feat: the length of binary/nchar type value is the same as that of column

include/libs/nodes/querynodes.h

@@ -81,6 +81,7 @@ typedef struct SValueNode {
   char*     literal;
   bool      isDuration;
   bool      translate;
+  bool      notReserved;
   int16_t   placeholderNo;
   union {
     bool     b;

source/libs/function/src/builtins.c

@@ -14,8 +14,8 @@
  */
 
 #include "builtins.h"
-#include "querynodes.h"
 #include "builtinsimpl.h"
+#include "querynodes.h"
 #include "scalar.h"
 #include "taoserror.h"
 #include "tdatablock.h"
@@ -215,7 +215,7 @@ static int32_t translateTop(SFunctionNode* pFunc, char* pErrBuf, int32_t len) {
     return invaildFuncParaTypeErrMsg(pErrBuf, len, pFunc->functionName);
   }
 
-  SValueNode* pValue = (SValueNode*) pParamNode;
+  SValueNode* pValue = (SValueNode*)pParamNode;
   if (pValue->node.resType.type != TSDB_DATA_TYPE_BIGINT) {
     return invaildFuncParaTypeErrMsg(pErrBuf, len, pFunc->functionName);
   }
@@ -224,6 +224,8 @@ static int32_t translateTop(SFunctionNode* pFunc, char* pErrBuf, int32_t len) {
     return invaildFuncParaValueErrMsg(pErrBuf, len, pFunc->functionName);
   }
 
+  pValue->notReserved = true;
+
   SDataType* pType = &((SExprNode*)nodesListGetNode(pFunc->pParameterList, 0))->resType;
   pFunc->node.resType = (SDataType){.bytes = pType->bytes, .type = pType->type};
   return TSDB_CODE_SUCCESS;
@@ -336,7 +338,7 @@ static int32_t translateStateCount(SFunctionNode* pFunc, char* pErrBuf, int32_t
     return invaildFuncParaTypeErrMsg(pErrBuf, len, pFunc->functionName);
   }
 
-  pFunc->node.resType = (SDataType) { .bytes = tDataTypes[TSDB_DATA_TYPE_BIGINT].bytes, .type = TSDB_DATA_TYPE_BIGINT };
+  pFunc->node.resType = (SDataType){.bytes = tDataTypes[TSDB_DATA_TYPE_BIGINT].bytes, .type = TSDB_DATA_TYPE_BIGINT};
   return TSDB_CODE_SUCCESS;
 }
 
@@ -361,7 +363,7 @@ static int32_t translateStateDuration(SFunctionNode* pFunc, char* pErrBuf, int32
     return invaildFuncParaTypeErrMsg(pErrBuf, len, pFunc->functionName);
   }
 
-  pFunc->node.resType = (SDataType) { .bytes = tDataTypes[TSDB_DATA_TYPE_BIGINT].bytes, .type = TSDB_DATA_TYPE_BIGINT };
+  pFunc->node.resType = (SDataType){.bytes = tDataTypes[TSDB_DATA_TYPE_BIGINT].bytes, .type = TSDB_DATA_TYPE_BIGINT};
   return TSDB_CODE_SUCCESS;
 }
 
@@ -392,7 +394,7 @@ static int32_t translateCsum(SFunctionNode* pFunc, char* pErrBuf, int32_t len) {
     }
   }
 
-  pFunc->node.resType = (SDataType) { .bytes = tDataTypes[resType].bytes, .type = resType};
+  pFunc->node.resType = (SDataType){.bytes = tDataTypes[resType].bytes, .type = resType};
   return TSDB_CODE_SUCCESS;
 }
 
@@ -434,7 +436,7 @@ static int32_t translateSample(SFunctionNode* pFunc, char* pErrBuf, int32_t len)
   }
 
   SExprNode* pCol = (SExprNode*)nodesListGetNode(pFunc->pParameterList, 0);
-  uint8_t colType = pCol->resType.type;
+  uint8_t    colType = pCol->resType.type;
   if (IS_VAR_DATA_TYPE(colType)) {
     pFunc->node.resType = (SDataType){.bytes = pCol->resType.bytes, .type = colType};
   } else {
@@ -463,7 +465,7 @@ static int32_t translateTail(SFunctionNode* pFunc, char* pErrBuf, int32_t len) {
   }
 
   SExprNode* pCol = (SExprNode*)nodesListGetNode(pFunc->pParameterList, 0);
-  uint8_t colType = pCol->resType.type;
+  uint8_t    colType = pCol->resType.type;
   if (IS_VAR_DATA_TYPE(colType)) {
     pFunc->node.resType = (SDataType){.bytes = pCol->resType.bytes, .type = colType};
   } else {
@@ -500,8 +502,7 @@ static int32_t translateUnique(SFunctionNode* pFunc, char* pErrBuf, int32_t len)
 
   SNode* pPara = nodesListGetNode(pFunc->pParameterList, 0);
   if (QUERY_NODE_COLUMN != nodeType(pPara)) {
-    return buildFuncErrMsg(pErrBuf, len, TSDB_CODE_FUNC_FUNTION_ERROR,
-                           "The parameters of UNIQUE can only be columns");
+    return buildFuncErrMsg(pErrBuf, len, TSDB_CODE_FUNC_FUNTION_ERROR, "The parameters of UNIQUE can only be columns");
   }
 
   pFunc->node.resType = ((SExprNode*)pPara)->resType;

source/libs/nodes/src/nodesCloneFuncs.c

@@ -112,6 +112,7 @@ static SNode* valueNodeCopy(const SValueNode* pSrc, SValueNode* pDst) {
   COPY_CHAR_POINT_FIELD(literal);
   COPY_SCALAR_FIELD(isDuration);
   COPY_SCALAR_FIELD(translate);
+  COPY_SCALAR_FIELD(notReserved);
   COPY_SCALAR_FIELD(placeholderNo);
   COPY_SCALAR_FIELD(typeData);
   COPY_SCALAR_FIELD(unit);

source/libs/parser/src/parTranslater.c

@@ -646,12 +646,13 @@ static EDealRes translateValueImpl(STranslateContext* pCxt, SValueNode* pVal, SD
       }
       case TSDB_DATA_TYPE_VARCHAR:
       case TSDB_DATA_TYPE_VARBINARY: {
-        pVal->datum.p = taosMemoryCalloc(1, targetDt.bytes + VARSTR_HEADER_SIZE + 1);
+        pVal->datum.p = taosMemoryCalloc(1, targetDt.bytes + 1);
         if (NULL == pVal->datum.p) {
           return generateDealNodeErrMsg(pCxt, TSDB_CODE_OUT_OF_MEMORY);
         }
-        varDataSetLen(pVal->datum.p, targetDt.bytes);
-        strncpy(varDataVal(pVal->datum.p), pVal->literal, targetDt.bytes);
+        int32_t len = TMIN(targetDt.bytes - VARSTR_HEADER_SIZE, pVal->node.resType.bytes);
+        varDataSetLen(pVal->datum.p, len);
+        strncpy(varDataVal(pVal->datum.p), pVal->literal, len);
         break;
       }
       case TSDB_DATA_TYPE_TIMESTAMP: {
@@ -662,19 +663,18 @@ static EDealRes translateValueImpl(STranslateContext* pCxt, SValueNode* pVal, SD
         break;
       }
       case TSDB_DATA_TYPE_NCHAR: {
-        int32_t bytes = targetDt.bytes * TSDB_NCHAR_SIZE;
-        pVal->datum.p = taosMemoryCalloc(1, bytes + VARSTR_HEADER_SIZE + 1);
+        pVal->datum.p = taosMemoryCalloc(1, targetDt.bytes + 1);
         if (NULL == pVal->datum.p) {
           return generateDealNodeErrMsg(pCxt, TSDB_CODE_OUT_OF_MEMORY);
           ;
         }
 
-        int32_t output = 0;
-        if (!taosMbsToUcs4(pVal->literal, pVal->node.resType.bytes, (TdUcs4*)varDataVal(pVal->datum.p), bytes,
-                           &output)) {
+        int32_t len = 0;
+        if (!taosMbsToUcs4(pVal->literal, pVal->node.resType.bytes, (TdUcs4*)varDataVal(pVal->datum.p),
+                           targetDt.bytes - VARSTR_HEADER_SIZE, &len)) {
           return generateDealNodeErrMsg(pCxt, TSDB_CODE_PAR_WRONG_VALUE_TYPE, pVal->literal);
         }
-        varDataSetLen(pVal->datum.p, output);
+        varDataSetLen(pVal->datum.p, len);
         break;
       }
       case TSDB_DATA_TYPE_JSON:
@@ -690,8 +690,20 @@ static EDealRes translateValueImpl(STranslateContext* pCxt, SValueNode* pVal, SD
   return DEAL_RES_CONTINUE;
 }
 
+static int32_t calcTypeBytes(SDataType dt) {
+  if (TSDB_DATA_TYPE_BINARY == dt.type) {
+    return dt.bytes + VARSTR_HEADER_SIZE;
+  } else if (TSDB_DATA_TYPE_NCHAR == dt.type) {
+    return dt.bytes * TSDB_NCHAR_SIZE + VARSTR_HEADER_SIZE;
+  } else {
+    return dt.bytes;
+  }
+}
+
 static EDealRes translateValue(STranslateContext* pCxt, SValueNode* pVal) {
-  return translateValueImpl(pCxt, pVal, pVal->node.resType);
+  SDataType dt = pVal->node.resType;
+  dt.bytes = calcTypeBytes(dt);
+  return translateValueImpl(pCxt, pVal, dt);
 }
 
 static bool isMultiResFunc(SNode* pNode) {
@@ -2343,16 +2355,6 @@ static int32_t translateAlterDatabase(STranslateContext* pCxt, SAlterDatabaseStm
   return buildCmdMsg(pCxt, TDMT_MND_ALTER_DB, (FSerializeFunc)tSerializeSAlterDbReq, &alterReq);
 }
 
-static int32_t calcTypeBytes(SDataType dt) {
-  if (TSDB_DATA_TYPE_BINARY == dt.type) {
-    return dt.bytes + VARSTR_HEADER_SIZE;
-  } else if (TSDB_DATA_TYPE_NCHAR == dt.type) {
-    return dt.bytes * TSDB_NCHAR_SIZE + VARSTR_HEADER_SIZE;
-  } else {
-    return dt.bytes;
-  }
-}
-
 static int32_t columnDefNodeToField(SNodeList* pList, SArray** pArray) {
   *pArray = taosArrayInit(LIST_LENGTH(pList), sizeof(SField));
   SNode* pNode;
@@ -4088,18 +4090,8 @@ static int32_t createValueFromFunction(STranslateContext* pCxt, SFunctionNode* p
   return scalarCalculateConstants((SNode*)pFunc, (SNode**)pVal);
 }
 
-static int32_t colDataBytesToValueDataBytes(uint8_t type, int32_t bytes) {
-  if (TSDB_DATA_TYPE_VARCHAR == type || TSDB_DATA_TYPE_BINARY == type || TSDB_DATA_TYPE_VARBINARY == type) {
-    return bytes - VARSTR_HEADER_SIZE;
-  } else if (TSDB_DATA_TYPE_NCHAR == type) {
-    return (bytes - VARSTR_HEADER_SIZE) / TSDB_NCHAR_SIZE;
-  }
-  return bytes;
-}
-
 static SDataType schemaToDataType(SSchema* pSchema) {
   SDataType dt = {.type = pSchema->type, .bytes = pSchema->bytes, .precision = 0, .scale = 0};
-  dt.bytes = colDataBytesToValueDataBytes(pSchema->type, pSchema->bytes);
   return dt;
 }
 

source/libs/planner/src/planPhysiCreater.c

@@ -650,6 +650,9 @@ static EDealRes doRewritePrecalcExprs(SNode** pNode, void* pContext) {
   SRewritePrecalcExprsCxt* pCxt = (SRewritePrecalcExprsCxt*)pContext;
   switch (nodeType(*pNode)) {
     case QUERY_NODE_VALUE: {
+      if (((SValueNode*)*pNode)->notReserved) {
+        break;
+      }
       pCxt->errCode = rewriteValueToOperator(pCxt, pNode);
       if (TSDB_CODE_SUCCESS != pCxt->errCode) {
         return DEAL_RES_ERROR;

source/libs/planner/test/planBasicTest.cpp

@@ -50,4 +50,6 @@ TEST_F(PlanBasicTest, func) {
   run("SELECT DIFF(c1) FROM t1");
 
   run("SELECT PERCENTILE(c1, 60) FROM t1");
+
+  run("SELECT TOP(c1, 60) FROM t1");
 }