本文整理了Java中org.apache.calcite.rel.RelNode.getChildExps方法的一些代码示例，展示了RelNode.getChildExps的具体用法。这些代码示例主要来源于Github/Stackoverflow/Maven等平台，是从一些精选项目中提取出来的代码，具有较强的参考意义，能在一定程度帮忙到你。RelNode.getChildExps方法的具体详情如下：
包路径：org.apache.calcite.rel.RelNode
类名称：RelNode
方法名：getChildExps

RelNode.getChildExps介绍

[英]Returns a list of this relational expression's child expressions. (These are scalar expressions, and so do not include the relational inputs that are returned by #getInputs.

The caller should treat the list as unmodifiable; typical implementations will return an immutable list. If there are no child expressions, returns an empty list, not null.
[中]返回此关系表达式的子表达式的列表。（这些是标量表达式，因此不包括#getInputs返回的关系输入。）。
调用方应将列表视为不可修改；典型的实现将返回一个不可变的列表。如果没有子表达式，则返回空列表，而不是null。

代码示例

代码示例来源：origin: apache/hive

private static boolean validGBParent(RelNode gbNode, RelNode parent) {
 boolean validParent = true;
 // TOODO: Verify GB having is not a seperate filter (if so we shouldn't
 // introduce derived table)
 if (parent instanceof Join || parent instanceof SetOp
   || parent instanceof Aggregate
   || (parent instanceof Filter && ((Aggregate) gbNode).getGroupSet().isEmpty())) {
  validParent = false;
 }
 if (parent instanceof Project) {
  for (RexNode child : parent.getChildExps()) {
   if (child instanceof RexOver || child instanceof RexWinAggCall) {
    // Hive can't handle select rank() over(order by sum(c1)/sum(c2)) from t1 group by c3
    // but can handle    select rank() over (order by c4) from
    // (select sum(c1)/sum(c2)  as c4 from t1 group by c3) t2;
    // so introduce a project on top of this gby.
    return false;
   }
  }
 }
 return validParent;
}

代码示例来源：origin: apache/hive

public static List<ExprNodeDesc> getExprNodes(List<Integer> inputRefs, RelNode inputRel,
  String inputTabAlias) {
 List<ExprNodeDesc> exprNodes = new ArrayList<ExprNodeDesc>();
 List<RexNode> rexInputRefs = getInputRef(inputRefs, inputRel);
 List<RexNode> exprs = inputRel.getChildExps();
 // TODO: Change ExprNodeConverter to be independent of Partition Expr
 ExprNodeConverter exprConv = new ExprNodeConverter(inputTabAlias, inputRel.getRowType(),
   new HashSet<Integer>(), inputRel.getCluster().getTypeFactory());
 for (int index = 0; index < rexInputRefs.size(); index++) {
  // The following check is only a guard against failures.
  // TODO: Knowing which expr is constant in GBY's aggregation function
  // arguments could be better done using Metadata provider of Calcite.
  //check the corresponding expression in exprs to see if it is literal
  if (exprs != null && index < exprs.size() && exprs.get(inputRefs.get(index)) instanceof RexLiteral) {
   //because rexInputRefs represent ref expr corresponding to value in inputRefs it is used to get
   //  corresponding index
   ExprNodeDesc exprNodeDesc = exprConv.visitLiteral((RexLiteral) exprs.get(inputRefs.get(index)));
   exprNodes.add(exprNodeDesc);
  } else {
   RexNode iRef = rexInputRefs.get(index);
   exprNodes.add(iRef.accept(exprConv));
  }
 }
 return exprNodes;
}

代码示例来源：origin: apache/drill

private static boolean validGBParent(RelNode gbNode, RelNode parent) {
 boolean validParent = true;
 // TOODO: Verify GB having is not a seperate filter (if so we shouldn't
 // introduce derived table)
 if (parent instanceof Join || parent instanceof SetOp
   || parent instanceof Aggregate
   || (parent instanceof Filter && ((Aggregate) gbNode).getGroupSet().isEmpty())) {
  validParent = false;
 }
 if (parent instanceof Project) {
  for (RexNode child : parent.getChildExps()) {
   if (child instanceof RexOver || child instanceof RexWinAggCall) {
    // Hive can't handle select rank() over(order by sum(c1)/sum(c2)) from t1 group by c3
    // but can handle    select rank() over (order by c4) from
    // (select sum(c1)/sum(c2)  as c4 from t1 group by c3) t2;
    // so introduce a project on top of this gby.
    return false;
   }
  }
 }
 return validParent;
}

代码示例来源：origin: apache/drill

public static List<ExprNodeDesc> getExprNodes(List<Integer> inputRefs, RelNode inputRel,
  String inputTabAlias) {
 List<ExprNodeDesc> exprNodes = new ArrayList<ExprNodeDesc>();
 List<RexNode> rexInputRefs = getInputRef(inputRefs, inputRel);
 List<RexNode> exprs = inputRel.getChildExps();
 // TODO: Change ExprNodeConverter to be independent of Partition Expr
 ExprNodeConverter exprConv = new ExprNodeConverter(inputTabAlias, inputRel.getRowType(),
   new HashSet<Integer>(), inputRel.getCluster().getTypeFactory());
 for (int index = 0; index < rexInputRefs.size(); index++) {
  // The following check is only a guard against failures.
  // TODO: Knowing which expr is constant in GBY's aggregation function
  // arguments could be better done using Metadata provider of Calcite.
  //check the corresponding expression in exprs to see if it is literal
  if (exprs != null && index < exprs.size() && exprs.get(inputRefs.get(index)) instanceof RexLiteral) {
   //because rexInputRefs represent ref expr corresponding to value in inputRefs it is used to get
   //  corresponding index
   ExprNodeDesc exprNodeDesc = exprConv.visitLiteral((RexLiteral) exprs.get(inputRefs.get(index)));
   exprNodes.add(exprNodeDesc);
  } else {
   RexNode iRef = rexInputRefs.get(index);
   exprNodes.add(iRef.accept(exprConv));
  }
 }
 return exprNodes;
}

代码示例来源：origin: apache/hive

private static void replaceEmptyGroupAggr(final RelNode rel, RelNode parent) {
  // If this function is called, the parent should only include constant
  List<RexNode> exps = parent.getChildExps();
  for (RexNode rexNode : exps) {
   if (!rexNode.accept(new HiveCalciteUtil.ConstantFinder())) {
    throw new RuntimeException("We expect " + parent.toString()
      + " to contain only constants. However, " + rexNode.toString() + " is "
      + rexNode.getKind());
   }
  }
  HiveAggregate oldAggRel = (HiveAggregate) rel;
  RelDataTypeFactory typeFactory = oldAggRel.getCluster().getTypeFactory();
  RelDataType longType = TypeConverter.convert(TypeInfoFactory.longTypeInfo, typeFactory);
  RelDataType intType = TypeConverter.convert(TypeInfoFactory.intTypeInfo, typeFactory);
  // Create the dummy aggregation.
  SqlAggFunction countFn = SqlFunctionConverter.getCalciteAggFn("count", false,
    ImmutableList.of(intType), longType);
  // TODO: Using 0 might be wrong; might need to walk down to find the
  // proper index of a dummy.
  List<Integer> argList = ImmutableList.of(0);
  AggregateCall dummyCall = new AggregateCall(countFn, false, argList, longType, null);
  Aggregate newAggRel = oldAggRel.copy(oldAggRel.getTraitSet(), oldAggRel.getInput(),
    oldAggRel.indicator, oldAggRel.getGroupSet(), oldAggRel.getGroupSets(),
    ImmutableList.of(dummyCall));
  RelNode select = introduceDerivedTable(newAggRel);
  parent.replaceInput(0, select);
 }
}

代码示例来源：origin: apache/hive

@Test
public void testIsNotNull() {
 // @formatter:off
 final RelNode basePlan = builder
    .scan("t")
    .filter(
      builder.call(SqlStdOperatorTable.IS_NOT_NULL,
         builder.field("_str")
         )
      )
    .build();
 // @formatter:on
 statObj.setNumNulls(0);
 planner.setRoot(basePlan);
 System.out.println(RelOptUtil.toString(basePlan));
 RelNode optimizedRelNode = planner.findBestExp();
 System.out.println(RelOptUtil.toString(optimizedRelNode));
 assertEquals("missing literal", SqlKind.LITERAL, optimizedRelNode.getChildExps().get(0).getKind());
 RexLiteral val = (RexLiteral) optimizedRelNode.getChildExps().get(0);
 assertEquals(true, val.getValue());
}

代码示例来源：origin: apache/hive

@Test
public void testIsNull_zero() {
 // @formatter:off
 final RelNode basePlan = builder
    .scan("t")
    .filter(
      builder.call(SqlStdOperatorTable.IS_NULL,
         builder.field("_str")
         )
      )
    .build();
 // @formatter:on
 statObj.setNumNulls(0);
 planner.setRoot(basePlan);
 System.out.println(RelOptUtil.toString(basePlan));
 RelNode optimizedRelNode = planner.findBestExp();
 System.out.println(RelOptUtil.toString(optimizedRelNode));
 assertEquals("missing literal", SqlKind.LITERAL, optimizedRelNode.getChildExps().get(0).getKind());
 RexLiteral val = (RexLiteral) optimizedRelNode.getChildExps().get(0);
 assertEquals(false, val.getValue());
}

代码示例来源：origin: apache/hive

final List<RexNode> newProjects = new ArrayList<>();
final List<RexNode> inputExprs = input.getChildExps();
if (inputExprs == null || inputExprs.isEmpty()) {
 return aggregate;

代码示例来源：origin: apache/hive

@Test
public void testGreaterThan_Below() {
 // @formatter:off
 final RelNode basePlan = builder
    .scan("t")
    .filter(
      builder.call(SqlStdOperatorTable.GREATER_THAN,
         builder.field("_int"), builder.literal(0)
         )
      )
    .build();
 // @formatter:on
 statObj.setRange(100, 200);
 planner.setRoot(basePlan);
 RelNode optimizedRelNode = planner.findBestExp();
 assertEquals("missing literal", SqlKind.LITERAL, optimizedRelNode.getChildExps().get(0).getKind());
 RexLiteral val = (RexLiteral) optimizedRelNode.getChildExps().get(0);
 assertEquals(true, val.getValue());
}

代码示例来源：origin: apache/hive

@Test
public void testIsNull_all() {
 // @formatter:off
 final RelNode basePlan = builder
    .scan("t")
    .filter(
      builder.call(SqlStdOperatorTable.IS_NULL,
         builder.field("_str")
         )
      )
    .build();
 // @formatter:on
 statObj.setNumNulls(3);
 planner.setRoot(basePlan);
 System.out.println(RelOptUtil.toString(basePlan));
 RelNode optimizedRelNode = planner.findBestExp();
 System.out.println(RelOptUtil.toString(optimizedRelNode));
 assertEquals("missing literal", SqlKind.LITERAL, optimizedRelNode.getChildExps().get(0).getKind());
 RexLiteral val = (RexLiteral) optimizedRelNode.getChildExps().get(0);
 assertEquals(true, val.getValue());
}

代码示例来源：origin: apache/hive

@Test
public void testIsNull_one() {
 // @formatter:off
 final RelNode basePlan = builder
    .scan("t")
    .filter(
      builder.call(SqlStdOperatorTable.IS_NULL,
         builder.field("_str")
         )
      )
    .build();
 // @formatter:on
 statObj.setNumNulls(1);
 planner.setRoot(basePlan);
 System.out.println(RelOptUtil.toString(basePlan));
 RelNode optimizedRelNode = planner.findBestExp();
 System.out.println(RelOptUtil.toString(optimizedRelNode));
 assertNotEquals("should not be a literal", SqlKind.LITERAL, optimizedRelNode.getChildExps().get(0).getKind());
}

代码示例来源：origin: apache/drill

private static void replaceEmptyGroupAggr(final RelNode rel, RelNode parent) {
  // If this function is called, the parent should only include constant
  List<RexNode> exps = parent.getChildExps();
  for (RexNode rexNode : exps) {
   if (!rexNode.accept(new HiveCalciteUtil.ConstantFinder())) {
    throw new RuntimeException("We expect " + parent.toString()
      + " to contain only constants. However, " + rexNode.toString() + " is "
      + rexNode.getKind());
   }
  }
  HiveAggregate oldAggRel = (HiveAggregate) rel;
  RelDataTypeFactory typeFactory = oldAggRel.getCluster().getTypeFactory();
  RelDataType longType = TypeConverter.convert(TypeInfoFactory.longTypeInfo, typeFactory);
  RelDataType intType = TypeConverter.convert(TypeInfoFactory.intTypeInfo, typeFactory);
  // Create the dummy aggregation.
  SqlAggFunction countFn = SqlFunctionConverter.getCalciteAggFn("count", false,
    ImmutableList.of(intType), longType);
  // TODO: Using 0 might be wrong; might need to walk down to find the
  // proper index of a dummy.
  List<Integer> argList = ImmutableList.of(0);
  AggregateCall dummyCall = new AggregateCall(countFn, false, argList, longType, null);
  Aggregate newAggRel = oldAggRel.copy(oldAggRel.getTraitSet(), oldAggRel.getInput(),
    oldAggRel.indicator, oldAggRel.getGroupSet(), oldAggRel.getGroupSets(),
    ImmutableList.of(dummyCall));
  RelNode select = introduceDerivedTable(newAggRel);
  parent.replaceInput(0, select);
 }
}

代码示例来源：origin: apache/drill

if (select.getChildExps().isEmpty()) {
 RexLiteral r = select.getCluster().getRexBuilder().makeExactLiteral(new BigDecimal(1));
 ASTNode selectExpr = ASTBuilder.selectExpr(ASTBuilder.literal(r), "1");
 int i = 0;
 for (RexNode r : select.getChildExps()) {
  if (RexUtil.isNull(r) && r.getType().getSqlTypeName() != SqlTypeName.NULL) {

代码示例来源：origin: dremio/dremio-oss

@Override
public RelWriter done(RelNode node) {
 int i = 0;
 if (values.size() > 0 && values.get(0).left.equals("subset")) {
  ++i;
 }
 for (RelNode input : node.getInputs()) {
  assert values.get(i).right == input;
  ++i;
 }
 for (RexNode expr : node.getChildExps()) {
  assert values.get(i).right == expr;
  ++i;
 }
 final List<Pair<String, Object>> valuesCopy =
   ImmutableList.copyOf(values);
 values.clear();
 explain_(node, valuesCopy);
 pw.flush();
 return this;
}

代码示例来源：origin: org.apache.drill.exec/drill-java-exec

public static boolean isProjectFlatten(RelNode project) {
 assert project instanceof Project : "Rel is NOT an instance of project!";
 for (RexNode rex : project.getChildExps()) {
  RexNode newExpr = rex;
  if (rex instanceof RexCall) {
   RexCall function = (RexCall) rex;
   String functionName = function.getOperator().getName();
   if (functionName.equalsIgnoreCase("flatten") ) {
    return true;
   }
  }
 }
 return false;
}

代码示例来源：origin: org.apache.drill.exec/drill-java-exec

@SuppressWarnings("deprecation")
public RelWriter done(RelNode node) {
 int i = 0;
 if (values.size() > 0 && values.get(0).left.equals("subset")) {
  ++i;
 }
 for (RelNode input : node.getInputs()) {
  assert values.get(i).right == input;
  ++i;
 }
 for (RexNode expr : node.getChildExps()) {
  assert values.get(i).right == expr;
  ++i;
 }
 final List<Pair<String, Object>> valuesCopy =
   ImmutableList.copyOf(values);
 values.clear();
 explain_(node, valuesCopy);
 pw.flush();
 return this;
}

代码示例来源：origin: dremio/dremio-oss

@Override
 public RelWriter explainTerms(RelWriter pw) {
  super.explainTerms(pw);
  List<RexNode> childExprs = input.getChildExps();

  List<String> convertFields = Lists.transform(conversions, new Function<ConversionColumn, String>() {
   @Override
   public String apply(ConversionColumn input) {
    return input.getInputField();
   }
  });

  for (Ord<RelDataTypeField> field : Ord.zip(rowType.getFieldList())) {
   String fieldName = field.e.getName();
   if (fieldName == null) {
    fieldName = "field#" + field.i;
   }
   if (convertFields.contains(fieldName)) {
    pw.item(fieldName, "CONVERT(" + fieldName + ")");
   } else {
    pw.item(fieldName, childExprs.get(field.i));
   }
  }

  pw.item("conversions", conversions);

  return pw;
 }
}

代码示例来源：origin: org.apache.drill.exec/drill-java-exec

List<RexNode> childExprs = curRel.getChildExps();
if (childExprs != null && childExprs.size() > 0) {
 if (childExprs.get(curIndex) instanceof RexInputRef) {

代码示例来源：origin: com.facebook.presto.hive/hive-apache

private static void replaceEmptyGroupAggr(final RelNode rel, RelNode parent) {
  // If this function is called, the parent should only include constant
  List<RexNode> exps = parent.getChildExps();
  for (RexNode rexNode : exps) {
   if (!rexNode.accept(new HiveCalciteUtil.ConstantFinder())) {
    throw new RuntimeException("We expect " + parent.toString()
      + " to contain only constants. However, " + rexNode.toString() + " is "
      + rexNode.getKind());
   }
  }
  HiveAggregate oldAggRel = (HiveAggregate) rel;
  RelDataTypeFactory typeFactory = oldAggRel.getCluster().getTypeFactory();
  RelDataType longType = TypeConverter.convert(TypeInfoFactory.longTypeInfo, typeFactory);
  RelDataType intType = TypeConverter.convert(TypeInfoFactory.intTypeInfo, typeFactory);
  // Create the dummy aggregation.
  SqlAggFunction countFn = SqlFunctionConverter.getCalciteAggFn("count",
    ImmutableList.of(intType), longType);
  // TODO: Using 0 might be wrong; might need to walk down to find the
  // proper index of a dummy.
  List<Integer> argList = ImmutableList.of(0);
  AggregateCall dummyCall = new AggregateCall(countFn, false, argList, longType, null);
  Aggregate newAggRel = oldAggRel.copy(oldAggRel.getTraitSet(), oldAggRel.getInput(),
    oldAggRel.indicator, oldAggRel.getGroupSet(), oldAggRel.getGroupSets(),
    ImmutableList.of(dummyCall));
  RelNode select = introduceDerivedTable(newAggRel);
  parent.replaceInput(0, select);
 }
}