'Error')], default='pending', help_text='Current processing status of the upload', max_length=20)),
                 ('content_type', models.CharField(blank=True, help_text='MIME type of the uploaded file', max_length=100)),
                 ('error_detail', models.TextField(blank=True, help_text='Error message if processing failed')),
@@ -1652,9 +1721,8 @@ class Migration(migrations.Migration):
             name='VideoFile',
             fields=[
                 ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
-                ('uuid', models.UUIDField(default=uuid.uuid4, editable=False, unique=True)),
-                ('raw_file', models.FileField(blank=True, null=True, upload_to='videos', validators=[django.core.validators.FileExtensionValidator(allowed_extensions=['mp4'])])),
-                ('processed_file', models.FileField(blank=True, null=True, upload_to='videos', validators=[django.core.validators.FileExtensionValidator(allowed_extensions=['mp4'])])),
+                ('raw_file', models.FileField(blank=True, null=True, upload_to='sensitive_videos', validators=[django.core.validators.FileExtensionValidator(allowed_extensions=['mp4'])])),
+                ('processed_file', models.FileField(blank=True, null=True, upload_to='processed_videos_final', validators=[django.core.validators.FileExtensionValidator(allowed_extensions=['mp4'])])),
                 ('video_hash', models.CharField(help_text='Hash of the raw video file.', max_length=255, unique=True)),
                 ('processed_video_hash', models.CharField(blank=True, help_text='Hash of the processed video file, unique if not null.', max_length=255, null=True, unique=True)),
                 ('original_file_name', models.CharField(blank=True, max_length=255, null=True)),
@@ -1690,19 +1758,10 @@ class Migration(migrations.Migration):
             name='video_file',
             field=models.ForeignKey(on_delete=django.db.models.deletion.CASCADE, related_name='label_video_segments', to='endoreg_db.videofile'),
         ),
-        migrations.CreateModel(
-            name='Frame',
-            fields=[
-                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
-                ('frame_number', models.PositiveIntegerField()),
-                ('relative_path', models.CharField(max_length=512)),
-                ('timestamp', models.FloatField(blank=True, null=True)),
-                ('is_extracted', models.BooleanField(default=False)),
-                ('video', models.ForeignKey(on_delete=django.db.models.deletion.CASCADE, related_name='frames', to='endoreg_db.videofile')),
-            ],
-            options={
-                'ordering': ['video', 'frame_number'],
-            },
+        migrations.AddField(
+            model_name='frame',
+            name='video',
+            field=models.ForeignKey(on_delete=django.db.models.deletion.CASCADE, related_name='frames', to='endoreg_db.videofile'),
         ),
         migrations.CreateModel(
             name='VideoMeta',
@@ -1815,23 +1874,13 @@ class Migration(migrations.Migration):
                 ('waste', models.ForeignKey(on_delete=django.db.models.deletion.CASCADE, to='endoreg_db.waste')),
             ],
         ),
-        migrations.CreateModel(
-            name='ImageClassificationAnnotation',
-            fields=[
-                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
-                ('value', models.BooleanField()),
-                ('float_value', models.FloatField(blank=True, null=True)),
-                ('annotator', models.CharField(blank=True, max_length=255, null=True)),
-                ('date_created', models.DateTimeField(auto_now_add=True)),
-                ('date_modified', models.DateTimeField(auto_now=True)),
-                ('frame', models.ForeignKey(on_delete=django.db.models.deletion.CASCADE, related_name='image_classification_annotations', to='endoreg_db.frame')),
-                ('information_source', models.ForeignKey(blank=True, default=None, null=True, on_delete=django.db.models.deletion.SET_NULL, related_name='image_classification_annotations', to='endoreg_db.informationsource')),
-                ('label', models.ForeignKey(on_delete=django.db.models.deletion.CASCADE, related_name='image_classification_annotations', to='endoreg_db.label')),
-                ('model_meta', models.ForeignKey(blank=True, default=None, null=True, on_delete=django.db.models.deletion.SET_NULL, related_name='image_classification_annotations', to='endoreg_db.modelmeta')),
-            ],
-            options={
-                'indexes': [models.Index(fields=['frame', 'label'], name='endoreg_db__frame_i_47f916_idx'), models.Index(fields=['frame'], name='endoreg_db__frame_i_92c75a_idx')],
-            },
+        migrations.AddIndex(
+            model_name='imageclassificationannotation',
+            index=models.Index(fields=['frame', 'label'], name='endoreg_db__frame_i_47f916_idx'),
+        ),
+        migrations.AddIndex(
+            model_name='imageclassificationannotation',
+            index=models.Index(fields=['frame'], name='endoreg_db__frame_i_92c75a_idx'),
         ),
         migrations.AddConstraint(
             model_name='patientexternalid',

endoreg_db/models/__init__.py

@@ -151,6 +151,12 @@ from .state import (
     VideoState,
 )
 
+# AI dataset model
+from .aidataset import (
+    AIDataSet,
+)
+
+
 __all__ = [
     ####### Administration ########
     # AI
@@ -311,4 +317,6 @@ __all__ = [
     "LabelVideoSegmentState",
     "AnonymizationState",
     "RawPdfState",
+    ###### AIDATASET ######
+    "AIDataSet",
 ]

endoreg_db/models/administration/ai/active_model.py

@@ -8,7 +8,8 @@ if TYPE_CHECKING:
 class ActiveModelManager(models.Manager):
     def get_by_natural_key(self, name):
         return self.get(name=name)
-    
+
+
 class ActiveModel(models.Model):
     """
     ActiveModel represents an active instance of a model within the application.
@@ -26,10 +27,9 @@ class ActiveModel(models.Model):
     """
 
     name = models.CharField(max_length=255, unique=True)
-    
-    model_meta:models.ForeignKey["ModelMeta|None"] = models.ForeignKey(
-        'ModelMeta', on_delete=models.SET_NULL, 
-        blank=True, null=True
+
+    model_meta: models.ForeignKey["ModelMeta|None"] = models.ForeignKey(
+        "ModelMeta", on_delete=models.SET_NULL, blank=True, null=True
     )
 
     objects = ActiveModelManager()