/*
Derby - Class org.apache.derby.impl.sql.compile.FromList
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.derby.impl.sql.compile;
import org.apache.derby.iapi.services.sanity.SanityManager;
import org.apache.derby.iapi.sql.compile.CompilerContext;
import org.apache.derby.iapi.sql.compile.Optimizable;
import org.apache.derby.iapi.sql.compile.OptimizableList;
import org.apache.derby.iapi.sql.compile.Optimizer;
import org.apache.derby.iapi.sql.compile.Visitable;
import org.apache.derby.iapi.sql.compile.Visitor;
import org.apache.derby.iapi.sql.compile.C_NodeTypes;
import org.apache.derby.iapi.sql.dictionary.DataDictionary;
import org.apache.derby.iapi.error.StandardException;
import org.apache.derby.iapi.reference.SQLState;
import org.apache.derby.iapi.util.JBitSet;
import org.apache.derby.iapi.util.StringUtil;
import java.util.Properties;
import java.util.Enumeration;
import java.util.Vector;
/**
* A FromList represents the list of tables in a FROM clause in a DML
* statement. It extends QueryTreeNodeVector.
*
*/
public class FromList extends QueryTreeNodeVector implements OptimizableList
{
Properties properties;
// RESOLVE: The default should be false
boolean fixedJoinOrder = true;
// true by default.
boolean useStatistics = true;
// FromList could have a view in it's list. If the view is defined in SESSION
// schema, then we do not want to cache the statement's plan. This boolean
// will help keep track of such a condition.
private boolean referencesSessionSchema;
/* Whether or not this FromList is transparent. A "transparent" FromList
* is one in which all FromTables are bound based on an outer query's
* FromList. This means that the FromTables in the transparent list are
* allowed to see and reference FromTables in the outer query's list.
* Or put differently, a FromTable which sits in a transparent FromList
* does not "see" the transparent FromList when binding; rather, it sees
* (and can therefore reference) the FromList of an outer query.
*/
private boolean isTransparent;
/**
* Window definitions used for resolving window functions not containing
* in-line window specifications, but referring window definitions
*/
private WindowList windows;
/** Initializer for a FromList */
public void init(Object optimizeJoinOrder)
{
fixedJoinOrder = ! (((Boolean) optimizeJoinOrder).booleanValue());
isTransparent = false;
}
/**
* Initializer for a FromList
*
* @exception StandardException Thrown on error
*/
public void init(Object optimizeJoinOrder, Object fromTable)
throws StandardException
{
init(optimizeJoinOrder);
addFromTable((FromTable) fromTable);
}
/*
* OptimizableList interface
*/
/**
* @see org.apache.derby.iapi.sql.compile.OptimizableList#getOptimizable
*/
public Optimizable getOptimizable(int index)
{
return (Optimizable) elementAt(index);
}
/**
* @see org.apache.derby.iapi.sql.compile.OptimizableList#setOptimizable
*/
public void setOptimizable(int index, Optimizable optimizable)
{
setElementAt((FromTable) optimizable, index);
}
/**
* @see OptimizableList#verifyProperties
* @exception StandardException Thrown on error
*/
public void verifyProperties(DataDictionary dDictionary) throws StandardException
{
int size = size();
for (int index = 0; index < size; index++)
{
((Optimizable) elementAt(index)).verifyProperties(dDictionary);
}
}
/**
* Add a table to the FROM list.
*
* @param fromTable A FromTable to add to the list
*
* @exception StandardException Thrown on error
*/
public void addFromTable(FromTable fromTable) throws StandardException
{
/* Don't worry about checking TableOperatorNodes since
* they don't have exposed names. This will potentially
* allow duplicate exposed names in some degenerate cases,
* but the binding of the ColumnReferences will catch those
* cases with a different error. If the query does not have
* any ColumnReferences from the duplicate exposed name, the
* user is executing a really dumb query and we won't throw
* and exception - consider it an ANSI extension.
*/
TableName leftTable = null;
TableName rightTable = null;
if (! (fromTable instanceof TableOperatorNode))
{
/* Check for duplicate table name in FROM list */
int size = size();
for (int index = 0; index < size; index++)
{
leftTable = fromTable.getTableName();
if(((FromTable) elementAt(index)) instanceof TableOperatorNode) {
continue;
}
else {
rightTable = ((FromTable) elementAt(index)).getTableName();
}
if(leftTable.equals(rightTable))
{
throw StandardException.newException(SQLState.LANG_FROM_LIST_DUPLICATE_TABLE_NAME, fromTable.getExposedName());
}
}
}
addElement(fromTable);
}
/**
* Search to see if a query references the specifed table name.
*
* @param name Table name (String) to search for.
* @param baseTable Whether or not name is for a base table
*
* @return true if found, else false
*
* @exception StandardException Thrown on error
*/
public boolean referencesTarget(String name, boolean baseTable)
throws StandardException
{
FromTable fromTable;
boolean found = false;
/* Check for table or VTI name in FROM list */
int size = size();
for (int index = 0; index < size; index++)
{
fromTable = (FromTable) elementAt(index);
if (fromTable.referencesTarget(name, baseTable))
{
found = true;
break;
}
}
return found;
}
/**
* Return true if the node references SESSION schema tables (temporary or permanent)
*
* @return true if references SESSION schema tables, else false
*
* @exception StandardException Thrown on error
*/
public boolean referencesSessionSchema()
throws StandardException
{
FromTable fromTable;
boolean found = false;
// Following if will return true if this FromList object had any VIEWs
// from SESSION schema as elements. This information is gathered during
// the bindTables method. At the end of the bindTables, we loose
// the information on VIEWs since they get replaced with their view
// definition. Hence, we need to intercept in the middle on the bindTables
// method and save that information in referencesSeesionSchema field.
if (referencesSessionSchema) return true;
/* Check for table or VTI name in FROM list */
int size = size();
for (int index = 0; index < size; index++)
{
fromTable = (FromTable) elementAt(index);
if (fromTable.referencesSessionSchema())
{
found = true;
break;
}
}
return found;
}
/**
* Determine whether or not the specified name is an exposed name in
* the current query block.
*
* @param name The specified name to search for as an exposed name.
* @param schemaName Schema name, if non-null.
* @param exactMatch Whether or not we need an exact match on specified schema and table
* names or match on table id.
*
* @return The FromTable, if any, with the exposed name.
*
* @exception StandardException Thrown on error
*/
protected FromTable getFromTableByName(String name, String schemaName, boolean exactMatch)
throws StandardException
{
boolean found = false;
FromTable fromTable;
FromTable result = null;
int size = size();
for (int index = 0; index < size; index++)
{
fromTable = (FromTable) elementAt(index);
result = fromTable.getFromTableByName(name, schemaName, exactMatch);
if (result != null)
{
return result;
}
}
return result;
}
/**
* Bind the tables in this FromList. This includes looking them up in
* the DataDictionary, getting their TableDescriptors and assigning the
* table numbers.
*
* @param dataDictionary The DataDictionary to use for binding
* @param fromListParam FromList to use/append to.
*
* @exception StandardException Thrown on error
*/
public void bindTables(DataDictionary dataDictionary,
FromList fromListParam)
throws StandardException
{
FromTable fromTable;
/* Now we bind the tables - this is a 2 step process.
* We first bind all of the non-VTIs, then we bind the VTIs.
* This enables us to handle the passing of correlation
* columns in VTI parameters.
* NOTE: We set the table numbers for all of the VTIs in the
* first step, when we find them, in order to avoid an ordering
* problem with join columns in parameters.
*/
int size = size();
for (int index = 0; index < size; index++)
{
fromTable = (FromTable) elementAt(index);
ResultSetNode newNode = fromTable.bindNonVTITables(dataDictionary, fromListParam);
// If the fromTable is a view in the SESSION schema, then we need to save that information
// in referencesSessionSchema element. The reason for this is that the view will get
// replaced by it's view definition and we will loose the information that the statement
// was referencing a SESSION schema object.
if (fromTable.referencesSessionSchema())
referencesSessionSchema = true;
setElementAt(newNode, index);
}
for (int index = 0; index < size; index++)
{
fromTable = (FromTable) elementAt(index);
ResultSetNode newNode = fromTable.bindVTITables(fromListParam);
if (fromTable.referencesSessionSchema())
referencesSessionSchema = true;
setElementAt(newNode, index);
}
}
/**
* Bind the expressions in this FromList. This means
* binding the sub-expressions, as well as figuring out what the return
* type is for each expression.
*
* @exception StandardException Thrown on error
*/
public void bindExpressions( FromList fromListParam )
throws StandardException
{
FromTable fromTable;
int size = size();
for (int index = 0; index < size; index++)
{
fromTable = (FromTable) elementAt(index);
/* If this FromList is transparent then its FromTables should
* be bound based on the outer query's FROM list.
*/
fromTable.bindExpressions(
isTransparent ? fromListParam : this);
}
}
/**
* Bind the result columns of the ResultSetNodes in this FromList when there is no
* base table to bind them to. This is useful for SELECT statements,
* where the result columns get their types from the expressions that
* live under them.
*
* @param fromListParam FromList to use/append to.
*
* @exception StandardException Thrown on error
*/
public void bindResultColumns(FromList fromListParam)
throws StandardException
{
FromTable fromTable;
int origList = fromListParam.size();
int size = size();
for (int index = 0; index < size; index++)
{
fromTable = (FromTable) elementAt(index);
if (fromTable.needsSpecialRCLBinding())
fromTable.bindResultColumns(fromListParam);
fromListParam.insertElementAt(fromTable, 0);
}
/* Remove all references added here */
while (fromListParam.size() > origList)
fromListParam.removeElementAt(0);
}
/**
* Expand a "*" into the appropriate ResultColumnList. If the "*"
* is unqualified it will expand into a list of all columns in all
* of the base tables in the from list at the current nesting level;
* otherwise it will expand into a list of all of the columns in the
* base table that matches the qualification.
*
* NOTE: Callers are responsible for ordering the FromList by nesting
* level, with tables at the deepest (current) nesting level first.
* We will expand the "*" into a list of all columns from all tables
* having the same nesting level as the first FromTable in this list.
* The check for nesting level exists because it's possible that this
* FromList includes FromTables from an outer query, which can happen
* if there is a "transparent" FromList above this one in the query
* tree. Ex:
*
* select j from onerow where exists
* (select 1 from somerow
* union select * from diffrow where onerow.j < diffrow.k)
*
* If "this" is the FromList for the right child of the UNION then it will
* contain both "diffrow" and "onerow", the latter of which was passed
* down via a transparent FromList (to allow binding of the WHERE clause).
* In that case the "*" should only expand the result columns of "diffrow";
* it should not expand the result columns of "onerow" because that table
* is from an outer query. We can achieve this selective expansion by
* looking at nesting levels.
*
* @param allTableName The qualification on the "*" as a String.
*
* @return ResultColumnList representing expansion
*
* @exception StandardException Thrown on error
*/
public ResultColumnList expandAll(TableName allTableName)
throws StandardException
{
ResultColumnList resultColumnList = null;
ResultColumnList tempRCList = null;
boolean matchfound = false;
FromTable fromTable;
/* Expand the "*" for the table that matches, if it is qualified
* (allTableName is not null) or for all tables in the list at the
* current nesting level if the "*" is not qualified (allTableName
* is null). Current nesting level is determined by the nesting
* level of the first FromTable in the list.
*/
int targetNestingLevel = ((FromTable)elementAt(0)).getLevel();
int size = size();
/* Make sure our assumption about nesting-based ordering
* has been satisified. I.e. that the list is ordered
* with the most deeply nested FromTables first.
*/
if (SanityManager.DEBUG)
{
int prevNL = targetNestingLevel;
for (int i = 1; i < size; i++)
{
int currNL = ((FromTable)elementAt(i)).getLevel();
SanityManager.ASSERT((prevNL >= currNL),
"FROM list should have been ordered by nesting " +
"level (deepest level first), but it was not.");
prevNL = currNL;
}
}
for (int index = 0; index < size; index++)
{
fromTable = (FromTable) elementAt(index);
if (targetNestingLevel != fromTable.getLevel())
{
/* We only expand result columns for tables at the
* target nesting level. Since the FromTables are
* sorted based on nesting level, we're done if we
* get here.
*/
break;
}
/* We let the FromTable decide if there is a match on
* the exposed name. (A JoinNode will not have an
* exposed name, so it will need to pass the info to its
* left and right children.)
*/
tempRCList = fromTable.getAllResultColumns(allTableName);
if (tempRCList == null)
{
continue;
}
/* Expand the column list and append to the list that
* we will return.
*/
if (resultColumnList == null)
{
resultColumnList = tempRCList;
}
else
{
resultColumnList.nondestructiveAppend(tempRCList);
}
/* If the "*" is qualified, then we can stop the
* expansion as soon as we find the matching table.
*/
if (allTableName != null)
{
matchfound = true;
}
}
/* Give an error if the qualification name did not match
* an exposed name
*/
if (resultColumnList == null)
{
throw StandardException.newException(SQLState.LANG_EXPOSED_NAME_NOT_FOUND, allTableName);
}
return resultColumnList;
}
/**
* Bind a column reference to one of the tables in this FromList. The column name
* must be unique within the tables in the FromList. An exception is thrown
* if a column name is not unique.
*
* NOTE: Callers are responsible for ordering the FromList by nesting level,
* with tables at the deepest (current) nesting level first. We will try to
* match against all FromTables at a given nesting level. If no match is
* found at a nesting level, then we proceed to the next level. We stop
* walking the list when the nesting level changes and we have found a match.
*
* NOTE: If the ColumnReference is qualified, then we will stop the search
* at the first nesting level where there is a match on the exposed table name.
* For example, s (a, b, c), t (d, e, f)
* select * from s where exists (select * from t s where s.c = a)
* will not find a match for s.c, which is the expected ANSI behavior.
*
* bindTables() must have already been called on this FromList before
* calling this method.
*
* @param columnReference The ColumnReference describing the column to bind
*
* @return ResultColumn The matching ResultColumn
*
* @exception StandardException Thrown on error
*/
public ResultColumn bindColumnReference(ColumnReference columnReference)
throws StandardException
{
boolean columnNameMatch = false;
boolean tableNameMatch = false;
FromTable fromTable;
int currentLevel = -1;
int previousLevel = -1;
ResultColumn matchingRC = null;
ResultColumn resultColumn;
String crTableName = columnReference.getTableName();
/*
** Find the first table with matching column name. If there
** is more than one table with a matching column name at the same
** nesting level, give an error.
*/
int size = size();
for (int index = 0; index < size; index++)
{
fromTable = (FromTable) elementAt(index);
/* We can stop if we've found a matching column or table name
* at the previous nesting level.
*/
currentLevel = fromTable.getLevel();
if (previousLevel != currentLevel)
{
if (columnNameMatch)
{
break;
}
if (tableNameMatch)
{
break;
}
}
/* Simpler to always set previousLevel then to test and set */
previousLevel = currentLevel;
resultColumn = fromTable.getMatchingColumn(columnReference);
if (resultColumn != null)
{
if (! columnNameMatch)
{
/* TableNumbers and column numbers are set in the CR in the
* underlying FromTable. This ensures that they get the
* table number/column number from the underlying table,
* not the join node. This is important for beging able to
* push predicates down through join nodes.
*/
matchingRC = resultColumn;
columnReference.setSource(resultColumn);
/* Set the nesting level at which the CR appears and the nesting level
* of its source RC.
*/
columnReference.setNestingLevel(((FromTable) elementAt(0)).getLevel());
columnReference.setSourceLevel(currentLevel);
columnNameMatch = true;
if (fromTable.isPrivilegeCollectionRequired())
getCompilerContext().addRequiredColumnPriv( resultColumn.getTableColumnDescriptor());
}
else
{
throw StandardException.newException(SQLState.LANG_AMBIGUOUS_COLUMN_NAME,
columnReference.getSQLColumnName());
}
}
/* Remember if we get a match on the exposed table name, so that
* we can stop at the beginning of the next level.
*/
tableNameMatch = tableNameMatch ||
(crTableName != null &&
crTableName.equals(fromTable.getExposedName()) );
}
return matchingRC;
}
/**
* Check for (and reject) all ? parameters directly under the ResultColumns.
* This is done for SELECT statements.
*
* @exception StandardException Thrown if a ? parameter found
* directly under a ResultColumn
*/
public void rejectParameters() throws StandardException
{
FromTable fromTable;
int size = size();
for (int index = 0; index < size; index++)
{
fromTable = (FromTable) elementAt(index);
fromTable.rejectParameters();
}
}
// This method reorders LOJs in the FROM clause.
// For now, we process only a LOJ. For example, "... from LOJ_1, LOJ2 ..."
// will not be processed.
public boolean LOJ_reorderable(int numTables) throws StandardException
{
boolean anyChange = false;
if (size() > 1) return anyChange;
FromTable ft = (FromTable) elementAt(0);
anyChange = ft.LOJ_reorderable(numTables);
return anyChange;
}
/**
* Preprocess the query tree - this currently means:
* o Generating a referenced table map for each ResultSetNode.
* o Putting the WHERE and HAVING clauses in conjunctive normal form (CNF).
* o Converting the WHERE and HAVING clauses into PredicateLists and
* classifying them.
* o Flatten those FromSubqueries which can be flattened.
* o Ensuring that a ProjectRestrictNode is generated on top of every
* FromBaseTable and generated in place of every FromSubquery which
* could not be flattened.
* o Pushing single table predicates down to the new ProjectRestrictNodes.
*
* @param numTables The number of tables in the DML Statement
* @param gbl The group by list, if any
*
* @exception StandardException Thrown on error
*/
public void preprocess(int numTables,
GroupByList gbl,
ValueNode predicateTree)
throws StandardException
{
int size = size();
/* Preprocess each FromTable in the list */
for (int index = 0; index < size; index++)
{
FromTable ft = (FromTable) elementAt(index);
/* Transform any outer joins to inner joins where appropriate */
ft = ft.transformOuterJoins(predicateTree, numTables);
/* Preprocess this FromTable */
setElementAt(ft.preprocess(numTables, gbl, this), index);
}
}
/**
* Flatten all the FromTables that are flattenable.
* RESOLVE - right now we just flatten FromSubqueries. We
* should also flatten flattenable JoinNodes here.
*
* @param rcl The RCL from the outer query
* @param predicateList The PredicateList from the outer query
* @param sql The SubqueryList from the outer query
* @param gbl The group by list, if any
* @param havingClause The HAVING clause, if any
*
* @exception StandardException Thrown on error
*/
public void flattenFromTables(ResultColumnList rcl,
PredicateList predicateList,
SubqueryList sql,
GroupByList gbl,
ValueNode havingClause)
throws StandardException
{
boolean flattened = true;
Vector flattenedTableNumbers = new Vector();
if (SanityManager.DEBUG)
{
SanityManager.ASSERT(rcl != null,
"rcl is expected to be non-null");
SanityManager.ASSERT(predicateList != null,
"predicateList is expected to be non-null");
SanityManager.ASSERT(sql != null,
"sql is expected to be non-null");
}
/* Loop until all flattenable entries are flattened.
* We restart the inner loop after flattening an in place
* to simplify the logic and so that we don't have to worry
* about walking a list while we are modifying it.
*/
while (flattened)
{
flattened = false;
for (int index = 0; index < size() && ! flattened; index++)
{
FromTable ft = (FromTable) elementAt(index);
/* Flatten FromSubquerys and flattenable JoinNodes */
if ((ft instanceof FromSubquery) ||
ft.isFlattenableJoinNode())
{
//save the table number of the node to be flattened
flattenedTableNumbers.add(new Integer(ft.getTableNumber()));
/* Remove the node from the list and insert its
* FromList here.
*/
FromList flatteningFL = ft.flatten(
rcl,
predicateList,
sql,
gbl,
havingClause);
if (SanityManager.DEBUG)
{
SanityManager.ASSERT(flatteningFL == null ||
flatteningFL.size() > 0,
"flatteningFL expected to be null or size > 0");
}
if (flatteningFL != null)
{
setElementAt(flatteningFL.elementAt(0), index);
int innerSize = flatteningFL.size();
for (int inner = 1; inner < innerSize; inner++)
{
insertElementAt(flatteningFL.elementAt(inner), index + inner);
}
}
else
{
/*
** If flatten returns null, that means it wants to
** be removed from the FromList.
*/
removeElementAt(index);
}
flattened = true;
}
}
}
/* fix up dependency maps for exists base tables since they might have a
* dependency on this join node
*/
if (flattenedTableNumbers.size() > 0)
{
for (int i = 0; i < size(); i++)
{
FromTable ft = (FromTable) elementAt(i);
if (ft instanceof ProjectRestrictNode)
{
ResultSetNode rst = ((ProjectRestrictNode)ft).getChildResult();
if (rst instanceof FromBaseTable)
{
((FromBaseTable)rst).clearDependency(flattenedTableNumbers);
}
}
}
}
}
/**
* Categorize and push the predicates that are pushable.
*
* @param predicateList The query's PredicateList
*
* @exception StandardException Thrown on error
*/
void pushPredicates(PredicateList predicateList)
throws StandardException
{
if (SanityManager.DEBUG)
{
SanityManager.ASSERT(predicateList != null,
"predicateList is expected to be non-null");
}
/* We can finally categorize each Predicate and try to push them down.
* NOTE: The PredicateList may be empty, but that's okay, we still
* call pushExpressions() for each entry in the FromList because that's
* where any outer join conditions will get pushed down.
*/
predicateList.categorize();
int size = size();
for (int index = 0; index < size; index++)
{
FromTable fromTable = (FromTable) elementAt(index);
fromTable.pushExpressions(predicateList);
}
}
/**
* Set the (query block) level (0-based) for the FromTables in this
* FromList.
*
* @param level The query block level for this table.
*/
public void setLevel(int level)
{
int size = size();
for (int index = 0; index < size; index++)
{
FromTable fromTable = (FromTable) elementAt(index);
fromTable.setLevel(level);
}
}
/**
* Get the FromTable from this list which has the specified ResultColumn in
* its RCL.
*
* @param rc The ResultColumn match on.
*
* @return FromTable The matching FromTable.
*/
public FromTable getFromTableByResultColumn(ResultColumn rc)
{
FromTable fromTable = null;
int size = size();
for (int index = 0; index < size; index++)
{
fromTable = (FromTable) elementAt(index);
if (fromTable.getResultColumns().indexOf(rc) != -1)
{
break;
}
}
if (SanityManager.DEBUG)
{
SanityManager.ASSERT(fromTable != null,
"No matching FromTable found");
}
return fromTable;
}
/**
* Set the Properties list for this FromList.
*
* @exception StandardException Thrown on error
*/
public void setProperties(Properties props) throws StandardException
{
properties = props;
/*
** Validate the properties list now. This is possible because
** there is nothing in this properties list that relies on binding
** or optimization to validate.
*/
Enumeration e = properties.keys();
while (e.hasMoreElements())
{
String key = (String) e.nextElement();
String value = (String) properties.get(key);
if (key.equals("joinOrder"))
{
if (StringUtil.SQLEqualsIgnoreCase(value,"fixed"))
{
fixedJoinOrder = true;
}
else if (StringUtil.SQLEqualsIgnoreCase(value,"unfixed"))
{
fixedJoinOrder = false;
}
else
{
throw StandardException.newException(SQLState.LANG_INVALID_JOIN_ORDER_SPEC, value);
}
}
else if (key.equals("useStatistics"))
{
if (StringUtil.SQLEqualsIgnoreCase(value,"true"))
{
useStatistics = true;
}
else if (StringUtil.SQLEqualsIgnoreCase(value,"false"))
{
useStatistics = false;
}
else
{
throw StandardException.newException(SQLState.LANG_INVALID_STATISTICS_SPEC, value);
}
}
else
{
throw StandardException.newException(SQLState.LANG_INVALID_FROM_LIST_PROPERTY, key, value);
}
}
}
/** @see OptimizableList#reOrder */
public void reOrder(int[] joinOrder)
{
int posn;
if (SanityManager.DEBUG)
{
if (joinOrder.length != size())
{
SanityManager.THROWASSERT("In reOrder(), size of FromList is " + size() + " while size of joinOrder array is " + joinOrder.length);
}
/*
** Determine that the values in the list are unique and in range.
** The easiest way to determine that they are unique is to add
** them all up and see whether the result is what's expected
** for that array size.
*/
int sum = 0;
for (int i = 0; i < joinOrder.length; i++)
{
if (joinOrder[i] < 0 || joinOrder[i] > (joinOrder.length - 1))
{
SanityManager.THROWASSERT("joinOrder[" + i + "] == " +
joinOrder[i] +
" is out of range - must be between 0 and " +
(joinOrder.length - 1) +
" inclusive.");
}
sum += joinOrder[i];
}
/*
** The sum of all integers from 0 through n is (n * (n - 1)) / 2.
*/
if (sum != ( ( joinOrder.length * (joinOrder.length - 1) ) / 2) )
{
String arrayVals = "";
for (int i = 0; i < joinOrder.length; i++)
arrayVals = arrayVals + joinOrder[i] + " ";
SanityManager.THROWASSERT("joinOrder array has some duplicate value: " + arrayVals);
}
}
/* Form a list that's in the order we want */
QueryTreeNode[] orderedFL = new FromTable[joinOrder.length];
for (posn = 0; posn < joinOrder.length; posn++)
{
/*
** Get the element at the i'th join order position from the
** current list and make it the next element of orderedList.
*/
orderedFL[posn] = elementAt(joinOrder[posn]);
}
/* Now orderedList has been built, so set this list to the same order */
for (posn = 0; posn < joinOrder.length; posn++)
{
setElementAt(orderedFL[posn], posn);
}
}
/** @see OptimizableList#useStatistics */
public boolean useStatistics()
{
return useStatistics;
}
/** @see OptimizableList#optimizeJoinOrder */
public boolean optimizeJoinOrder()
{
return ! fixedJoinOrder;
}
/** @see OptimizableList#legalJoinOrder */
public boolean legalJoinOrder(int numTablesInQuery)
{
JBitSet assignedTableMap = new JBitSet(numTablesInQuery);
int size = size();
for (int index = 0; index < size; index++)
{
FromTable ft = (FromTable) elementAt(index);
assignedTableMap.or(ft.getReferencedTableMap());
if ( ! ft.legalJoinOrder(assignedTableMap))
{
return false;
}
}
return true;
}
/** @see OptimizableList#initAccessPaths */
public void initAccessPaths(Optimizer optimizer)
{
int size = size();
for (int index = 0; index < size; index++)
{
FromTable ft = (FromTable) elementAt(index);
ft.initAccessPaths(optimizer);
}
}
/**
* Bind any untyped null nodes to the types in the given ResultColumnList.
*
* @param bindingRCL The ResultColumnList with the types to bind to.
*
* @exception StandardException Thrown on error
*/
public void bindUntypedNullsToResultColumns(ResultColumnList bindingRCL)
throws StandardException
{
int size = size();
for (int index = 0; index < size; index++)
{
FromTable fromTable = (FromTable) elementAt(index);
fromTable.bindUntypedNullsToResultColumns(bindingRCL);
}
}
/**
* Decrement (query block) level (0-based) for
* all of the tables in this from list.
* This is useful when flattening a subquery.
*
* @param decrement The amount to decrement by.
*/
void decrementLevel(int decrement)
{
int size = size();
for (int index = 0; index < size; index++)
{
FromTable fromTable = (FromTable) elementAt(index);
fromTable.decrementLevel(decrement);
/* Decrement the level of any CRs in single table
* predicates that are interesting to transitive
* closure.
*/
ProjectRestrictNode prn = (ProjectRestrictNode) fromTable;
PredicateList pl = prn.getRestrictionList();
if (pl != null)
{
pl.decrementLevel(this, decrement);
}
}
}
/**
* This method is used for both subquery flattening and distinct
* elimination based on a uniqueness condition. For subquery
* flattening we want to make sure that the query block
* will return at most 1 row. For distinct elimination we
* want to make sure that the query block will not return
* any duplicates.
* This is true if every table in the from list is
* (a base table and the set of columns from the table that
* are in equality comparisons with expressions that do not include columns
* from the same table is a superset of any unique index
* on the table) or an EXISTS FBT. In addition, at least 1 of the tables
* in the list has a set of columns in equality comparisons with expressions
* that do not include column references from the same query block
* is a superset of a unique index
* on that table. (This ensures that the query block will onlyr
* return a single row.)
* This method is expected to be called after normalization and
* after the from list has been preprocessed.
* It can be called both before and after the predicates have
* been pulled from the where clause.
* The algorithm for this is as follows
*
* If any table in the query block is not a base table, give up.
* For each table in the query
* Ignore exists table since they can only produce one row
*
* create a matrix of tables and columns from the table (tableColMap)
* (this is used to keep track of the join columns and constants
* that can be used to figure out whether the rows from a join
* or in a select list are distinct based on unique indexes)
*
* create an array of columns from the table(eqOuterCol)
* (this is used to determine that only one row will be returned
* from a join)
*
* if the current table is the table for the result columns
* set the result columns in the eqOuterCol and tableColMap
* (if these columns are a superset of a unique index and
* all joining tables result in only one row, the
* results will be distinct)
* go through all the predicates and update tableColMap and
* eqOuterCol with join columns and correlation variables,
* parameters and constants
* since setting constants, correlation variables and parameters,
* reduces the number of columns required for uniqueness in a
* multi-column index, they are set for all the tables (if the
* table is not the result table, in this case only the column of the
* result table is set)
* join columns are just updated for the column in the row of the
* joining table.
*
* check if the marked columns in tableColMap are a superset of a unique
* index
* (This means that the join will only produce 1 row when joined
* with 1 row of another table)
* check that there is a least one table for which the columns in
* eqOuterCol(i.e. constant values) are a superset of a unique index
* (This quarantees that there will be only one row selected
* from this table).
*
* Once all tables have been evaluated, check that all the tables can be
* joined by unique index or will have only one row
*
*
*
* @param rcl If non-null, the RCL from the query block.
* If non-null for subqueries, then entry can
* be considered as part of an = comparison.
* @param whereClause The WHERE clause to consider.
* @param wherePredicates The predicates that have already been
* pulled from the WHERE clause.
* @param dd The DataDictionary to use.
*
* @return Whether or not query block will return
* at most 1 row for a subquery, no duplicates
* for a distinct.
*
* @exception StandardException Thrown on error
*/
boolean returnsAtMostSingleRow(ResultColumnList rcl,
ValueNode whereClause,
PredicateList wherePredicates,
DataDictionary dd)
throws StandardException
{
boolean satisfiesOuter = false;
int[] tableNumbers;
ColumnReference additionalCR = null;
PredicateList predicatesTemp;
predicatesTemp = (PredicateList) getNodeFactory().getNode(
C_NodeTypes.PREDICATE_LIST, getContextManager());
int wherePredicatesSize = wherePredicates.size();
for (int index = 0; index < wherePredicatesSize; index++)
predicatesTemp.addPredicate((Predicate)wherePredicates.elementAt(index));
/* When considering subquery flattening, we are interested
* in the 1st (and only) entry in the RCL. (The RCL will be
* null if result column is not of interest for subquery flattening.)
* We are interested in all entries in the RCL for distinct
* elimination.
*/
if (rcl != null)
{
ResultColumn rc = (ResultColumn) rcl.elementAt(0);
if (rc.getExpression() instanceof ColumnReference)
{
additionalCR = (ColumnReference) rc.getExpression();
}
}
/* First see if all entries are FromBaseTables. No point
* in continuing if not.
*/
int size = size();
for (int index = 0; index < size; index++)
{
FromTable fromTable = (FromTable) elementAt(index);
if (! (fromTable instanceof ProjectRestrictNode))
{
return false;
}
ProjectRestrictNode prn = (ProjectRestrictNode) fromTable;
if (! (prn.getChildResult() instanceof FromBaseTable))
{
return false;
}
FromBaseTable fbt = (FromBaseTable) prn.getChildResult();
//Following for loop code is to take care of Derby-251 (DISTINCT returns
//duplicate rows).
//Derby-251 returned duplicate rows because we were looking at predicates
//that belong to existsTable to determine DISTINCT elimination
//
//(Check method level comments to understand DISTINCT elimination rules.)
//
//For one specific example, consider the query below
//select distinct q1."NO1" from IDEPT q1, IDEPT q2
//where ( q2."DISCRIM_DEPT" = 'HardwareDept')
//and ( q1."DISCRIM_DEPT" = 'SoftwareDept') and ( q1."NO1" <> ALL
//(select q3."NO1" from IDEPT q3 where (q3."REPORTTO_NO" = q2."NO1")))
//(select q3."NO1" from IDEPT q3 where ( ABS(q3."REPORTTO_NO") = q2."NO1")))
//
//Table IDEPT in the query above has a primary key defined on column "NO1"
//This query gets converted to following during optimization
//
//select distinct q1."NO1" from IDEPT q1, IDEPT q2
//where ( q2."DISCRIM_DEPT" = 'HardwareDept')
//and ( q1."DISCRIM_DEPT" = 'SoftwareDept') and not exists (
//(select q3."NO1" from IDEPT q3 where
//( ( ABS(q3."REPORTTO_NO") = q2."NO1") and q3."NO1" = q1."NO1") ) ) ;
//
//For the optimized query above, Derby generates following predicates.
//ABS(q3.reportto_no) = q2.no1
//q2.discrim_dept = 'HardwareDept'
//q1.descrim_dept = 'SoftwareDept'
//q1.no1 = q3.no1
//The predicate ABS(q3."NO1") = q1."NO1" should not be considered when trying
//to determine if q1 in the outer query has equality comparisons.
//Similarly, the predicate q3.reportto_no = q2.no1 should not be
//considered when trying to determine if q2 in the outer query has
//equality comparisons. To achieve this, predicates based on exists base
//table q3 (the first and the last predicate) should be removed while
//evaluating outer query for uniqueness.
//
if (fbt.getExistsBaseTable())
{
int existsTableNumber = fbt.getTableNumber();
int predicatesTempSize = predicatesTemp.size();
for (int predicatesTempIndex = predicatesTempSize-1;
predicatesTempIndex >= 0; predicatesTempIndex--)
{
AndNode topAndNode = (AndNode)
((Predicate) predicatesTemp.elementAt(predicatesTempIndex)).getAndNode();
for (ValueNode whereWalker = topAndNode; whereWalker instanceof AndNode;
whereWalker = ((AndNode) whereWalker).getRightOperand())
{
// See if this is a candidate =
AndNode and = (AndNode) whereWalker;
//we only need to worry about equality predicates because only those
//predicates are considered during DISTINCT elimination.
if (!and.getLeftOperand().isRelationalOperator() ||
!(((RelationalOperator)(and.getLeftOperand())).getOperator() ==
RelationalOperator.EQUALS_RELOP))
{
continue;
}
JBitSet referencedTables = and.getLeftOperand().getTablesReferenced();
if (referencedTables.get(existsTableNumber))
{
predicatesTemp.removeElementAt(predicatesTempIndex);
break;
}
}
}
}
}
/* Build an array of tableNumbers from this query block.
* We will use that array to find out if we have at least
* one table with a uniqueness condition based only on
* constants, parameters and correlation columns.
*/
tableNumbers = getTableNumbers();
JBitSet[][] tableColMap = new JBitSet[size][size];
boolean[] oneRow = new boolean[size];
boolean oneRowResult = false;
/* See if each table has a uniqueness condition */
for (int index = 0; index < size; index++)
{
ProjectRestrictNode prn = (ProjectRestrictNode) elementAt(index);
FromBaseTable fbt = (FromBaseTable) prn.getChildResult();
// Skip over EXISTS FBT since they cannot introduce duplicates
if (fbt.getExistsBaseTable())
{
oneRow[index] = true;
continue;
}
int numColumns = fbt.getTableDescriptor().getNumberOfColumns();
boolean[] eqOuterCols = new boolean[numColumns + 1];
int tableNumber = fbt.getTableNumber();
boolean resultColTable = false;
for (int i = 0; i < size; i++)
tableColMap[index][i] = new JBitSet(numColumns + 1);
if (additionalCR != null &&
additionalCR.getTableNumber() == tableNumber)
{
rcl.recordColumnReferences(eqOuterCols, tableColMap[index], index);
resultColTable = true;
}
/* Now see if there are any equality conditions
* of interest in the where clause.
*/
if (whereClause != null)
{
whereClause.checkTopPredicatesForEqualsConditions(
tableNumber, eqOuterCols, tableNumbers,
tableColMap[index], resultColTable);
}
/* Now see if there are any equality conditions
* of interest in the where predicates.
*/
predicatesTemp.checkTopPredicatesForEqualsConditions(
tableNumber, eqOuterCols, tableNumbers,
tableColMap[index], resultColTable);
/* Now see if there are any equality conditions
* of interest that were already pushed down to the
* PRN above the FBT. (Single table predicates.)
*/
if (prn.getRestrictionList() != null)
{
prn.getRestrictionList().checkTopPredicatesForEqualsConditions(
tableNumber, eqOuterCols, tableNumbers,
tableColMap[index], resultColTable);
}
/* We can finally check to see if the marked columns
* are a superset of any unique index.
*/
if (! fbt.supersetOfUniqueIndex(tableColMap[index]))
{
return false;
}
/* Do we have at least 1 table whose equality condition
* is based solely on constants, parameters and correlation columns.
*/
oneRowResult = fbt.supersetOfUniqueIndex(eqOuterCols);
if (oneRowResult)
{
oneRow[index] = true;
satisfiesOuter = true;
}
}
/* Have we met all of the criteria */
if (satisfiesOuter)
{
/* check that all the tables are joined by unique indexes
* or only produce 1 row
*/
boolean foundOneRow = true;
while (foundOneRow)
{
foundOneRow = false;
for (int index = 0; index < size; index++)
{
if (oneRow[index])
{
for (int i = 0; i < size; i++)
{
/* unique key join - exists tables already marked as
* 1 row - so don't need to look at them
*/
if (!oneRow[i] && tableColMap[i][index].get(0))
{
oneRow[i] = true;
foundOneRow = true;
}
}
}
}
}
/* does any table produce more than one row */
for (int index = 0; index < size; index++)
{
if (!oneRow[index])
{
satisfiesOuter = false;
break;
}
}
}
return satisfiesOuter;
}
int[] getTableNumbers()
{
int size = size();
int[] tableNumbers = new int[size];
for (int index = 0; index < size; index++)
{
ProjectRestrictNode prn = (ProjectRestrictNode) elementAt(index);
if (! (prn.getChildResult() instanceof FromTable))
{
continue;
}
FromTable ft = (FromTable) prn.getChildResult();
tableNumbers[index] = ft.getTableNumber();
}
return tableNumbers;
}
/**
* Mark all of the FromBaseTables in the list as EXISTS FBTs.
* Each EBT has the same dependency list - those tables that are referenced
* minus the tables in the from list.
*
* @param referencedTableMap The referenced table map.
* @param outerFromList FromList from outer query block
* @param isNotExists Whether or not for NOT EXISTS
*
* @exception StandardException Thrown on error
*/
void genExistsBaseTables(JBitSet referencedTableMap, FromList outerFromList,
boolean isNotExists)
throws StandardException
{
JBitSet dependencyMap = (JBitSet) referencedTableMap.clone();
// We currently only flatten single table from lists
if (SanityManager.DEBUG)
{
if (size() != 1)
{
SanityManager.THROWASSERT(
"size() expected to be 1, not " + size());
}
}
/* Create the dependency map */
int size = size();
for (int index = 0; index < size; index++)
{
ResultSetNode ft = ((ProjectRestrictNode) elementAt(index)).getChildResult();
if (ft instanceof FromTable)
{
dependencyMap.clear(((FromTable) ft).getTableNumber());
}
}
/* Degenerate case - If flattening a non-correlated EXISTS subquery
* then we need to make the table that is getting flattened dependendent on
* all of the tables in the outer query block. Gross but true. Otherwise
* that table can get chosen as an outer table and introduce duplicates.
* The reason that duplicates can be introduced is that we do special processing
* in the join to make sure only one qualified row from the right side is
* returned. If the exists table is on the left, we can return all the
* qualified rows.
*/
if (dependencyMap.getFirstSetBit() == -1)
{
int outerSize = outerFromList.size();
for (int outer = 0; outer < outerSize; outer++)
dependencyMap.or(((FromTable) outerFromList.elementAt(outer)).getReferencedTableMap());
}
/* Do the marking */
for (int index = 0; index < size; index++)
{
FromTable fromTable = (FromTable) elementAt(index);
if (fromTable instanceof ProjectRestrictNode)
{
ProjectRestrictNode prn = (ProjectRestrictNode) fromTable;
if (prn.getChildResult() instanceof FromBaseTable)
{
FromBaseTable fbt = (FromBaseTable) prn.getChildResult();
fbt.setExistsBaseTable(true, (JBitSet) dependencyMap.clone(), isNotExists);
}
}
}
}
/**
* determine whether this table is NOT EXISTS.
*
* This routine searches for the indicated table number in the fromlist
* and returns TRUE if the table is present in the from list and is
* marked NOT EXISTS, false otherwise.
*
* A table may be present in the from list for NOT EXISTS if it is used
* as a correlated NOT EXISTS subquery. In such a situation, when the
* subquery is flattened, it is important that we remember that this is
* a NOT EXISTS subquery, because the join semantics are different
* (we're looking for rows that do NOT match, rather than rows
* that do). And since the join semantics are different, we cannot
* include this table into a transitive closure of equijoins
* (See DERBY-3033 for a situation where this occurs).
*
* @param tableNumber which table to check
* @return true if this table is in the from list as NOT EXISTS
*/
boolean tableNumberIsNotExists(int tableNumber)
throws StandardException
{
int size = size();
for (int index = 0; index < size; index++)
{
ProjectRestrictNode prn = (ProjectRestrictNode) elementAt(index);
if (! (prn.getChildResult() instanceof FromTable))
{
continue;
}
FromTable ft = (FromTable) prn.getChildResult();
if (ft.getTableNumber() == tableNumber)
return ft.isNotExists();
}
return false;
}
/**
* Get the lock mode for the target of an update statement
* (a delete or update). The update mode will always be row for
* CurrentOfNodes. It will be table if there is no where clause.
*
* @return The lock mode
*/
public int updateTargetLockMode()
{
if (SanityManager.DEBUG)
{
if (size() != 1)
{
SanityManager.THROWASSERT(
"size() expected to be 1");
}
}
return ((ResultSetNode) elementAt(0)).updateTargetLockMode();
}
/**
* Return whether or not the user specified a hash join for any of the
* tables in this list.
*
* @return Whether or not the user specified a hash join for any of the
* tables in this list.
*/
boolean hashJoinSpecified()
{
int size = size();
for (int index = 0; index < size; index++)
{
FromTable ft = (FromTable) elementAt(index);
String joinStrategy = ft.getUserSpecifiedJoinStrategy();
if (joinStrategy != null && StringUtil.SQLToUpperCase(joinStrategy).equals("HASH"))
{
return true;
}
}
return false;
}
/**
* Indicate that this FromList is "transparent", which means that
* its FromTables should be bound to tables from an outer query.
* Generally this is not allowed, but there are exceptions. See
* SetOperatorNode.setResultToBooleanTrueNode() for more.
*/
void markAsTransparent()
{
isTransparent = true;
}
/**
* Set windows field to the supplied value.
* @param windows list of window definitions associated with a SELECT.
*/
public void setWindows(WindowList windows) {
this.windows = windows;
}
/**
* @return list of window definitions associated with a SELECT.
*/
public WindowList getWindows() {
return windows;
}
}