Merge pull request #14955 from taosdata/feat/agg_client_api

feat(query): add percentile/apercentile scalar version

include/libs/scalar/scalar.h

@@ -104,6 +104,9 @@ int32_t maxScalarFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *
 int32_t avgScalarFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput);
 int32_t stddevScalarFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput);
 int32_t leastSQRScalarFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput);
+int32_t percentileScalarFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput);
+int32_t apercentileScalarFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput);
+int32_t spreadScalarFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput);
 
 #ifdef __cplusplus
 }

source/libs/function/src/builtins.c

@@ -2034,6 +2034,7 @@ const SBuiltinFuncDefinition funcMgtBuiltins[] = {
     .getEnvFunc   = getPercentileFuncEnv,
     .initFunc     = percentileFunctionSetup,
     .processFunc  = percentileFunction,
+    .sprocessFunc = percentileScalarFunction,
     .finalizeFunc = percentileFinalize,
     .invertFunc   = NULL,
     .combineFunc  = NULL,
@@ -2046,6 +2047,7 @@ const SBuiltinFuncDefinition funcMgtBuiltins[] = {
     .getEnvFunc   = getApercentileFuncEnv,
     .initFunc     = apercentileFunctionSetup,
     .processFunc  = apercentileFunction,
+    .sprocessFunc = apercentileScalarFunction,
     .finalizeFunc = apercentileFinalize,
     .invertFunc   = NULL,
     .combineFunc  = apercentileCombine,
@@ -2113,6 +2115,7 @@ const SBuiltinFuncDefinition funcMgtBuiltins[] = {
     .getEnvFunc   = getSpreadFuncEnv,
     .initFunc     = spreadFunctionSetup,
     .processFunc  = spreadFunction,
+    .sprocessFunc = spreadScalarFunction,
     .finalizeFunc = spreadFinalize,
     .invertFunc   = NULL,
     .combineFunc  = spreadCombine,

source/libs/parser/src/parTranslater.c

@@ -1167,7 +1167,11 @@ static int32_t translateRepeatScanFunc(STranslateContext* pCxt, SFunctionNode* p
   if (!fmIsRepeatScanFunc(pFunc->funcId)) {
     return TSDB_CODE_SUCCESS;
   }
-  if (isSelectStmt(pCxt->pCurrStmt) && NULL != ((SSelectStmt*)pCxt->pCurrStmt)->pFromTable) {
+  if (isSelectStmt(pCxt->pCurrStmt)) {
+    //select percentile() without from clause is also valid
+    if (NULL == ((SSelectStmt*)pCxt->pCurrStmt)->pFromTable) {
+      return TSDB_CODE_SUCCESS;
+    }
     SNode* pTable = ((SSelectStmt*)pCxt->pCurrStmt)->pFromTable;
     if (QUERY_NODE_REAL_TABLE == nodeType(pTable) &&
         (TSDB_CHILD_TABLE == ((SRealTableNode*)pTable)->pMeta->tableType ||

source/libs/scalar/src/scalar.c

@@ -700,7 +700,7 @@ EDealRes sclRewriteNonConstOperator(SNode** pNode, SScalarCtx *ctx) {
 EDealRes sclRewriteFunction(SNode** pNode, SScalarCtx *ctx) {
   SFunctionNode *node = (SFunctionNode *)*pNode;
   SNode* tnode = NULL;
-  if (!fmIsScalarFunc(node->funcId)) {
+  if (!fmIsScalarFunc(node->funcId) && (!ctx->dual)) {
     return DEAL_RES_CONTINUE;
   }
 

source/libs/scalar/src/sclfunc.c

@@ -2307,3 +2307,76 @@ int32_t leastSQRScalarFunction(SScalarParam *pInput, int32_t inputNum, SScalarPa
   pOutput->numOfRows = 1;
   return TSDB_CODE_SUCCESS;
 }
+
+int32_t percentileScalarFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
+  SColumnInfoData *pInputData  = pInput->columnData;
+  SColumnInfoData *pOutputData = pOutput->columnData;
+
+  int32_t type = GET_PARAM_TYPE(pInput);
+
+  double val;
+  bool hasNull = false;
+  for (int32_t i = 0; i < pInput->numOfRows; ++i) {
+    if (colDataIsNull_s(pInputData, i)) {
+      hasNull = true;
+      break;
+    }
+    char *in = pInputData->pData;
+    GET_TYPED_DATA(val, double, type, in);
+  }
+
+  if (hasNull) {
+    colDataAppendNULL(pOutputData, 0);
+  } else {
+    colDataAppend(pOutputData, 0, (char *)&val, false);
+  }
+
+  pOutput->numOfRows = 1;
+  return TSDB_CODE_SUCCESS;
+}
+
+int32_t apercentileScalarFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
+  return percentileScalarFunction(pInput, inputNum, pOutput);
+}
+
+int32_t spreadScalarFunction(SScalarParam *pInput, int32_t inputNum, SScalarParam *pOutput) {
+  SColumnInfoData *pInputData  = pInput->columnData;
+  SColumnInfoData *pOutputData = pOutput->columnData;
+
+  int32_t type = GET_PARAM_TYPE(pInput);
+
+  double min, max;
+  SET_DOUBLE_VAL(&min, DBL_MAX);
+  SET_DOUBLE_VAL(&max, -DBL_MAX);
+
+  bool hasNull = false;
+  for (int32_t i = 0; i < pInput->numOfRows; ++i) {
+    if (colDataIsNull_s(pInputData, i)) {
+      hasNull = true;
+      break;
+    }
+
+    char *in = pInputData->pData;
+
+    double val = 0;
+    GET_TYPED_DATA(val, double, type, in);
+
+    if (val < GET_DOUBLE_VAL(&min)) {
+      SET_DOUBLE_VAL(&min, val);
+    }
+
+    if (val > GET_DOUBLE_VAL(&max)) {
+      SET_DOUBLE_VAL(&max, val);
+    }
+  }
+
+  if (hasNull) {
+    colDataAppendNULL(pOutputData, 0);
+  } else {
+    double result = max - min;
+    colDataAppend(pOutputData, 0, (char *)&result, false);
+  }
+
+  pOutput->numOfRows = 1;
+  return TSDB_CODE_SUCCESS;
+}