Merge pull request #15828 from taosdata/feature/3.0_query_optimize_wxy

enh: increase explain output, enhance debugging function

source/libs/command/inc/commandInt.h

@@ -43,6 +43,7 @@ extern "C" {
 #define EXPLAIN_GROUP_SORT_FORMAT "Group Sort"
 #define EXPLAIN_INTERVAL_FORMAT "Interval on Column %s"
 #define EXPLAIN_MERGE_INTERVAL_FORMAT "Merge Interval on Column %s"
+#define EXPLAIN_MERGE_ALIGNED_INTERVAL_FORMAT "Merge Aligned Interval on Column %s"
 #define EXPLAIN_FILL_FORMAT "Fill"
 #define EXPLAIN_SESSION_FORMAT "Session"
 #define EXPLAIN_STATE_WINDOW_FORMAT "StateWindow on Column %s"
@@ -59,6 +60,8 @@ extern "C" {
 #define EXPLAIN_RATIO_TIME_FORMAT "Ratio: %f"
 #define EXPLAIN_MERGE_FORMAT "SortMerge"
 #define EXPLAIN_MERGE_KEYS_FORMAT "Merge Key: "
+#define EXPLAIN_PARTITION_KETS_FORMAT "Partition Key: "
+#define EXPLAIN_INTERP_FORMAT "Interp"
 
 #define EXPLAIN_PLANNING_TIME_FORMAT "Planning Time: %.3f ms"
 #define EXPLAIN_EXEC_TIME_FORMAT "Execution Time: %.3f ms"
@@ -69,6 +72,7 @@ extern "C" {
 #define EXPLAIN_LEFT_PARENTHESIS_FORMAT " ("
 #define EXPLAIN_RIGHT_PARENTHESIS_FORMAT ")"
 #define EXPLAIN_BLANK_FORMAT " "
+#define EXPLAIN_COMMA_FORMAT ", "
 #define EXPLAIN_COST_FORMAT "cost=%.2f..%.2f"
 #define EXPLAIN_ROWS_FORMAT "rows=%" PRIu64
 #define EXPLAIN_COLUMNS_FORMAT "columns=%d"
@@ -86,6 +90,7 @@ extern "C" {
 #define EXPLAIN_OUTPUT_ORDER_TYPE_FORMAT "output_order=%s"
 #define EXPLAIN_OFFSET_FORMAT "offset=%d"
 #define EXPLAIN_SOFFSET_FORMAT "soffset=%d"
+#define EXPLAIN_PARTITIONS_FORMAT "partitions=%d"
 
 #define COMMAND_RESET_LOG "resetLog"
 #define COMMAND_SCHEDULE_POLICY "schedulePolicy"

source/libs/command/src/explain.c

@@ -535,6 +535,13 @@ int32_t qExplainResNodeToRowsImpl(SExplainResNode *pResNode, SExplainCtx *ctx, i
         EXPLAIN_ROW_END();
         QRY_ERR_RET(qExplainResAppendRow(ctx, tbuf, tlen, level + 1));
 
+        if (NULL != pTblScanNode->pGroupTags) {
+          EXPLAIN_ROW_NEW(level + 1, EXPLAIN_PARTITION_KETS_FORMAT);
+          EXPLAIN_ROW_APPEND(EXPLAIN_PARTITIONS_FORMAT, pTblScanNode->pGroupTags->length);
+          EXPLAIN_ROW_END();
+          QRY_ERR_RET(qExplainResAppendRow(ctx, tbuf, tlen, level + 1));
+        }
+
         if (pTblScanNode->scan.node.pConditions) {
           EXPLAIN_ROW_NEW(level + 1, EXPLAIN_FILTER_FORMAT);
           QRY_ERR_RET(nodesNodeToSQL(pTblScanNode->scan.node.pConditions, tbuf + VARSTR_HEADER_SIZE,
@@ -612,6 +619,10 @@ int32_t qExplainResNodeToRowsImpl(SExplainResNode *pResNode, SExplainCtx *ctx, i
         EXPLAIN_ROW_END();
         QRY_ERR_RET(qExplainResAppendRow(ctx, tbuf, tlen, level + 1));
 
+        EXPLAIN_ROW_NEW(level + 1, EXPLAIN_MERGEBLOCKS_FORMAT, pPrjNode->mergeDataBlock? "True":"False");
+        EXPLAIN_ROW_END();
+        QRY_ERR_RET(qExplainResAppendRow(ctx, tbuf, tlen, level + 1));  
+
         if (pPrjNode->node.pConditions) {
           EXPLAIN_ROW_NEW(level + 1, EXPLAIN_FILTER_FORMAT);
           QRY_ERR_RET(nodesNodeToSQL(pPrjNode->node.pConditions, tbuf + VARSTR_HEADER_SIZE,
@@ -915,7 +926,7 @@ int32_t qExplainResNodeToRowsImpl(SExplainResNode *pResNode, SExplainCtx *ctx, i
     }
     case QUERY_NODE_PHYSICAL_PLAN_MERGE_ALIGNED_INTERVAL: {
       SMergeAlignedIntervalPhysiNode *pIntNode = (SMergeAlignedIntervalPhysiNode *)pNode;
-      EXPLAIN_ROW_NEW(level, EXPLAIN_INTERVAL_FORMAT, nodesGetNameFromColumnNode(pIntNode->window.pTspk));
+      EXPLAIN_ROW_NEW(level, EXPLAIN_MERGE_ALIGNED_INTERVAL_FORMAT, nodesGetNameFromColumnNode(pIntNode->window.pTspk));
       EXPLAIN_ROW_APPEND(EXPLAIN_LEFT_PARENTHESIS_FORMAT);
       if (pResNode->pExecInfo) {
         QRY_ERR_RET(qExplainBufAppendExecInfo(pResNode->pExecInfo, tbuf, &tlen));
@@ -1129,6 +1140,11 @@ int32_t qExplainResNodeToRowsImpl(SExplainResNode *pResNode, SExplainCtx *ctx, i
         EXPLAIN_ROW_END();
         QRY_ERR_RET(qExplainResAppendRow(ctx, tbuf, tlen, level + 1));
 
+        EXPLAIN_ROW_NEW(level + 1, EXPLAIN_PARTITION_KETS_FORMAT);
+        EXPLAIN_ROW_APPEND(EXPLAIN_PARTITIONS_FORMAT, pPartNode->pPartitionKeys->length);
+        EXPLAIN_ROW_END();
+        QRY_ERR_RET(qExplainResAppendRow(ctx, tbuf, tlen, level + 1));
+
         if (pPartNode->node.pConditions) {
           EXPLAIN_ROW_NEW(level + 1, EXPLAIN_FILTER_FORMAT);
           QRY_ERR_RET(nodesNodeToSQL(pPartNode->node.pConditions, tbuf + VARSTR_HEADER_SIZE,
@@ -1201,11 +1217,20 @@ int32_t qExplainResNodeToRowsImpl(SExplainResNode *pResNode, SExplainCtx *ctx, i
         QRY_ERR_RET(qExplainResAppendRow(ctx, tbuf, tlen, level + 1));
 
         EXPLAIN_ROW_NEW(level + 1, EXPLAIN_MERGE_KEYS_FORMAT);
+        if (pMergeNode->groupSort) {
+          EXPLAIN_ROW_APPEND(EXPLAIN_STRING_TYPE_FORMAT, "_group_id asc");
+          if (LIST_LENGTH(pMergeNode->pMergeKeys) > 0) {
+            EXPLAIN_ROW_APPEND(EXPLAIN_COMMA_FORMAT);
+          }
+        }
         for (int32_t i = 0; i < LIST_LENGTH(pMergeNode->pMergeKeys); ++i) {
           SOrderByExprNode *ptn = (SOrderByExprNode *)nodesListGetNode(pMergeNode->pMergeKeys, i);
           EXPLAIN_ROW_APPEND(EXPLAIN_STRING_TYPE_FORMAT, nodesGetNameFromColumnNode(ptn->pExpr));
           EXPLAIN_ROW_APPEND(EXPLAIN_BLANK_FORMAT);
           EXPLAIN_ROW_APPEND(EXPLAIN_STRING_TYPE_FORMAT, EXPLAIN_ORDER_STRING(ptn->order));
+          if (i != LIST_LENGTH(pMergeNode->pMergeKeys) - 1) {
+            EXPLAIN_ROW_APPEND(EXPLAIN_COMMA_FORMAT);
+          }
         }
         EXPLAIN_ROW_END();
         QRY_ERR_RET(qExplainResAppendRow(ctx, tbuf, tlen, level + 1));
@@ -1417,7 +1442,7 @@ int32_t qExplainResNodeToRowsImpl(SExplainResNode *pResNode, SExplainCtx *ctx, i
     }
     case QUERY_NODE_PHYSICAL_PLAN_INTERP_FUNC: {
       SInterpFuncPhysiNode *pInterpNode = (SInterpFuncPhysiNode *)pNode;
-      EXPLAIN_ROW_NEW(level, EXPLAIN_AGG_FORMAT);
+      EXPLAIN_ROW_NEW(level, EXPLAIN_INTERP_FORMAT);
       EXPLAIN_ROW_APPEND(EXPLAIN_LEFT_PARENTHESIS_FORMAT);
       if (pResNode->pExecInfo) {
         QRY_ERR_RET(qExplainBufAppendExecInfo(pResNode->pExecInfo, tbuf, &tlen));

source/libs/function/src/builtins.c

@@ -1575,7 +1575,8 @@ static int32_t translateDiff(SFunctionNode* pFunc, char* pErrBuf, int32_t len) {
   }
 
   uint8_t colType = ((SExprNode*)nodesListGetNode(pFunc->pParameterList, 0))->resType.type;
-  if (!IS_SIGNED_NUMERIC_TYPE(colType) && !IS_FLOAT_TYPE(colType) && TSDB_DATA_TYPE_BOOL != colType && TSDB_DATA_TYPE_TIMESTAMP != colType) {
+  if (!IS_SIGNED_NUMERIC_TYPE(colType) && !IS_FLOAT_TYPE(colType) && TSDB_DATA_TYPE_BOOL != colType &&
+      TSDB_DATA_TYPE_TIMESTAMP != colType) {
     return invaildFuncParaTypeErrMsg(pErrBuf, len, pFunc->functionName);
   }
 
@@ -2150,7 +2151,7 @@ const SBuiltinFuncDefinition funcMgtBuiltins[] = {
   },
   {
     .name = "_apercentile_partial",
-    .type = FUNCTION_TYPE_APERCENTILE_PARTIAL | FUNC_MGT_TIMELINE_FUNC,
+    .type = FUNCTION_TYPE_APERCENTILE_PARTIAL,
     .classification = FUNC_MGT_AGG_FUNC,
     .translateFunc = translateApercentilePartial,
     .getEnvFunc   = getApercentileFuncEnv,
@@ -2163,7 +2164,7 @@ const SBuiltinFuncDefinition funcMgtBuiltins[] = {
   {
     .name = "_apercentile_merge",
     .type = FUNCTION_TYPE_APERCENTILE_MERGE,
-    .classification = FUNC_MGT_AGG_FUNC | FUNC_MGT_TIMELINE_FUNC,
+    .classification = FUNC_MGT_AGG_FUNC,
     .translateFunc = translateApercentileMerge,
     .getEnvFunc   = getApercentileFuncEnv,
     .initFunc     = apercentileFunctionSetup,

source/libs/parser/src/parInsertData.c

@@ -486,11 +486,11 @@ static int32_t tdBlockRowMerge(STableMeta* pTableMeta, SBlockKeyTuple* pEndKeyTp
   void* pDestRow = NULL;
   ++((*pBlkRowMerger)->index);
   if ((*pBlkRowMerger)->index < taosArrayGetSize((*pBlkRowMerger)->rowArray)) {
-    void* pAlloc = *(void**)taosArrayGet((*pBlkRowMerger)->rowArray, (*pBlkRowMerger)->index);
-    if (tRealloc((uint8_t**)&pAlloc, rowSize) != 0) {
+    void** pAlloc = (void**)taosArrayGet((*pBlkRowMerger)->rowArray, (*pBlkRowMerger)->index);
+    if (tRealloc((uint8_t**)pAlloc, rowSize) != 0) {
       return TSDB_CODE_FAILED;
     }
-    pDestRow = pAlloc;
+    pDestRow = *pAlloc;
   } else {
     if (tRealloc((uint8_t**)&pDestRow, rowSize) != 0) {
       return TSDB_CODE_FAILED;

source/libs/parser/src/parTranslater.c

@@ -2543,7 +2543,6 @@ static int32_t translateWindow(STranslateContext* pCxt, SSelectStmt* pSelect) {
   if (NULL == pSelect->pWindow) {
     return TSDB_CODE_SUCCESS;
   }
-  pSelect->isTimeLineResult = true;
   pCxt->currClause = SQL_CLAUSE_WINDOW;
   int32_t code = translateExpr(pCxt, &pSelect->pWindow);
   if (TSDB_CODE_SUCCESS == code) {
@@ -2614,7 +2613,6 @@ static int32_t translatePartitionBy(STranslateContext* pCxt, SSelectStmt* pSelec
   if (NULL == pSelect->pPartitionByList) {
     return TSDB_CODE_SUCCESS;
   }
-  pSelect->isTimeLineResult = false;
   pCxt->currClause = SQL_CLAUSE_PARTITION_BY;
   return translateExprList(pCxt, pSelect->pPartitionByList);
 }

source/libs/planner/src/planLogicCreater.c

@@ -342,7 +342,7 @@ static int32_t createJoinLogicNode(SLogicPlanContext* pCxt, SSelectStmt* pSelect
   pJoin->inputTsOrder = ORDER_ASC;
   pJoin->node.groupAction = GROUP_ACTION_CLEAR;
   pJoin->node.requireDataOrder = DATA_ORDER_LEVEL_GLOBAL;
-  pJoin->node.requireDataOrder = DATA_ORDER_LEVEL_GLOBAL;
+  pJoin->node.resultDataOrder = DATA_ORDER_LEVEL_GLOBAL;
 
   int32_t code = TSDB_CODE_SUCCESS;
 
@@ -472,6 +472,11 @@ static EGroupAction getGroupAction(SLogicPlanContext* pCxt, SSelectStmt* pSelect
   return (pCxt->pPlanCxt->streamQuery || NULL != pSelect->pLimit) ? GROUP_ACTION_KEEP : GROUP_ACTION_NONE;
 }
 
+static EDataOrderLevel getRequireDataOrder(bool needTimeline, SSelectStmt* pSelect) {
+  return needTimeline ? (NULL != pSelect->pPartitionByList ? DATA_ORDER_LEVEL_IN_GROUP : DATA_ORDER_LEVEL_GLOBAL)
+                      : DATA_ORDER_LEVEL_NONE;
+}
+
 static int32_t createAggLogicNode(SLogicPlanContext* pCxt, SSelectStmt* pSelect, SLogicNode** pLogicNode) {
   if (!pSelect->hasAggFuncs && NULL == pSelect->pGroupByList) {
     return TSDB_CODE_SUCCESS;
@@ -486,8 +491,8 @@ static int32_t createAggLogicNode(SLogicPlanContext* pCxt, SSelectStmt* pSelect,
   pAgg->hasTimeLineFunc = pSelect->hasTimeLineFunc;
   pAgg->onlyHasKeepOrderFunc = pSelect->onlyHasKeepOrderFunc;
   pAgg->node.groupAction = getGroupAction(pCxt, pSelect);
-  pAgg->node.requireDataOrder = pAgg->hasTimeLineFunc ? DATA_ORDER_LEVEL_IN_GROUP : DATA_ORDER_LEVEL_NONE;
-  pAgg->node.resultDataOrder = DATA_ORDER_LEVEL_NONE;
+  pAgg->node.requireDataOrder = getRequireDataOrder(pAgg->hasTimeLineFunc, pSelect);
+  pAgg->node.resultDataOrder = pAgg->onlyHasKeepOrderFunc ? pAgg->node.requireDataOrder : DATA_ORDER_LEVEL_NONE;
 
   int32_t code = TSDB_CODE_SUCCESS;
 
@@ -557,8 +562,7 @@ static int32_t createIndefRowsFuncLogicNode(SLogicPlanContext* pCxt, SSelectStmt
   pIdfRowsFunc->isUniqueFunc = pSelect->hasUniqueFunc;
   pIdfRowsFunc->isTimeLineFunc = pSelect->hasTimeLineFunc;
   pIdfRowsFunc->node.groupAction = getGroupAction(pCxt, pSelect);
-  pIdfRowsFunc->node.requireDataOrder =
-      pIdfRowsFunc->isTimeLineFunc ? DATA_ORDER_LEVEL_IN_GROUP : DATA_ORDER_LEVEL_NONE;
+  pIdfRowsFunc->node.requireDataOrder = getRequireDataOrder(pIdfRowsFunc->isTimeLineFunc, pSelect);
   pIdfRowsFunc->node.resultDataOrder = pIdfRowsFunc->node.requireDataOrder;
 
   // indefinite rows functions and _select_values functions
@@ -592,7 +596,7 @@ static int32_t createInterpFuncLogicNode(SLogicPlanContext* pCxt, SSelectStmt* p
   }
 
   pInterpFunc->node.groupAction = getGroupAction(pCxt, pSelect);
-  pInterpFunc->node.requireDataOrder = DATA_ORDER_LEVEL_IN_GROUP;
+  pInterpFunc->node.requireDataOrder = getRequireDataOrder(true, pSelect);
   pInterpFunc->node.resultDataOrder = pInterpFunc->node.requireDataOrder;
 
   int32_t code = nodesCollectFuncs(pSelect, SQL_CLAUSE_SELECT, fmIsInterpFunc, &pInterpFunc->pFuncs);
@@ -668,8 +672,10 @@ static int32_t createWindowLogicNodeByState(SLogicPlanContext* pCxt, SStateWindo
 
   pWindow->winType = WINDOW_TYPE_STATE;
   pWindow->node.groupAction = getGroupAction(pCxt, pSelect);
-  pWindow->node.requireDataOrder = pCxt->pPlanCxt->streamQuery ? DATA_ORDER_LEVEL_IN_BLOCK : DATA_ORDER_LEVEL_IN_GROUP;
-  pWindow->node.resultDataOrder = DATA_ORDER_LEVEL_IN_GROUP;
+  pWindow->node.requireDataOrder =
+      pCxt->pPlanCxt->streamQuery ? DATA_ORDER_LEVEL_IN_BLOCK : getRequireDataOrder(true, pSelect);
+  pWindow->node.resultDataOrder =
+      pCxt->pPlanCxt->streamQuery ? DATA_ORDER_LEVEL_GLOBAL : pWindow->node.requireDataOrder;
   pWindow->pStateExpr = nodesCloneNode(pState->pExpr);
   pWindow->pTspk = nodesCloneNode(pState->pCol);
   if (NULL == pWindow->pStateExpr || NULL == pWindow->pTspk) {
@@ -691,8 +697,10 @@ static int32_t createWindowLogicNodeBySession(SLogicPlanContext* pCxt, SSessionW
   pWindow->sessionGap = ((SValueNode*)pSession->pGap)->datum.i;
   pWindow->windowAlgo = pCxt->pPlanCxt->streamQuery ? SESSION_ALGO_STREAM_SINGLE : SESSION_ALGO_MERGE;
   pWindow->node.groupAction = getGroupAction(pCxt, pSelect);
-  pWindow->node.requireDataOrder = pCxt->pPlanCxt->streamQuery ? DATA_ORDER_LEVEL_IN_BLOCK : DATA_ORDER_LEVEL_IN_GROUP;
-  pWindow->node.resultDataOrder = DATA_ORDER_LEVEL_IN_GROUP;
+  pWindow->node.requireDataOrder =
+      pCxt->pPlanCxt->streamQuery ? DATA_ORDER_LEVEL_IN_BLOCK : getRequireDataOrder(true, pSelect);
+  pWindow->node.resultDataOrder =
+      pCxt->pPlanCxt->streamQuery ? DATA_ORDER_LEVEL_GLOBAL : pWindow->node.requireDataOrder;
 
   pWindow->pTspk = nodesCloneNode((SNode*)pSession->pCol);
   if (NULL == pWindow->pTspk) {
@@ -720,8 +728,12 @@ static int32_t createWindowLogicNodeByInterval(SLogicPlanContext* pCxt, SInterva
       (NULL != pInterval->pSliding ? ((SValueNode*)pInterval->pSliding)->unit : pWindow->intervalUnit);
   pWindow->windowAlgo = pCxt->pPlanCxt->streamQuery ? INTERVAL_ALGO_STREAM_SINGLE : INTERVAL_ALGO_HASH;
   pWindow->node.groupAction = (NULL != pInterval->pFill ? GROUP_ACTION_KEEP : getGroupAction(pCxt, pSelect));
-  pWindow->node.requireDataOrder = pSelect->hasTimeLineFunc ? DATA_ORDER_LEVEL_IN_GROUP : DATA_ORDER_LEVEL_IN_BLOCK;
-  pWindow->node.resultDataOrder = DATA_ORDER_LEVEL_IN_GROUP;
+  pWindow->node.requireDataOrder =
+      pCxt->pPlanCxt->streamQuery
+          ? DATA_ORDER_LEVEL_IN_BLOCK
+          : (pSelect->hasTimeLineFunc ? getRequireDataOrder(true, pSelect) : DATA_ORDER_LEVEL_IN_BLOCK);
+  pWindow->node.resultDataOrder =
+      pCxt->pPlanCxt->streamQuery ? DATA_ORDER_LEVEL_GLOBAL : getRequireDataOrder(true, pSelect);
 
   pWindow->pTspk = nodesCloneNode(pInterval->pCol);
   if (NULL == pWindow->pTspk) {
@@ -768,8 +780,8 @@ static int32_t createFillLogicNode(SLogicPlanContext* pCxt, SSelectStmt* pSelect
   }
 
   pFill->node.groupAction = getGroupAction(pCxt, pSelect);
-  pFill->node.requireDataOrder = DATA_ORDER_LEVEL_IN_GROUP;
-  pFill->node.resultDataOrder = DATA_ORDER_LEVEL_IN_GROUP;
+  pFill->node.requireDataOrder = getRequireDataOrder(true, pSelect);
+  pFill->node.resultDataOrder = pFill->node.requireDataOrder;
   pFill->inputTsOrder = ORDER_ASC;
 
   int32_t code = nodesCollectColumns(pSelect, SQL_CLAUSE_WINDOW, NULL, COLLECT_COL_TYPE_ALL, &pFill->node.pTargets);

source/libs/planner/src/planPhysiCreater.c

@@ -985,7 +985,7 @@ static bool projectCanMergeDataBlock(SProjectLogicNode* pProject) {
     return true;
   }
   if (1 != LIST_LENGTH(pProject->node.pChildren)) {
-    return false;
+    return true;
   }
   SLogicNode* pChild = (SLogicNode*)nodesListGetNode(pProject->node.pChildren, 0);
   return DATA_ORDER_LEVEL_GLOBAL == pChild->resultDataOrder ? true : false;