fix: last/first must return not null values

include/libs/function/functionMgt.h

@@ -202,6 +202,7 @@ bool fmIsForbidStreamFunc(int32_t funcId);
 bool fmIsIntervalInterpoFunc(int32_t funcId);
 bool fmIsInterpFunc(int32_t funcId);
 bool fmIsLastRowFunc(int32_t funcId);
+bool fmIsReturnNotNullFunc(int32_t funcId);
 bool fmIsSelectValueFunc(int32_t funcId);
 bool fmIsSystemInfoFunc(int32_t funcId);
 bool fmIsImplicitTsFunc(int32_t funcId);

source/libs/executor/src/executorimpl.c

@@ -1437,7 +1437,8 @@ static void setExecutionContext(SOperatorInfo* pOperator, int32_t numOfOutput, u
   pAggInfo->groupId = groupId;
 }
 
-static void doUpdateNumOfRows(SResultRow* pRow, int32_t numOfExprs, const int32_t* rowCellOffset) {
+static void doUpdateNumOfRows(SqlFunctionCtx* pCtx, SResultRow* pRow, int32_t numOfExprs, const int32_t* rowCellOffset) {
+  bool returnNotNull = false;
   for (int32_t j = 0; j < numOfExprs; ++j) {
     struct SResultRowEntryInfo* pResInfo = getResultEntryInfo(pRow, j, rowCellOffset);
     if (!isRowEntryInitialized(pResInfo)) {
@@ -1447,9 +1448,13 @@ static void doUpdateNumOfRows(SResultRow* pRow, int32_t numOfExprs, const int32_
     if (pRow->numOfRows < pResInfo->numOfRes) {
       pRow->numOfRows = pResInfo->numOfRes;
     }
+
+    if (fmIsReturnNotNullFunc(pCtx[j].functionId)) {
+      returnNotNull = true;
+    }
   }
   // TODO: if all expr skips all blocks, e.g. all null inputs for max function, output one row in final result.
-  if (pRow->numOfRows == 0) {
+  if (pRow->numOfRows == 0 && !returnNotNull) {
     pRow->numOfRows = 1;
   }
 }
@@ -1462,7 +1467,7 @@ int32_t finalizeResultRowIntoResultDataBlock(SDiskbasedBuf* pBuf, SResultRowPosi
   SFilePage*  page = getBufPage(pBuf, resultRowPosition->pageId);
   SResultRow* pRow = (SResultRow*)((char*)page + resultRowPosition->offset);
 
-  doUpdateNumOfRows(pRow, numOfExprs, rowCellOffset);
+  doUpdateNumOfRows(pCtx, pRow, numOfExprs, rowCellOffset);
   if (pRow->numOfRows == 0) {
     releaseBufPage(pBuf, page);
     return 0;
@@ -1518,7 +1523,7 @@ int32_t doCopyToSDataBlock(SExecTaskInfo* pTaskInfo, SSDataBlock* pBlock, SExprI
 
     SResultRow* pRow = (SResultRow*)((char*)page + pPos->pos.offset);
 
-    doUpdateNumOfRows(pRow, numOfExprs, rowCellOffset);
+    doUpdateNumOfRows(pCtx, pRow, numOfExprs, rowCellOffset);
     if (pRow->numOfRows == 0) {
       pGroupResInfo->index += 1;
       releaseBufPage(pBuf, page);

source/libs/function/src/functionMgt.c

@@ -221,6 +221,18 @@ bool fmIsLastRowFunc(int32_t funcId) {
   return FUNCTION_TYPE_LAST_ROW == funcMgtBuiltins[funcId].type;
 }
 
+bool fmIsReturnNotNullFunc(int32_t funcId) {
+  if (funcId < 0 || funcId >= funcMgtBuiltinsNum) {
+    return false;
+  }
+  return FUNCTION_TYPE_LAST == funcMgtBuiltins[funcId].type ||
+         FUNCTION_TYPE_LAST_PARTIAL == funcMgtBuiltins[funcId].type ||
+         FUNCTION_TYPE_LAST_MERGE == funcMgtBuiltins[funcId].type ||
+         FUNCTION_TYPE_FIRST == funcMgtBuiltins[funcId].type ||
+         FUNCTION_TYPE_FIRST_PARTIAL == funcMgtBuiltins[funcId].type ||
+         FUNCTION_TYPE_FIRST_MERGE == funcMgtBuiltins[funcId].type;
+}
+
 bool fmIsSelectValueFunc(int32_t funcId) {
   if (funcId < 0 || funcId >= funcMgtBuiltinsNum) {
     return false;

tests/script/tsim/parser/function.sim

@@ -327,8 +327,8 @@ endi
 
 print =================>td-2610
 sql select stddev(k) from tm2 where ts='2020-12-29 18:46:19.109'
-if $rows != 0 then
+if $rows != 1 then
-  print expect 0, actual:$rows
+  print expect 1, actual:$rows
   return -1
 endi
 sql select twa(k) from tm2 where ts='2020-12-29 18:46:19.109'
@@ -406,7 +406,7 @@ if $data00 != 1.378704626 then
 endi
 
 sql select stddev(c) from m1
-if $rows != 0 then
+if $rows != 1 then
   return -1
 endi
 

tests/system-test/2-query/avg.py

@@ -295,7 +295,7 @@ class TDTestCase:
         tdSql.checkData(0, 0, 4.500000000)
 
         tdSql.query(f" select avg(c1) from {dbname}.stb1 where c1 is null ")
-        tdSql.checkRows(0)
+        tdSql.checkRows(1)
 
 
     def avg_func_filter(self, dbname="db"):

tests/system-test/2-query/distribute_agg_apercentile.py

@@ -86,7 +86,7 @@ class TDTestCase:
     def distribute_agg_query(self, dbname="testdb"):
         # basic filter
         tdSql.query(f"select apercentile(c1 , 20) from {dbname}.stb1 where c1 is null")
-        tdSql.checkRows(0)
+        tdSql.checkRows(1)
 
         tdSql.query(f"select apercentile(c1 , 20) from {dbname}.stb1 where t1=1")
         tdSql.checkData(0,0,2.800000000)

tests/system-test/2-query/distribute_agg_max.py

@@ -168,7 +168,7 @@ class TDTestCase:
     def distribute_agg_query(self, dbname="testdb"):
         # basic filter
         tdSql.query(f"select max(c1) from {dbname}.stb1 where c1 is null")
-        tdSql.checkRows(0)
+        tdSql.checkRows(1)
 
         tdSql.query(f"select max(c1) from {dbname}.stb1 where t1=1")
         tdSql.checkData(0,0,10)

tests/system-test/2-query/distribute_agg_min.py

@@ -167,7 +167,7 @@ class TDTestCase:
     def distribute_agg_query(self, dbname="testdb"):
         # basic filter
         tdSql.query(f"select min(c1) from {dbname}.stb1 where c1 is null")
-        tdSql.checkRows(0)
+        tdSql.checkRows(1)
 
         tdSql.query(f"select min(c1) from {dbname}.stb1 where t1=1")
         tdSql.checkData(0,0,2)

tests/system-test/2-query/distribute_agg_spread.py

@@ -195,7 +195,7 @@ class TDTestCase:
     def distribute_agg_query(self):
         # basic filter
         tdSql.query("select spread(c1) from stb1 where c1 is null")
-        tdSql.checkRows(0)
+        tdSql.checkRows(1)
 
         tdSql.query("select spread(c1) from stb1 where t1=1")
         tdSql.checkData(0,0,8.000000000)

tests/system-test/2-query/sample.py

@@ -743,7 +743,7 @@ class TDTestCase:
         # filter data
 
         tdSql.query(" select sample(c1, 20 ) from t1 where c1 is null ")
-        tdSql.checkRows(0)
+        tdSql.checkRows(1)
 
         tdSql.query(" select sample(c1, 20 ) from t1 where c1 =6 ")
         tdSql.checkRows(1)
@@ -891,4 +891,4 @@ class TDTestCase:
         tdLog.success("%s successfully executed" % __file__)
 
 tdCases.addWindows(__file__, TDTestCase())
-tdCases.addLinux(__file__, TDTestCase())
+tdCases.addLinux(__file__, TDTestCase())