feat: handle file content logic when creating next component version

openedx_learning/apps/authoring/components/api.py

@@ -12,6 +12,7 @@
 """
 from __future__ import annotations
 
+import mimetypes
 from datetime import datetime, timezone
 from enum import StrEnum, auto
 from logging import getLogger
@@ -129,7 +130,7 @@ def create_component_version(
 def create_next_component_version(
     component_pk: int,
     /,
-    content_to_replace: dict[str, int | None],
+    content_to_replace: dict[str, int | None | bytes],
     created: datetime,
     title: str | None = None,
     created_by: int | None = None,
@@ -191,6 +192,20 @@ def create_next_component_version(
             # content represented by our key from the next version. Otherwise,
             # we add our key->content_pk mapping to the next version.
             if content_pk is not None:
+                if isinstance(content_pk, bytes):
+                    file_path, file_content = key, content_pk
+                    media_type_str, _encoding = mimetypes.guess_type(file_path)
+                    # We use "application/octet-stream" as a generic fallback media type, per
+                    # RFC 2046: https://datatracker.ietf.org/doc/html/rfc2046
+                    media_type_str = media_type_str or "application/octet-stream"
+                    media_type = contents_api.get_or_create_media_type(media_type_str)
+                    content = contents_api.get_or_create_file_content(
+                        component.publishable_entity.learning_package.id,
+                        media_type.id,
+                        data=file_content,
+                        created=created,
+                    )
+                    content_pk = content.pk
                 ComponentVersionContent.objects.create(
                     content_id=content_pk,
                     component_version=component_version,

openedx_learning/apps/authoring/components/management/commands/add_assets_to_component.py

@@ -4,14 +4,12 @@
 This is mostly meant to be a debugging tool to let us to easily load some test
 asset data into the system.
 """
-import mimetypes
 import pathlib
 from datetime import datetime, timezone
 
 from django.core.management.base import BaseCommand
 
 from ....components.api import create_component_version_content
-from ....contents.api import get_or_create_file_content, get_or_create_media_type
 from ....publishing.api import get_learning_package_by_key
 from ...api import create_next_component_version, get_component_by_key
 
@@ -70,7 +68,7 @@ def handle(self, *args, **options):
 
         created = datetime.now(tz=timezone.utc)
         keys_to_remove = set()
-        local_keys_to_content = {}
+        local_keys_to_raw_content = {}
 
         for file_mapping in file_mappings:
             local_key, file_path = file_mapping.split(":", 1)
@@ -80,22 +78,14 @@ def handle(self, *args, **options):
                 keys_to_remove.add(local_key)
                 continue
 
-            media_type_str, _encoding = mimetypes.guess_type(file_path)
-            media_type = get_or_create_media_type(media_type_str)
-            content = get_or_create_file_content(
-                learning_package.id,
-                media_type.id,
-                data=pathlib.Path(file_path).read_bytes(),
-                created=created,
-            )
-            local_keys_to_content[local_key] = content.id
+            local_keys_to_raw_content[local_key] = pathlib.Path(file_path).read_bytes()
 
         next_version = create_next_component_version(
             component.pk,
             content_to_replace={local_key: None for local_key in keys_to_remove},
             created=created,
         )
-        for local_key, content_id in sorted(local_keys_to_content.items()):
+        for local_key, content_id in sorted(local_keys_to_raw_content.items()):
             create_component_version_content(
                 next_version.pk,
                 content_id,

tests/openedx_learning/apps/authoring/components/test_api.py

@@ -442,13 +442,14 @@ def test_multiple_versions(self):
             content_to_replace={
                 "hello.txt": hello_content.pk,
                 "goodbye.txt": goodbye_content.pk,
+                "raw.txt": b'raw content',
             },
             created=self.now,
         )
         assert version_1.version_num == 1
         assert version_1.title == "Problem Version 1"
         version_1_contents = list(version_1.contents.all())
-        assert len(version_1_contents) == 2
+        assert len(version_1_contents) == 3
         assert (
             hello_content ==
             version_1.contents
@@ -472,7 +473,7 @@ def test_multiple_versions(self):
             created=self.now,
         )
         assert version_2.version_num == 2
-        assert version_2.contents.count() == 3
+        assert version_2.contents.count() == 4
         assert (
             blank_content ==
             version_2.contents
@@ -503,7 +504,7 @@ def test_multiple_versions(self):
             created=self.now,
         )
         assert version_3.version_num == 3
-        assert version_3.contents.count() == 1
+        assert version_3.contents.count() == 2
         assert (
             hello_content ==
             version_3.contents