From af17800bba8d8fd39817b6ad7bbe804cf3b7ceba Mon Sep 17 00:00:00 2001
From: GroovieGermanikus <groovie@mango.markets>
Date: Fri, 7 Jun 2024 14:11:12 +0200
Subject: [PATCH] xor+lz4

---
 Cargo.lock                             | 26 +++++++++++++++++++++
 Cargo.toml                             |  2 ++
 examples/bench_geyser_grpc_accounts.rs | 32 ++++++++++++++++++++++++--
 3 files changed, 58 insertions(+), 2 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index e906d14..13a96b5 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -1372,6 +1372,7 @@ dependencies = [
  "futures",
  "itertools 0.10.5",
  "log",
+ "lz4_flex",
  "merge-streams",
  "regex",
  "solana-account-decoder",
@@ -1846,6 +1847,15 @@ version = "0.4.21"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "90ed8c1e510134f979dbc4f070f87d4313098b704861a105fe34231c70a3901c"
 
+[[package]]
+name = "lz4_flex"
+version = "0.11.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "75761162ae2b0e580d7e7c390558127e5f01b4194debd6221fd8c207fc80e3f5"
+dependencies = [
+ "twox-hash",
+]
+
 [[package]]
 name = "matchit"
 version = "0.7.3"
@@ -3493,6 +3503,12 @@ dependencies = [
  "spl-program-error",
 ]
 
+[[package]]
+name = "static_assertions"
+version = "1.1.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a2eb9349b6444b326872e140eb1cf5e7c522154d69e7a0ffb0fb81c06b37543f"
+
 [[package]]
 name = "strsim"
 version = "0.10.0"
@@ -3902,6 +3918,16 @@ version = "0.2.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "e421abadd41a4225275504ea4d6566923418b7f05506fbc9c0fe86ba7396114b"
 
+[[package]]
+name = "twox-hash"
+version = "1.6.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "97fee6b57c6a41524a810daee9286c02d7752c4253064d0b05472833a438f675"
+dependencies = [
+ "cfg-if",
+ "static_assertions",
+]
+
 [[package]]
 name = "typenum"
 version = "1.17.0"
diff --git a/Cargo.toml b/Cargo.toml
index 177d7cd..f5699ea 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -38,6 +38,8 @@ tonic-health = "0.10.2"
 regex = "1.10.4"
 clap = { version = "4.2", features = ["derive"] }
 
+lz4_flex = "0.11.3"
+
 [dev-dependencies]
 tracing-subscriber = "0.3.16"
 solana-logger = "1"
diff --git a/examples/bench_geyser_grpc_accounts.rs b/examples/bench_geyser_grpc_accounts.rs
index 3b88d78..c82e2d5 100644
--- a/examples/bench_geyser_grpc_accounts.rs
+++ b/examples/bench_geyser_grpc_accounts.rs
@@ -1,3 +1,4 @@
+use std::cmp::min;
 use std::collections::{HashMap, VecDeque};
 use futures::{Stream, StreamExt};
 use log::{debug, info};
@@ -225,10 +226,15 @@ fn start_tracking_account_consumer(mut geyser_messages_rx: Receiver<Message>, cu
                             info!("got account update!!! {} - {:?} - {} bytes - {}",
                                 slot, account_pk, account_info.data.len(), account_info.lamports);
 
-                            if let Some(data) = last_account_data {
-                                let hash1 = hash(&data);
+                            if let Some(prev_data) = last_account_data {
+                                let hash1 = hash(&prev_data);
                                 let hash2 = hash(&account_info.data);
                                 info!("diff: {} {}", hash1, hash2);
+                                
+                                if hash1 != hash2 {
+                                    delta_compress(&prev_data, &account_info.data);
+                                }
+                                
                             }
 
                             last_account_data = Some(account_info.data.clone());
@@ -302,6 +308,28 @@ fn start_tracking_account_consumer(mut geyser_messages_rx: Receiver<Message>, cu
     });
 }
 
+fn delta_compress(prev_data: &Vec<u8>, data: &Vec<u8>) {
+
+    let xor_region = min(prev_data.len(), data.len());
+    let mut xor_diff = vec![0u8; xor_region];
+
+    for i in 0..xor_region {
+        xor_diff[i] = prev_data[i] ^ data[i];
+    }
+
+    // TODO https://users.rust-lang.org/t/how-to-find-common-prefix-of-two-byte-slices-effectively/25815/3
+    let count_non_zero = xor_diff.iter().filter(|&x| *x != 0).count();
+    info!("count_non_zero={} xor_region={}", count_non_zero, xor_region);
+    // info!("hex {:02X?}", xor_data);
+
+    let compressed_xor = lz4_flex::compress_prepend_size(&xor_diff);
+    info!("compressed size of xor: {} (was {})", compressed_xor.len(), xor_diff.len());
+
+    let compressed_data = lz4_flex::compress_prepend_size(&data);
+    info!("compressed size of data: {} (was {})", compressed_data.len(), data.len());
+
+}
+
 fn get_epoch_sec() -> UnixTimestamp {
     SystemTime::now()
         .duration_since(UNIX_EPOCH)