Extract media duration on upload

contentcuration/contentcuration/frontend/shared/data/resources.js

@@ -1510,7 +1510,7 @@ export const File = new Resource({
   tableName: TABLE_NAMES.FILE,
   urlName: 'file',
   indexFields: ['contentnode'],
-  uploadUrl({ checksum, size, type, name, file_format, preset }) {
+  uploadUrl({ checksum, size, type, name, file_format, preset, duration = null }) {
     return client
       .post(this.getUrlFunction('upload_url')(), {
         checksum,
@@ -1519,6 +1519,7 @@ export const File = new Resource({
         name,
         file_format,
         preset,
+        duration,
       })
       .then(response => {
         if (!response) {

contentcuration/contentcuration/frontend/shared/vuex/file/actions.js

@@ -1,4 +1,4 @@
-import { getHash, inferPreset, storageUrl } from './utils';
+import { getHash, extractMetadata, storageUrl } from './utils';
 import { File } from 'shared/data/resources';
 import client from 'shared/client';
 import { fileErrors, NOVALUE } from 'shared/constants';
@@ -166,8 +166,8 @@ export function uploadFileToStorage(
 export function uploadFile(context, { file, preset = null } = {}) {
   return new Promise((resolve, reject) => {
     // 1. Get the checksum of the file
-    Promise.all([getHash(file), preset ? Promise.resolve() : inferPreset(file)])
-      .then(([checksum, presetId]) => {
+    Promise.all([getHash(file), extractMetadata(file, preset)])
+      .then(([checksum, metadata]) => {
         const file_format = file.name
           .split('.')
           .pop()
@@ -179,7 +179,7 @@ export function uploadFile(context, { file, preset = null } = {}) {
           type: file.type,
           name: file.name,
           file_format,
-          preset: preset || presetId,
+          ...metadata,
         })
           .then(data => {
             const fileObject = {
@@ -235,7 +235,7 @@ export function uploadFile(context, { file, preset = null } = {}) {
               file_size: file.size,
               original_filename: file.name,
               file_format,
-              preset: presetId,
+              preset: metadata.preset,
               error: errorType,
             };
             context.commit('ADD_FILE', fileObject);

contentcuration/contentcuration/frontend/shared/vuex/file/utils.js

@@ -3,6 +3,12 @@ import { FormatPresetsList, FormatPresetsNames } from 'shared/leUtils/FormatPres
 
 const BLOB_SLICE = File.prototype.slice || File.prototype.mozSlice || File.prototype.webkitSlice;
 const CHUNK_SIZE = 2097152;
+const MEDIA_PRESETS = [
+  FormatPresetsNames.AUDIO,
+  FormatPresetsNames.HIGH_RES_VIDEO,
+  FormatPresetsNames.LOW_RES_VIDEO,
+];
+const VIDEO_PRESETS = [FormatPresetsNames.HIGH_RES_VIDEO, FormatPresetsNames.LOW_RES_VIDEO];
 
 export function getHash(file) {
   return new Promise((resolve, reject) => {
@@ -55,40 +61,48 @@ export function storageUrl(checksum, file_format) {
   return `${window.storageBaseUrl}${checksum[0]}/${checksum[1]}/${checksum}.${file_format}`;
 }
 
-export function inferPreset(file) {
-  return new Promise(resolve => {
-    if (file.preset) {
-      resolve(file.preset);
-    }
-    const file_format = file.name
+/**
+ * @param {{name: String, preset: String}} file
+ * @param {String|null} preset
+ * @return {Promise<{preset: String, duration:Number|null}>}
+ */
+export function extractMetadata(file, preset = null) {
+  const metadata = {
+    preset: file.preset || preset,
+  };
+
+  if (!metadata.preset) {
+    const fileFormat = file.name
       .split('.')
       .pop()
       .toLowerCase();
-    const inferredPresets = extensionPresetMap[file_format];
-    if (inferredPresets && inferredPresets.length > 1) {
-      // Special processing for inferring preset of videos
-      if (
-        inferredPresets.length === 2 &&
-        inferredPresets.includes(FormatPresetsNames.HIGH_RES_VIDEO) &&
-        inferredPresets.includes(FormatPresetsNames.LOW_RES_VIDEO)
-      ) {
-        const videoElement = document.createElement('video');
-        const videoSource = URL.createObjectURL(file);
-        // Add a listener to read the height from the video once
-        // the metadata has loaded.
-        videoElement.addEventListener('loadedmetadata', () => {
-          if (videoElement.videoHeight >= 720) {
-            resolve(FormatPresetsNames.HIGH_RES_VIDEO);
-          } else {
-            resolve(FormatPresetsNames.LOW_RES_VIDEO);
-          }
-        });
-        // Set the src url on the video element
-        videoElement.src = videoSource;
-        // Return here to prevent subsequent processing
-        return;
+    // Default to whatever the first preset is
+    metadata.preset = extensionPresetMap[fileFormat][0];
+  }
+
+  // End here if not audio or video
+  if (!MEDIA_PRESETS.includes(metadata.preset)) {
+    return Promise.resolve(metadata);
+  }
+
+  // Extract additional media metadata
+  const isVideo = VIDEO_PRESETS.includes(metadata.preset);
+
+  return new Promise(resolve => {
+    const mediaElement = document.createElement(isVideo ? 'video' : 'audio');
+    // Add a listener to read the metadata once it has loaded.
+    mediaElement.addEventListener('loadedmetadata', () => {
+      metadata.duration = mediaElement.duration;
+      // Override preset based off video resolution
+      if (isVideo) {
+        metadata.preset =
+          mediaElement.videoHeight >= 720
+            ? FormatPresetsNames.HIGH_RES_VIDEO
+            : FormatPresetsNames.LOW_RES_VIDEO;
       }
-    }
-    resolve(inferredPresets && inferredPresets[0]);
+      resolve(metadata);
+    });
+    // Set the src url on the media element
+    mediaElement.src = URL.createObjectURL(file);
   });
 }

contentcuration/contentcuration/migrations/0133_auto_20220124_2149.py

@@ -0,0 +1,22 @@
+# Generated by Django 3.2.5 on 2022-01-24 21:49
+from django.db import migrations
+from django.db import models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('contentcuration', '0132_auto_20210708_0011'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='file',
+            name='duration',
+            field=models.IntegerField(blank=True, null=True),
+        ),
+        migrations.AddConstraint(
+            model_name='file',
+            constraint=models.CheckConstraint(check=models.Q(models.Q(('duration__gt', 0), ('preset__in', ['audio', 'high_res_video', 'low_res_video'])), ('duration__isnull', True), _connector='OR'), name='file_media_duration_int'),
+        ),
+    ]

contentcuration/contentcuration/models.py

@@ -1922,6 +1922,8 @@ class StagedFile(models.Model):
 
 FILE_DISTINCT_INDEX_NAME = "file_checksum_file_size_idx"
 FILE_MODIFIED_DESC_INDEX_NAME = "file_modified_desc_idx"
+FILE_DURATION_CONSTRAINT = "file_media_duration_int"
+MEDIA_PRESETS = [format_presets.AUDIO, format_presets.VIDEO_HIGH_RES, format_presets.VIDEO_LOW_RES]
 
 
 class File(models.Model):
@@ -1945,6 +1947,7 @@ class File(models.Model):
     uploaded_by = models.ForeignKey(User, related_name='files', blank=True, null=True, on_delete=models.SET_NULL)
 
     modified = models.DateTimeField(auto_now=True, verbose_name="modified", null=True)
+    duration = models.IntegerField(blank=True, null=True)
 
     objects = CustomManager()
 
@@ -2056,6 +2059,9 @@ class Meta:
             models.Index(fields=['checksum', 'file_size'], name=FILE_DISTINCT_INDEX_NAME),
             models.Index(fields=["-modified"], name=FILE_MODIFIED_DESC_INDEX_NAME),
         ]
+        constraints = [
+            models.CheckConstraint(check=(Q(preset__in=MEDIA_PRESETS, duration__gt=0) | Q(duration__isnull=True)), name=FILE_DURATION_CONSTRAINT)
+        ]
 
 
 @receiver(models.signals.post_delete, sender=File)
@@ -2095,7 +2101,7 @@ def clean(self, *args, **kwargs):
             raise IntegrityError('Cannot self reference as prerequisite.')
         # immediate cyclic exception
         if PrerequisiteContentRelationship.objects.using(self._state.db) \
-                        .filter(target_node=self.prerequisite, prerequisite=self.target_node):
+                .filter(target_node=self.prerequisite, prerequisite=self.target_node):
             raise IntegrityError(
                 'Note: Prerequisite relationship is directional! %s and %s cannot be prerequisite of each other!'
                 % (self.target_node, self.prerequisite))
@@ -2128,7 +2134,7 @@ def save(self, *args, **kwargs):
             raise IntegrityError('Cannot self reference as related.')
         # handle immediate cyclic
         if RelatedContentRelationship.objects.using(self._state.db) \
-                        .filter(contentnode_1=self.contentnode_2, contentnode_2=self.contentnode_1):
+                .filter(contentnode_1=self.contentnode_2, contentnode_2=self.contentnode_1):
             return  # silently cancel the save
         super(RelatedContentRelationship, self).save(*args, **kwargs)
 

contentcuration/contentcuration/tests/test_models.py

@@ -5,11 +5,13 @@
 from django.conf import settings
 from django.db.utils import IntegrityError
 from le_utils.constants import content_kinds
+from le_utils.constants import format_presets
 
 from contentcuration.models import AssessmentItem
 from contentcuration.models import Channel
 from contentcuration.models import ContentNode
 from contentcuration.models import File
+from contentcuration.models import FILE_DURATION_CONSTRAINT
 from contentcuration.models import generate_object_storage_name
 from contentcuration.models import Invitation
 from contentcuration.models import object_storage_name
@@ -569,6 +571,37 @@ def test_filter_edit_queryset__uploaded_by(self):
         queryset = File.filter_edit_queryset(self.base_queryset, user=user)
         self.assertQuerysetContains(queryset, pk=node_file.id)
 
+    def test_duration_check_constraint__acceptable(self):
+        channel = testdata.channel()
+        File.objects.create(
+            contentnode=create_contentnode(channel.main_tree_id),
+            preset_id=format_presets.AUDIO,
+            duration=10,
+        )
+        File.objects.create(
+            contentnode=create_contentnode(channel.main_tree_id),
+            preset_id=format_presets.VIDEO_HIGH_RES,
+            duration=1123123,
+        )
+
+    def test_duration_check_constraint__negative(self):
+        channel = testdata.channel()
+        with self.assertRaises(IntegrityError, msg=FILE_DURATION_CONSTRAINT):
+            File.objects.create(
+                contentnode=create_contentnode(channel.main_tree_id),
+                preset_id=format_presets.AUDIO,
+                duration=-10,
+            )
+
+    def test_duration_check_constraint__not_media(self):
+        channel = testdata.channel()
+        with self.assertRaises(IntegrityError, msg=FILE_DURATION_CONSTRAINT):
+            File.objects.create(
+                contentnode=create_contentnode(channel.main_tree_id),
+                preset_id=format_presets.EPUB,
+                duration=10,
+            )
+
 
 class AssessmentItemFilePermissionTestCase(PermissionQuerysetTestCase):
     @property

contentcuration/contentcuration/utils/publish.py

@@ -20,6 +20,7 @@
 from django.core.management import call_command
 from django.db import transaction
 from django.db.models import Count
+from django.db.models import Max
 from django.db.models import Q
 from django.db.models import Sum
 from django.db.utils import IntegrityError
@@ -225,6 +226,11 @@ def create_bare_contentnode(ccnode, default_language, channel_id, channel_name):
     if ccnode.language or default_language:
         language, _new = get_or_create_language(ccnode.language or default_language)
 
+    duration = None
+    if ccnode.kind_id in [content_kinds.AUDIO, content_kinds.VIDEO]:
+        # aggregate duration from associated files, choosing maximum if there are multiple, like hi and lo res videos
+        duration = ccnode.files.aggregate(duration=Max("duration")).get("duration")
+
     options = {}
     if ccnode.extra_fields and 'options' in ccnode.extra_fields:
         options = ccnode.extra_fields['options']
@@ -247,6 +253,7 @@ def create_bare_contentnode(ccnode, default_language, channel_id, channel_name):
             'license_name': kolibri_license.license_name if kolibri_license is not None else None,
             'license_description': kolibri_license.license_description if kolibri_license is not None else None,
             'coach_content': ccnode.role_visibility == roles.COACH,
+            'duration': duration,
             'options': json.dumps(options)
         }
     )

contentcuration/contentcuration/viewsets/file.py

@@ -71,6 +71,7 @@ class Meta:
             "contentnode",
             "assessment_item",
             "preset",
+            "duration"
         )
         list_serializer_class = BulkListSerializer
 
@@ -98,6 +99,7 @@ class FileViewSet(BulkDeleteMixin, BulkUpdateMixin, ReadOnlyValuesViewset):
         "language_id",
         "original_filename",
         "uploaded_by",
+        "duration"
     )
 
     field_map = {
@@ -164,6 +166,7 @@ def upload_url(self, request):
             file_format_id=file_format,
             preset_id=preset,
             uploaded_by=request.user,
+            duration=request.data.get("duration"),
         )
 
         # Avoid using our file_on_disk attribute for checks

contentcuration/kolibri_content/migrations/0016_contentnode_duration.py

@@ -0,0 +1,18 @@
+# Generated by Django 3.2.5 on 2022-01-27 18:56
+from django.db import migrations
+from django.db import models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('content', '0015_auto_20210707_1606'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='contentnode',
+            name='duration',
+            field=models.IntegerField(blank=True, null=True),
+        ),
+    ]

contentcuration/kolibri_content/models.py

@@ -122,6 +122,9 @@ class ContentNode(MPTTModel):
     # A JSON Dictionary of properties to configure loading, rendering, etc. the file
     options = models.TextField(default="{}")
 
+    # If media, the duration in seconds
+    duration = models.IntegerField(null=True, blank=True)
+
     class Meta:
         ordering = ("lft",)
         index_together = [