Normalize tag names

lbry/lbry/schema/attrs.py

@@ -13,6 +13,7 @@ from torba.client.constants import COIN
 
 from lbry.schema.mime_types import guess_media_type
 from lbry.schema.base import Metadata, BaseMessageList
+from lbry.schema.tags import clean_tags, normalize_tag
 from lbry.schema.types.v2.claim_pb2 import (
     Fee as FeeMessage,
     Location as LocationMessage,
@@ -533,3 +534,22 @@ class LocationList(BaseMessageList[Location]):
 
     def append(self, value):
         self.add().from_value(value)
+
+
+class TagList(BaseMessageList[str]):
+    __slots__ = ()
+    item_class = str
+
+    def __init__(self, message):
+        message[:] = clean_tags(message)
+        self.message = message
+
+    def append(self, tag: str):
+        tag = normalize_tag(tag)
+        if tag not in self.message:
+            self.message.append(tag)
+
+    def extend(self, tags: List[str]):
+        tags = clean_tags(tags)
+        tags = list(set(tags).difference(set(self.message)))
+        self.message.extend(tags)

lbry/lbry/schema/base.py

@@ -119,3 +119,6 @@ class BaseMessageList(Metadata, Generic[I]):
 
     def __delitem__(self, key):
         del self._message[key]
+
+    def __eq__(self, values: List[str]) -> bool:
+        return self._message == values

lbry/lbry/schema/claim.py

@@ -13,7 +13,7 @@ from lbry.schema.base import Signable
 from lbry.schema.mime_types import guess_media_type, guess_stream_type
 from lbry.schema.attrs import (
     Source, Playable, Dimmensional, Fee, Image, Video, Audio,
-    LanguageList, LocationList, ClaimList, ClaimReference
+    LanguageList, LocationList, ClaimList, ClaimReference, TagList
 )
 from lbry.schema.types.v2.claim_pb2 import Claim as ClaimMessage
 
@@ -164,8 +164,8 @@ class BaseClaim:
         return Source(self.claim.message.thumbnail)
 
     @property
-    def tags(self) -> List:
-        return self.claim.message.tags
+    def tags(self) -> List[str]:
+        return TagList(self.claim.message.tags)
 
     @property
     def languages(self) -> LanguageList:

lbry/tests/integration/test_claim_commands.py

@@ -523,6 +523,19 @@ class ChannelCommands(CommandTestCase):
         self.assertEqual(result[0]['amount'], '3.0')
         await self.channel_abandon(self.get_claim_id(channel))
 
+    async def test_tag_normalization(self):
+        tx1 = await self.channel_create('@abc', '1.0', tags=['aBc', ' ABC ', 'xYZ ', 'xyz'])
+        claim_id = self.get_claim_id(tx1)
+        self.assertCountEqual(tx1['outputs'][0]['value']['tags'], ['abc', 'xyz'])
+
+        tx2 = await self.channel_update(claim_id, tags=[' pqr', 'PQr '])
+        self.assertCountEqual(tx2['outputs'][0]['value']['tags'], ['abc', 'xyz', 'pqr'])
+
+        tx3 = await self.channel_update(claim_id, tags=' pqr')
+        self.assertCountEqual(tx3['outputs'][0]['value']['tags'], ['abc', 'xyz', 'pqr'])
+
+        tx4 = await self.channel_update(claim_id, tags=[' pqr', 'PQr '], clear_tags=True)
+        self.assertEqual(tx4['outputs'][0]['value']['tags'], ['pqr'])
 
 class StreamCommands(ClaimTestCase):
 
@@ -1101,13 +1114,14 @@ class StreamCommands(ClaimTestCase):
         )
 
         # publishing again clears channel
-        tx4 = await self.publish('foo', languages='uk-UA')
+        tx4 = await self.publish('foo', languages='uk-UA', tags=['Anime', 'anime '])
         self.assertEqual(2, len(self.daemon.jsonrpc_file_list()))
         r = await self.resolve('lbry://foo')
         claim = r['lbry://foo']
         self.assertEqual(claim['txid'], tx4['outputs'][0]['txid'])
         self.assertNotIn('signing_channel', claim)
         self.assertEqual(claim['value']['languages'], ['uk-UA'])
+        self.assertEqual(claim['value']['tags'], ['anime'])
 
 
 class SupportCommands(CommandTestCase):

lbry/tests/unit/schema/test_models.py

@@ -97,6 +97,27 @@ class TestLanguages(TestCase):
             stream.languages.append('en-Zzz-US')
 
 
+class TestTags(TestCase):
+
+    def test_normalize_tags(self):
+        claim = Claim()
+
+        claim.channel.update(tags=['Anime', 'anime', ' aNiMe', 'maNGA '])
+        self.assertCountEqual(claim.channel.tags, ['anime', 'manga'])
+
+        claim.channel.update(tags=['Juri', 'juRi'])
+        self.assertCountEqual(claim.channel.tags, ['anime', 'manga', 'juri'])
+
+        claim.channel.update(tags='Anime')
+        self.assertCountEqual(claim.channel.tags, ['anime', 'manga', 'juri'])
+
+        claim.channel.update(clear_tags=True)
+        self.assertEqual(len(claim.channel.tags), 0)
+
+        claim.channel.update(tags='Anime')
+        self.assertEqual(claim.channel.tags, ['anime'])
+
+
 class TestLocations(TestCase):
 
     def test_location_successful_parsing(self):