See #432: Moved tag validation in the tags app

api/funkwhale_api/music/metadata.py

@@ -2,7 +2,6 @@ import base64
 import datetime
 import logging
 import pendulum
-import re
 
 import mutagen._util
 import mutagen.oggtheora
@@ -12,6 +11,8 @@ import mutagen.flac
 from rest_framework import serializers
 from rest_framework.compat import Mapping
 
+from funkwhale_api.tags import models as tags_models
+
 logger = logging.getLogger(__name__)
 NODEFAULT = object()
 # default title used when imported tracks miss the `Album` tag, see #122
@@ -491,9 +492,6 @@ class PermissiveDateField(serializers.CharField):
         return None
 
 
-TAG_REGEX = re.compile(r"^((\w+)([\d_]*))$")
-
-
 def extract_tags_from_genre(string):
     tags = []
     delimiter = "@@@@@"
@@ -511,7 +509,7 @@ def extract_tags_from_genre(string):
         if not tag:
             continue
         final_tag = ""
-        if not TAG_REGEX.match(tag.replace(" ", "")):
+        if not tags_models.TAG_REGEX.match(tag.replace(" ", "")):
             # the string contains some non words chars ($, €, etc.), right now
             # we simply skip such tags
             continue

api/funkwhale_api/tags/models.py

@@ -1,3 +1,5 @@
+import re
+
 from django.contrib.contenttypes.fields import GenericForeignKey
 from django.contrib.contenttypes.models import ContentType
 from django.contrib.postgres.fields import CICharField
@@ -8,6 +10,9 @@ from django.utils import timezone
 from django.utils.translation import gettext_lazy as _
 
 
+TAG_REGEX = re.compile(r"^((\w+)([\d_]*))$")
+
+
 class Tag(models.Model):
     name = CICharField(max_length=100, unique=True)
     creation_date = models.DateTimeField(default=timezone.now)

api/funkwhale_api/tags/serializers.py

@@ -7,3 +7,11 @@ class TagSerializer(serializers.ModelSerializer):
     class Meta:
         model = models.Tag
         fields = ["name", "creation_date"]
+
+
+class TagNameField(serializers.CharField):
+    def to_internal_value(self, value):
+        value = super().to_internal_value(value)
+        if not models.TAG_REGEX.match(value):
+            raise serializers.ValidationError('Invalid tag "{}"'.format(value))
+        return value

api/tests/tags/test_serializers.py

@@ -1,3 +1,5 @@
+import pytest
+
 from funkwhale_api.tags import serializers
 
 
@@ -12,3 +14,18 @@ def test_tag_serializer(factories):
     }
 
     assert serializer.data == expected
+
+
+@pytest.mark.parametrize(
+    "name",
+    [
+        "",
+        "invalid because spaces",
+        "invalid-because-dashes",
+        "invalid because non breaking spaces",
+    ],
+)
+def test_tag_name_field_validation(name):
+    field = serializers.TagNameField()
+    with pytest.raises(serializers.serializers.ValidationError):
+        field.to_internal_value(name)