automatically batch large resolve requests

lbry/wallet/ledger.py

@@ -752,7 +752,11 @@ class Ledger(metaclass=LedgerRegistry):
 
     async def resolve(self, accounts, urls, **kwargs):
         resolve = partial(self.network.retriable_call, self.network.resolve)
-        txos = (await self._inflate_outputs(resolve(urls), accounts, **kwargs))[0]
+        urls_copy = list(urls)
+        txos = []
+        while urls_copy:
+            batch, urls_copy = urls_copy[:500], urls_copy[500:]
+            txos.extend((await self._inflate_outputs(resolve(batch), accounts, **kwargs))[0])
         assert len(urls) == len(txos), "Mismatch between urls requested for resolve and responses received."
         result = {}
         for url, txo in zip(urls, txos):

tests/integration/blockchain/test_claim_commands.py

@@ -1,6 +1,7 @@
 import os.path
 import tempfile
 import logging
+import asyncio
 from binascii import unhexlify
 from urllib.request import urlopen
 
@@ -79,6 +80,12 @@ class ClaimSearchCommand(ClaimTestCase):
         ] * 23828
         self.assertListEqual([], await self.claim_search(claim_ids=claim_ids))
 
+        # this should do nothing... if the resolve (which is retried) results in the server disconnecting,
+        # it kerplodes
+        await asyncio.wait_for(self.daemon.jsonrpc_resolve([
+            f'0000000000000000000000000000000000000000{i}' for i in range(30000)
+        ]), 30)
+
         # 23829 claim ids makes the request just large enough
         claim_ids = [
             '0000000000000000000000000000000000000000',