Merge pull request #12172 from taosdata/feature/3.0_glzhao

feat(query): add histogram function

source/libs/function/inc/builtinsimpl.h

@@ -73,6 +73,11 @@ bool spreadFunctionSetup(SqlFunctionCtx *pCtx, SResultRowEntryInfo* pResultInfo)
 int32_t spreadFunction(SqlFunctionCtx* pCtx);
 int32_t spreadFinalize(SqlFunctionCtx* pCtx, SSDataBlock* pBlock);
 
+bool getHistogramFuncEnv(struct SFunctionNode* pFunc, SFuncExecEnv* pEnv);
+bool histogramFunctionSetup(SqlFunctionCtx *pCtx, SResultRowEntryInfo* pResultInfo);
+int32_t histogramFunction(SqlFunctionCtx* pCtx);
+int32_t histogramFinalize(SqlFunctionCtx* pCtx, SSDataBlock* pBlock);
+
 #ifdef __cplusplus
 }
 #endif

source/libs/function/src/builtins.c

@@ -225,6 +225,26 @@ static int32_t translateSpread(SFunctionNode* pFunc, char* pErrBuf, int32_t len)
   return TSDB_CODE_SUCCESS;
 }
 
+static int32_t translateHistogram(SFunctionNode* pFunc, char* pErrBuf, int32_t len) {
+  if (4 != LIST_LENGTH(pFunc->pParameterList)) {
+    return invaildFuncParaNumErrMsg(pErrBuf, len, pFunc->functionName);
+  }
+
+  uint8_t colType = ((SExprNode*)nodesListGetNode(pFunc->pParameterList, 0))->resType.type;
+  if (!IS_NUMERIC_TYPE(colType)) {
+    return invaildFuncParaTypeErrMsg(pErrBuf, len, pFunc->functionName);
+  }
+
+  if (((SExprNode*)nodesListGetNode(pFunc->pParameterList, 1))->resType.type != TSDB_DATA_TYPE_BINARY ||
+      ((SExprNode*)nodesListGetNode(pFunc->pParameterList, 2))->resType.type != TSDB_DATA_TYPE_BINARY ||
+      ((SExprNode*)nodesListGetNode(pFunc->pParameterList, 3))->resType.type != TSDB_DATA_TYPE_BIGINT) {
+    return invaildFuncParaTypeErrMsg(pErrBuf, len, pFunc->functionName);
+  }
+
+  pFunc->node.resType = (SDataType) { .bytes = tDataTypes[TSDB_DATA_TYPE_DOUBLE].bytes, .type = TSDB_DATA_TYPE_DOUBLE };
+  return TSDB_CODE_SUCCESS;
+}
+
 static int32_t translateLastRow(SFunctionNode* pFunc, char* pErrBuf, int32_t len) {
   // todo
   return TSDB_CODE_SUCCESS;
@@ -599,6 +619,16 @@ const SBuiltinFuncDefinition funcMgtBuiltins[] = {
     .processFunc  = diffFunction,
     .finalizeFunc = functionFinalize
   },
+  {
+    .name = "histogram",
+    .type = FUNCTION_TYPE_HISTOGRAM,
+    .classification = FUNC_MGT_AGG_FUNC,
+    .translateFunc = translateHistogram,
+    .getEnvFunc   = getHistogramFuncEnv,
+    .initFunc     = histogramFunctionSetup,
+    .processFunc  = histogramFunction,
+    .finalizeFunc = histogramFinalize
+  },
   {
     .name = "abs",
     .type = FUNCTION_TYPE_ABS,

source/libs/function/src/builtinsimpl.c

@@ -20,6 +20,8 @@
 #include "tdatablock.h"
 #include "tpercentile.h"
 
+#define HISTOGRAM_MAX_BINS_NUM 100
+
 typedef struct SSumRes {
   union {
     int64_t  isum;
@@ -89,6 +91,22 @@ typedef struct SSpreadInfo {
   double max;
 } SSpreadInfo;
 
+typedef struct SHistoFuncBin {
+  double lower;
+  double upper;
+  union {
+    int64_t count;
+    double  percentage;
+  };
+} SHistoFuncBin;
+
+typedef struct SHistoFuncInfo {
+  int32_t numOfBins;
+  bool    normalized;
+  SHistoFuncBin bins[];
+} SHistoFuncInfo;
+
+
 #define SET_VAL(_info, numOfElem, res) \
   do {                                 \
     if ((numOfElem) <= 0) {            \
@@ -1777,3 +1795,34 @@ int32_t spreadFinalize(SqlFunctionCtx* pCtx, SSDataBlock* pBlock) {
   }
   return functionFinalize(pCtx, pBlock);
 }
+
+bool getHistogramFuncEnv(SFunctionNode* UNUSED_PARAM(pFunc), SFuncExecEnv* pEnv) {
+  pEnv->calcMemSize = sizeof(SHistoFuncInfo) + HISTOGRAM_MAX_BINS_NUM * sizeof(SHistoFuncBin);
+  return true;
+}
+
+bool histogramFunctionSetup(SqlFunctionCtx *pCtx, SResultRowEntryInfo* pResultInfo) {
+  if (!functionSetup(pCtx, pResultInfo)) {
+    return false;
+  }
+
+  SHistoFuncInfo* pInfo = GET_ROWCELL_INTERBUF(pResultInfo);
+  char* binType = pCtx->param[1].param.pz;
+  char* binDesc = pCtx->param[2].param.pz;
+  int64_t nornalized = pCtx->param[3].param.i;
+
+
+  return true;
+}
+
+int32_t histogramFunction(SqlFunctionCtx *pCtx) {
+  return TSDB_CODE_SUCCESS;
+}
+
+int32_t histogramFinalize(SqlFunctionCtx* pCtx, SSDataBlock* pBlock) {
+  SHistoFuncInfo* pInfo = GET_ROWCELL_INTERBUF(GET_RES_INFO(pCtx));
+  //if (pInfo->hasResult == true) {
+  //  SET_DOUBLE_VAL(&pInfo->result, pInfo->max - pInfo->min);
+  //}
+  return functionFinalize(pCtx, pBlock);
+}