Merge pull request ClickHouse#36467 from olevino/wyhash

Wyhash
SipeiZhou · May 11, 2022 · 6d3a54a · 6d3a54a
2 parents 3215c98 + e04dc44
commit 6d3a54a
Show file tree

Hide file tree

Showing 10 changed files with 47 additions and 0 deletions.
diff --git a/.gitmodules b/.gitmodules
@@ -262,3 +262,6 @@
 [submodule "contrib/minizip-ng"]
 	path = contrib/minizip-ng
 	url = https://github.com/zlib-ng/minizip-ng
+[submodule "contrib/wyhash"]
+	path = contrib/wyhash
+	url = https://github.com/wangyi-fudan/wyhash.git
diff --git a/contrib/CMakeLists.txt b/contrib/CMakeLists.txt
@@ -94,6 +94,7 @@ if (ENABLE_FUZZING)
     add_contrib (libprotobuf-mutator-cmake libprotobuf-mutator)
 endif()
 
+add_contrib (wyhash-cmake wyhash)
 add_contrib (cityhash102)
 add_contrib (libfarmhash)
 add_contrib (icu-cmake icu)

diff --git a/contrib/wyhash b/contrib/wyhash
diff --git a/contrib/wyhash-cmake/CMakeLists.txt b/contrib/wyhash-cmake/CMakeLists.txt
@@ -0,0 +1,3 @@
+add_library(wyhash INTERFACE)
+target_include_directories(wyhash SYSTEM BEFORE INTERFACE "${ClickHouse_SOURCE_DIR}/contrib/wyhash")
+add_library(ch_contrib::wyhash ALIAS wyhash)
diff --git a/docker/test/fasttest/run.sh b/docker/test/fasttest/run.sh
@@ -176,6 +176,7 @@ function clone_submodules
             contrib/NuRaft
             contrib/jemalloc
             contrib/replxx
+            contrib/wyhash
         )
 
         git submodule sync

diff --git a/src/Functions/CMakeLists.txt b/src/Functions/CMakeLists.txt
@@ -13,6 +13,7 @@ add_library(clickhouse_functions ${clickhouse_functions_sources})
 
 target_link_libraries(clickhouse_functions
     PUBLIC
+        ch_contrib::wyhash
         ch_contrib::cityhash
         ch_contrib::farmhash
         clickhouse_dictionaries

diff --git a/src/Functions/FunctionsHashing.cpp b/src/Functions/FunctionsHashing.cpp
@@ -39,5 +39,7 @@ void registerFunctionsHashing(FunctionFactory & factory)
 
     factory.registerFunction<FunctionXxHash32>();
     factory.registerFunction<FunctionXxHash64>();
+
+    factory.registerFunction<FunctionWyHash64>();
 }
 }
diff --git a/src/Functions/FunctionsHashing.h b/src/Functions/FunctionsHashing.h
@@ -5,6 +5,7 @@
 #include <metrohash.h>
 #include <MurmurHash2.h>
 #include <MurmurHash3.h>
+#include <wyhash.h>
 
 #include "config_functions.h"
 #include "config_core.h"
@@ -1369,6 +1370,29 @@ class FunctionURLHash : public IFunction
     }
 };
 
+struct ImplWyHash64
+{
+    static constexpr auto name = "wyHash64";
+    using ReturnType = UInt64;
+
+    static UInt64 apply(const char * s, const size_t len)
+    {
+        return wyhash(s, len, 0, _wyp);
+    }
+    static UInt64 combineHashes(UInt64 h1, UInt64 h2)
+    {
+        union
+        {
+            UInt64 u64[2];
+            char chars[16];
+        };
+        u64[0] = h1;
+        u64[1] = h2;
+        return apply(chars, 16);
+    }
+
+    static constexpr bool use_int_hash_for_pods = false;
+};
 
 struct NameIntHash32 { static constexpr auto name = "intHash32"; };
 struct NameIntHash64 { static constexpr auto name = "intHash64"; };
@@ -1406,4 +1430,6 @@ using FunctionHiveHash = FunctionAnyHash<HiveHashImpl>;
 using FunctionXxHash32 = FunctionAnyHash<ImplXxHash32>;
 using FunctionXxHash64 = FunctionAnyHash<ImplXxHash64>;
 
+using FunctionWyHash64 = FunctionAnyHash<ImplWyHash64>;
+
 }
diff --git a/tests/queries/0_stateless/02286_function_wyhash.reference b/tests/queries/0_stateless/02286_function_wyhash.reference
@@ -0,0 +1,4 @@
+\N
+4808886099364463827
+10557035923789874751
+10561902096955922022
diff --git a/tests/queries/0_stateless/02286_function_wyhash.sql b/tests/queries/0_stateless/02286_function_wyhash.sql
@@ -0,0 +1,5 @@
+SELECT wyHash64(NULL);
+SELECT wyHash64('');
+SELECT wyHash64(' ');
+SELECT wyHash64('qwerty');
+