using --has_source with claim_search now properly handles reposts and channels

lbry/wallet/server/db/elasticsearch/constants.py

@@ -42,7 +42,7 @@
           'claims_in_channel', 'channel_join', 'signature_valid', 'effective_amount', 'support_amount',
           'trending_group', 'trending_mixed', 'trending_local', 'trending_global', 'channel_id', 'tx_id', 'tx_nout',
           'signature', 'signature_digest', 'public_key_bytes', 'public_key_hash', 'public_key_id', '_id', 'tags',
-          'reposted_claim_id', 'has_source'}
+          'reposted_claim_id'}
 TEXT_FIELDS = {'author', 'canonical_url', 'channel_id', 'claim_name', 'description', 'claim_id',
                'media_type', 'normalized', 'public_key_bytes', 'public_key_hash', 'short_url', 'signature',
                'signature_digest', 'stream_type', 'title', 'tx_id', 'fee_currency', 'reposted_claim_id', 'tags'}

lbry/wallet/server/db/elasticsearch/search.py

@@ -429,6 +429,13 @@ def expand_query(**kwargs):
         query["minimum_should_match"] = 1
         query['should'].append({"bool": {"must_not": {"exists": {"field": "signature_digest"}}}})
         query['should'].append({"term": {"signature_valid": bool(kwargs["signature_valid"])}})
+    if 'has_source' in kwargs:
+        query.setdefault('should', [])
+        query["minimum_should_match"] = 1
+        is_stream_or_repost = {"terms": {"claim_type": [CLAIM_TYPES['stream'], CLAIM_TYPES['repost']]}}
+        query['should'].append(
+            {"bool": {"must": [{"match": {"has_source": kwargs['has_source']}}, is_stream_or_repost]}})
+        query['should'].append({"bool": {"must_not": [is_stream_or_repost]}})
     if kwargs.get('text'):
         query['must'].append(
                     {"simple_query_string":

lbry/wallet/server/db/elasticsearch/sync.py

@@ -28,12 +28,14 @@ def exec_factory(cursor, statement, bindings):
        claimtrie.last_take_over_height,
        (select group_concat(tag, ',,') from tag where tag.claim_hash in (claim.claim_hash, claim.reposted_claim_hash)) as tags,
        (select group_concat(language, ' ') from language where language.claim_hash in (claim.claim_hash, claim.reposted_claim_hash)) as languages,
+       (select cr.has_source from claim cr where cr.claim_hash = claim.reposted_claim_hash) as reposted_has_source,
        claim.*
 FROM claim LEFT JOIN claimtrie USING (claim_hash)
 WHERE claim.height % {shards_total} = {shard_num}
 ORDER BY claim.height desc
 """)):
         claim = dict(claim._asdict())
+        claim['has_source'] = bool(claim.pop('reposted_has_source') or claim['has_source'])
         claim['censor_type'] = 0
         claim['censoring_channel_hash'] = None
         claim['tags'] = claim['tags'].split(',,') if claim['tags'] else []

lbry/wallet/server/db/writer.py

@@ -827,6 +827,7 @@ def enqueue_changes(self):
                claimtrie.last_take_over_height,
                (select group_concat(tag, ',,') from tag where tag.claim_hash in (claim.claim_hash, claim.reposted_claim_hash)) as tags,
                (select group_concat(language, ' ') from language where language.claim_hash in (claim.claim_hash, claim.reposted_claim_hash)) as languages,
+               (select cr.has_source from claim cr where cr.claim_hash = claim.reposted_claim_hash) as reposted_has_source,
                claim.*
         FROM claim LEFT JOIN claimtrie USING (claim_hash)
         WHERE claim.claim_hash in (SELECT claim_hash FROM changelog)
@@ -835,6 +836,7 @@ def enqueue_changes(self):
             id_set = set(filter(None, (claim['claim_hash'], claim['channel_hash'], claim['reposted_claim_hash'])))
             claim['censor_type'] = 0
             claim['censoring_channel_hash'] = None
+            claim['has_source'] = bool(claim.pop('reposted_has_source') or claim['has_source'])
             for reason_id in id_set:
                 if reason_id in self.blocked_streams:
                     claim['censor_type'] = 2

tests/integration/blockchain/test_claim_commands.py

@@ -180,11 +180,14 @@ async def test_basic_claim_search(self):
         await self.assertFindsClaims([three], claim_id=self.get_claim_id(three), text='*')
 
     async def test_source_filter(self):
-        no_source = await self.stream_create('no_source', data=None)
+        channel = await self.channel_create('@abc')
+        no_source = await self.stream_create('no-source', data=None)
         normal = await self.stream_create('normal', data=b'normal')
-        await self.assertFindsClaims([no_source], has_no_source=True)
-        await self.assertFindsClaims([normal], has_source=True)
-        await self.assertFindsClaims([normal, no_source])
+        normal_repost = await self.stream_repost(self.get_claim_id(normal), 'normal-repost')
+        no_source_repost = await self.stream_repost(self.get_claim_id(no_source), 'no-source-repost')
+        await self.assertFindsClaims([no_source_repost, no_source, channel], has_no_source=True)
+        await self.assertFindsClaims([normal_repost, normal, channel], has_source=True)
+        await self.assertFindsClaims([no_source_repost, normal_repost, normal, no_source, channel])
 
     async def test_pagination(self):
         await self.create_channel()