From 1eb59a2fedc3a960c6478cba1865924311f54938 Mon Sep 17 00:00:00 2001
From: Ganlin Zhao <ganlinzhao@hotmail.com>
Date: Tue, 17 May 2022 10:34:59 +0800
Subject: [PATCH] feat(query): add hll function

---
 include/libs/function/functionMgt.h     |  1 +
 source/libs/function/inc/builtinsimpl.h |  4 ++++
 source/libs/function/src/builtins.c     | 25 +++++++++++++++++++++++++
 source/libs/function/src/builtinsimpl.c |  4 ++--
 4 files changed, 32 insertions(+), 2 deletions(-)

diff --git a/include/libs/function/functionMgt.h b/include/libs/function/functionMgt.h
index 2b58ed7c0b..aec1476663 100644
--- a/include/libs/function/functionMgt.h
+++ b/include/libs/function/functionMgt.h
@@ -41,6 +41,7 @@ typedef enum EFunctionType {
   FUNCTION_TYPE_SUM,
   FUNCTION_TYPE_TWA,
   FUNCTION_TYPE_HISTOGRAM,
+  FUNCTION_TYPE_HYPERLOGLOG,
 
   // nonstandard SQL function
   FUNCTION_TYPE_BOTTOM = 500,
diff --git a/source/libs/function/inc/builtinsimpl.h b/source/libs/function/inc/builtinsimpl.h
index c25d74911c..99313675a5 100644
--- a/source/libs/function/inc/builtinsimpl.h
+++ b/source/libs/function/inc/builtinsimpl.h
@@ -90,6 +90,10 @@ bool histogramFunctionSetup(SqlFunctionCtx *pCtx, SResultRowEntryInfo* pResultIn
 int32_t histogramFunction(SqlFunctionCtx* pCtx);
 int32_t histogramFinalize(SqlFunctionCtx* pCtx, SSDataBlock* pBlock);
 
+bool getHLLFuncEnv(struct SFunctionNode* pFunc, SFuncExecEnv* pEnv);
+int32_t hllFunction(SqlFunctionCtx* pCtx);
+int32_t hllFinalize(SqlFunctionCtx* pCtx, SSDataBlock* pBlock);
+
 bool getStateFuncEnv(struct SFunctionNode* pFunc, SFuncExecEnv* pEnv);
 bool stateFunctionSetup(SqlFunctionCtx *pCtx, SResultRowEntryInfo* pResultInfo);
 int32_t stateCountFunction(SqlFunctionCtx* pCtx);
diff --git a/source/libs/function/src/builtins.c b/source/libs/function/src/builtins.c
index e41e3c7c39..48165fdd99 100644
--- a/source/libs/function/src/builtins.c
+++ b/source/libs/function/src/builtins.c
@@ -263,6 +263,21 @@ static int32_t translateHistogram(SFunctionNode* pFunc, char* pErrBuf, int32_t l
   return TSDB_CODE_SUCCESS;
 }
 
+static int32_t translateHLL(SFunctionNode* pFunc, char* pErrBuf, int32_t len) {
+  if (1 != LIST_LENGTH(pFunc->pParameterList)) {
+    return invaildFuncParaNumErrMsg(pErrBuf, len, pFunc->functionName);
+  }
+
+  SNode* pPara = nodesListGetNode(pFunc->pParameterList, 0);
+  if (QUERY_NODE_COLUMN != nodeType(pPara)) {
+    return buildFuncErrMsg(pErrBuf, len, TSDB_CODE_FUNC_FUNTION_ERROR,
+                           "The input parameter of HYPERLOGLOG function can only be column");
+  }
+
+  pFunc->node.resType = (SDataType){.bytes = tDataTypes[TSDB_DATA_TYPE_UBIGINT].bytes, .type = TSDB_DATA_TYPE_UBIGINT};
+  return TSDB_CODE_SUCCESS;
+}
+
 static int32_t translateStateCount(SFunctionNode* pFunc, char* pErrBuf, int32_t len) {
   if (3 != LIST_LENGTH(pFunc->pParameterList)) {
     return invaildFuncParaNumErrMsg(pErrBuf, len, pFunc->functionName);
@@ -829,6 +844,16 @@ const SBuiltinFuncDefinition funcMgtBuiltins[] = {
     .processFunc  = histogramFunction,
     .finalizeFunc = histogramFinalize
   },
+  {
+    .name = "hyperloglog",
+    .type = FUNCTION_TYPE_HYPERLOGLOG,
+    .classification = FUNC_MGT_AGG_FUNC,
+    .translateFunc = translateHLL,
+    .getEnvFunc   = getHLLFuncEnv,
+    .initFunc     = functionSetup,
+    .processFunc  = hllFunction,
+    .finalizeFunc = hllFinalize
+  },
   {
     .name = "state_count",
     .type = FUNCTION_TYPE_STATE_COUNT,
diff --git a/source/libs/function/src/builtinsimpl.c b/source/libs/function/src/builtinsimpl.c
index 556015a2ac..7ff3b6fb05 100644
--- a/source/libs/function/src/builtinsimpl.c
+++ b/source/libs/function/src/builtinsimpl.c
@@ -26,7 +26,7 @@
 #define MAVG_MAX_POINTS_NUM      1000
 #define SAMPLE_MAX_POINTS_NUM    1000
 #define TAIL_MAX_POINTS_NUM      100
-#define TAIL_MAX_OFFSET          10
+#define TAIL_MAX_OFFSET          100
 
 #define HLL_BUCKET_BITS 14 // The bits of the bucket
 #define HLL_DATA_BITS (64-HLL_BUCKET_BITS)
@@ -2849,8 +2849,8 @@ int32_t hllFunction(SqlFunctionCtx *pCtx) {
 
     char* data = colDataGetData(pCol, i);
     if (IS_VAR_DATA_TYPE(type)) {
+      bytes = varDataLen(data);
       data = varDataVal(data);
-      bytes -= VARSTR_HEADER_SIZE;
     }
 
     int32_t index = 0;
-- 
GitLab