tree 20bf3bc2dfffab24c6e5c50ada9930297b3ee866
parent 9e2cf0d8d187dd222e5456ac40c8a4a4c347a812
author David Morgan <davidmorgan@google.com> 1615222602 +0100
committer GitHub <noreply@github.com> 1615222602 -0800
gpgsig -----BEGIN PGP SIGNATURE-----
 
 wsBcBAABCAAQBQJgRldKCRBK7hj4Ov3rIwAAdHIIAISaWRldae5yZ5NgghU2gRzP
 U5NSQLTPbQXWp9W5LCivGkUstunHMVkfyqmb22kfEJTHMpS+s5EgLaAPMJyt1oAS
 Fz+Z+KMubhTqxN2ZMhBRkwwWcJP90Sc4sk+zO2OpCaXAYIwMtRkpJXqHRW6HLsYJ
 3r5nd5HjL72iCwyydyyV8JJ44WN7trFxshFgyMpc60cq51AqVwY5aZ0eLG0ppssu
 g+LHyjmhyU5X63xBf3WgCgVGNksK3jiFU2QszLWcli1IVEgpoYTSx7X5X0+2kB9j
 rSN03UyKERsc3+dU6ggk2j3Vd+9rhzMP3OFT7P4QkdNMOP06Cb4ppzafjTp98o0=
 =8a6C
 -----END PGP SIGNATURE-----
 

Optimize `AsyncMessageGrouper`. (#54)

Add a benchmark for a large (~600k) request.

Before this change, grouping the message takes around 14ms. After this change it takes around 1ms.

Deserializing is not affected, but is benchmarked too as it is always needed and so is useful to compare. It takes around 2ms.