org.apache.calcite.adapter.splunk.SplunkPushDownRule Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of calcite-splunk Show documentation
Show all versions of calcite-splunk Show documentation
Splunk adapter for Calcite; also a JDBC driver for Splunk
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to you under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.calcite.adapter.splunk;
import org.apache.calcite.adapter.splunk.util.StringUtils;
import org.apache.calcite.plan.RelOptCluster;
import org.apache.calcite.plan.RelOptRule;
import org.apache.calcite.plan.RelOptRuleCall;
import org.apache.calcite.plan.RelOptRuleOperand;
import org.apache.calcite.rel.RelNode;
import org.apache.calcite.rel.logical.LogicalFilter;
import org.apache.calcite.rel.logical.LogicalProject;
import org.apache.calcite.rel.type.RelDataType;
import org.apache.calcite.rel.type.RelDataTypeField;
import org.apache.calcite.rex.RexCall;
import org.apache.calcite.rex.RexInputRef;
import org.apache.calcite.rex.RexLiteral;
import org.apache.calcite.rex.RexNode;
import org.apache.calcite.rex.RexSlot;
import org.apache.calcite.sql.SqlBinaryOperator;
import org.apache.calcite.sql.SqlKind;
import org.apache.calcite.sql.SqlOperator;
import org.apache.calcite.sql.fun.SqlStdOperatorTable;
import org.apache.calcite.sql.type.SqlTypeName;
import org.apache.calcite.util.NlsString;
import org.apache.calcite.util.Pair;
import com.google.common.collect.ImmutableSet;
import org.slf4j.Logger;
import java.util.ArrayList;
import java.util.LinkedList;
import java.util.List;
import java.util.Set;
/**
* Planner rule to push filters and projections to Splunk.
*/
public class SplunkPushDownRule
extends RelOptRule {
private static final Logger LOGGER =
StringUtils.getClassTracer(SplunkPushDownRule.class);
private static final Set SUPPORTED_OPS =
ImmutableSet.of(
SqlKind.CAST,
SqlKind.EQUALS,
SqlKind.LESS_THAN,
SqlKind.LESS_THAN_OR_EQUAL,
SqlKind.GREATER_THAN,
SqlKind.GREATER_THAN_OR_EQUAL,
SqlKind.NOT_EQUALS,
SqlKind.LIKE,
SqlKind.AND,
SqlKind.OR,
SqlKind.NOT);
public static final SplunkPushDownRule PROJECT_ON_FILTER =
new SplunkPushDownRule(
operand(
LogicalProject.class,
operand(
LogicalFilter.class,
operand(
LogicalProject.class,
operand(SplunkTableScan.class, none())))),
"proj on filter on proj");
public static final SplunkPushDownRule FILTER_ON_PROJECT =
new SplunkPushDownRule(
operand(
LogicalFilter.class,
operand(
LogicalProject.class,
operand(SplunkTableScan.class, none()))),
"filter on proj");
public static final SplunkPushDownRule FILTER =
new SplunkPushDownRule(
operand(
LogicalFilter.class, operand(SplunkTableScan.class, none())),
"filter");
public static final SplunkPushDownRule PROJECT =
new SplunkPushDownRule(
operand(
LogicalProject.class,
operand(SplunkTableScan.class, none())),
"proj");
/** Creates a SplunkPushDownRule. */
protected SplunkPushDownRule(RelOptRuleOperand rule, String id) {
super(rule, "SplunkPushDownRule: " + id);
}
// ~ Methods --------------------------------------------------------------
// implement RelOptRule
public void onMatch(RelOptRuleCall call) {
LOGGER.debug(description);
int relLength = call.rels.length;
SplunkTableScan splunkRel =
(SplunkTableScan) call.rels[relLength - 1];
LogicalFilter filter;
LogicalProject topProj = null;
LogicalProject bottomProj = null;
RelDataType topRow = splunkRel.getRowType();
int filterIdx = 2;
if (call.rels[relLength - 2] instanceof LogicalProject) {
bottomProj = (LogicalProject) call.rels[relLength - 2];
filterIdx = 3;
// bottom projection will change the field count/order
topRow = bottomProj.getRowType();
}
String filterString;
if (filterIdx <= relLength
&& call.rels[relLength - filterIdx] instanceof LogicalFilter) {
filter = (LogicalFilter) call.rels[relLength - filterIdx];
int topProjIdx = filterIdx + 1;
if (topProjIdx <= relLength
&& call.rels[relLength - topProjIdx] instanceof LogicalProject) {
topProj = (LogicalProject) call.rels[relLength - topProjIdx];
}
RexCall filterCall = (RexCall) filter.getCondition();
SqlOperator op = filterCall.getOperator();
List operands = filterCall.getOperands();
LOGGER.debug("fieldNames: {}", getFieldsString(topRow));
final StringBuilder buf = new StringBuilder();
if (getFilter(op, operands, buf, topRow.getFieldNames())) {
filterString = buf.toString();
} else {
return; // can't handle
}
} else {
filterString = "";
}
// top projection will change the field count/order
if (topProj != null) {
topRow = topProj.getRowType();
}
LOGGER.debug("pre transformTo fieldNames: {}", getFieldsString(topRow));
call.transformTo(
appendSearchString(
filterString, splunkRel, topProj, bottomProj,
topRow, null));
}
/**
* Appends a search string.
*
* @param toAppend Search string to append
* @param splunkRel Relational expression
* @param topProj Top projection
* @param bottomProj Bottom projection
*/
protected RelNode appendSearchString(
String toAppend,
SplunkTableScan splunkRel,
LogicalProject topProj,
LogicalProject bottomProj,
RelDataType topRow,
RelDataType bottomRow) {
final RelOptCluster cluster = splunkRel.getCluster();
StringBuilder updateSearchStr = new StringBuilder(splunkRel.search);
if (!toAppend.isEmpty()) {
updateSearchStr.append(" ").append(toAppend);
}
List bottomFields =
bottomRow == null ? null : bottomRow.getFieldList();
List topFields =
topRow == null ? null : topRow.getFieldList();
if (bottomFields == null) {
bottomFields = splunkRel.getRowType().getFieldList();
}
// handle bottom projection (ie choose a subset of the table fields)
if (bottomProj != null) {
List tmp =
new ArrayList();
List dRow = bottomProj.getRowType().getFieldList();
for (RexNode rn : bottomProj.getProjects()) {
RelDataTypeField rdtf;
if (rn instanceof RexSlot) {
RexSlot rs = (RexSlot) rn;
rdtf = bottomFields.get(rs.getIndex());
} else {
rdtf = dRow.get(tmp.size());
}
tmp.add(rdtf);
}
bottomFields = tmp;
}
// field renaming: to -> from
List> renames =
new LinkedList>();
// handle top projection (ie reordering and renaming)
List newFields = bottomFields;
if (topProj != null) {
LOGGER.debug("topProj: {}", String.valueOf(topProj.getPermutation()));
newFields = new ArrayList();
int i = 0;
for (RexNode rn : topProj.getProjects()) {
RexInputRef rif = (RexInputRef) rn;
RelDataTypeField field = bottomFields.get(rif.getIndex());
if (!bottomFields.get(rif.getIndex()).getName()
.equals(topFields.get(i).getName())) {
renames.add(
Pair.of(
bottomFields.get(rif.getIndex()).getName(),
topFields.get(i).getName()));
field = topFields.get(i);
}
newFields.add(field);
}
}
if (!renames.isEmpty()) {
updateSearchStr.append("| rename ");
for (Pair p : renames) {
updateSearchStr.append(p.left).append(" AS ")
.append(p.right).append(" ");
}
}
RelDataType resultType =
cluster.getTypeFactory().createStructType(newFields);
String searchWithFilter = updateSearchStr.toString();
RelNode rel =
new SplunkTableScan(
cluster,
splunkRel.getTable(),
splunkRel.splunkTable,
searchWithFilter,
splunkRel.earliest,
splunkRel.latest,
resultType.getFieldNames());
LOGGER.debug("end of appendSearchString fieldNames: {}",
rel.getRowType().getFieldNames());
return rel;
}
// ~ Private Methods ------------------------------------------------------
private static RelNode addProjectionRule(LogicalProject proj, RelNode rel) {
if (proj == null) {
return rel;
}
return LogicalProject.create(rel, proj.getProjects(), proj.getRowType());
}
// TODO: use StringBuilder instead of String
// TODO: refactor this to use more tree like parsing, need to also
// make sure we use parens properly - currently precedence
// rules are simply left to right
private boolean getFilter(SqlOperator op, List operands,
StringBuilder s, List fieldNames) {
if (!valid(op.getKind())) {
return false;
}
boolean like = false;
switch (op.getKind()) {
case NOT:
// NOT op pre-pended
s = s.append(" NOT ");
break;
case CAST:
return asd(false, operands, s, fieldNames, 0);
case LIKE:
like = true;
break;
}
for (int i = 0; i < operands.size(); i++) {
if (!asd(like, operands, s, fieldNames, i)) {
return false;
}
if (op instanceof SqlBinaryOperator && i == 0) {
s.append(" ").append(op).append(" ");
}
}
return true;
}
private boolean asd(boolean like, List operands, StringBuilder s,
List fieldNames, int i) {
RexNode operand = operands.get(i);
if (operand instanceof RexCall) {
s.append("(");
final RexCall call = (RexCall) operand;
boolean b =
getFilter(
call.getOperator(),
call.getOperands(),
s,
fieldNames);
if (!b) {
return false;
}
s.append(")");
} else {
if (operand instanceof RexInputRef) {
if (i != 0) {
return false;
}
int fieldIndex = ((RexInputRef) operand).getIndex();
String name = fieldNames.get(fieldIndex);
s.append(name);
} else { // RexLiteral
String tmp = toString(like, (RexLiteral) operand);
if (tmp == null) {
return false;
}
s.append(tmp);
}
}
return true;
}
private boolean valid(SqlKind kind) {
return SUPPORTED_OPS.contains(kind);
}
private String toString(SqlOperator op) {
if (op.equals(SqlStdOperatorTable.LIKE)) {
return SqlStdOperatorTable.EQUALS.toString();
} else if (op.equals(SqlStdOperatorTable.NOT_EQUALS)) {
return "!=";
}
return op.toString();
}
public static String searchEscape(String str) {
if (str.isEmpty()) {
return "\"\"";
}
StringBuilder sb = new StringBuilder(str.length());
boolean quote = false;
for (int i = 0; i < str.length(); i++) {
char c = str.charAt(i);
if (c == '"' || c == '\\') {
sb.append('\\');
}
sb.append(c);
quote |= !(Character.isLetterOrDigit(c) || c == '_');
}
if (quote || sb.length() != str.length()) {
sb.insert(0, '"');
sb.append('"');
return sb.toString();
}
return str;
}
private String toString(boolean like, RexLiteral literal) {
String value = null;
SqlTypeName litSqlType = literal.getTypeName();
if (SqlTypeName.NUMERIC_TYPES.contains(litSqlType)) {
value = literal.getValue().toString();
} else if (litSqlType.equals(SqlTypeName.CHAR)) {
value = ((NlsString) literal.getValue()).getValue();
if (like) {
value = value.replaceAll("%", "*");
}
value = searchEscape(value);
}
return value;
}
// transform the call from SplunkUdxRel to FarragoJavaUdxRel
// usually used to stop the optimizer from calling us
protected void transformToFarragoUdxRel(
RelOptRuleCall call,
SplunkTableScan splunkRel,
LogicalFilter filter,
LogicalProject topProj,
LogicalProject bottomProj) {
assert false;
/*
RelNode rel =
new EnumerableTableScan(
udxRel.getCluster(),
udxRel.getTable(),
udxRel.getRowType(),
udxRel.getServerMofId());
rel = RelOptUtil.createCastRel(rel, udxRel.getRowType(), true);
rel = addProjectionRule(bottomProj, rel);
if (filter != null) {
rel =
new LogicalFilter(filter.getCluster(), rel, filter.getCondition());
}
rel = addProjectionRule(topProj, rel);
call.transformTo(rel);
*/
}
public static String getFieldsString(RelDataType row) {
return row.getFieldNames().toString();
}
}
// End SplunkPushDownRule.java