HSQLDB example source code file (Select.java)
The HSQLDB Select.java source code/*
* For work developed by the HSQL Development Group:
*
* Copyright (c) 2001-2010, The HSQL Development Group
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are met:
*
* Redistributions of source code must retain the above copyright notice, this
* list of conditions and the following disclaimer.
*
* Redistributions in binary form must reproduce the above copyright notice,
* this list of conditions and the following disclaimer in the documentation
* and/or other materials provided with the distribution.
*
* Neither the name of the HSQL Development Group nor the names of its
* contributors may be used to endorse or promote products derived from this
* software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
* AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL HSQL DEVELOPMENT GROUP, HSQLDB.ORG,
* OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
* LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
* ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
* SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*
*
*
* For work originally developed by the Hypersonic SQL Group:
*
* Copyright (c) 1995-2000, The Hypersonic SQL Group.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are met:
*
* Redistributions of source code must retain the above copyright notice, this
* list of conditions and the following disclaimer.
*
* Redistributions in binary form must reproduce the above copyright notice,
* this list of conditions and the following disclaimer in the documentation
* and/or other materials provided with the distribution.
*
* Neither the name of the Hypersonic SQL Group nor the names of its
* contributors may be used to endorse or promote products derived from this
* software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
* AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE HYPERSONIC SQL GROUP,
* OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
* LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
* ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
* SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*
* This software consists of voluntary contributions made by many individuals
* on behalf of the Hypersonic SQL Group.
*/
package org.hsqldb;
import org.hsqldb.HsqlNameManager.HsqlName;
import org.hsqldb.lib.HashMap;
import org.hsqldb.lib.HashSet;
import org.hsqldb.lib.HsqlArrayList;
import org.hsqldb.lib.IntKeyHashMap;
import org.hsqldb.lib.Iterator;
// fredt@users 20010701 - patch 1.6.1 by hybris
// basic implementation of LIMIT n m
// fredt@users 20020130 - patch 471710 by fredt - LIMIT rewritten
// for SELECT LIMIT n m DISTINCT
// fredt@users 20020221 - patch 513005 by sqlbob@users (RMP)
// type and logging attributes of sIntotable
// fredt@users 20020230 - patch 495938 by johnhobs@users - GROUP BY order
// fred@users 20020522 - patch 1.7.0 - aggregate functions with DISTINCT
// rougier@users 20020522 - patch 552830 - COUNT(DISTINCT)
// fredt@users 20020804 - patch 580347 by dkkopp - view speedup
// tony_lai@users 20021020 - patch 1.7.2 - improved aggregates and HAVING
// boucherb@users 20030811 - patch 1.7.2 - prepared statement support
// fredt@users 20031012 - patch 1.7.2 - better OUTER JOIN implementation
// fredt@users 20031012 - patch 1.7.2 - SQL standard ORDER BY with UNION and other set queries
// fredt@users 200408xx - patch 1.7.2 - correct evaluation of the precedence of nested UNION and other set query
/**
* The compiled representation of an SQL SELECT.
*
* Extended in successive versions of HSQLDB.
*
* @author Thomas Mueller (Hypersonic SQL Group)
* @version 1.8.0
* @since Hypersonic SQL
*/
class Select {
boolean isDistinctSelect;
boolean isAggregated;
private boolean isGrouped;
private HashSet groupColumnNames;
TableFilter[] tFilter;
Expression limitCondition;
Expression queryCondition; // null means no condition
Expression havingCondition; // null means none
Expression[] exprColumns; // 'result', 'group' and 'order' columns
int iResultLen; // number of columns that are 'result'
int iGroupLen; // number of columns that are 'group'
int iHavingLen; // number of columns that are 'group'
int iOrderLen; // number of columns that are 'order'
int[] sortOrder;
int[] sortDirection;
boolean sortUnion; // if true, sort the result of the full union
HsqlName sIntoTable; // null means not select..into
int intoType;
Select[] unionArray; // only set in the first Select in a union chain
int unionMaxDepth; // max unionDepth in chain
Select unionSelect; // null means no union select
int unionType;
int unionDepth;
static final int NOUNION = 0,
UNION = 1,
UNIONALL = 2,
INTERSECT = 3,
EXCEPT = 4;
private boolean simpleLimit; // true if maxrows can be uses as is
Result.ResultMetaData resultMetaData;
IntKeyHashMap asteriskPositions;
/**
* Experimental.
*
* Map the column aliases to expressions in order to resolve alias names
* in WHERE clauses
*
*/
HashMap getColumnAliases() {
HashMap aliasMap = new HashMap();
for (int i = 0; i < iResultLen; i++) {
String alias = exprColumns[i].getAlias();
if (alias != null) {
aliasMap.put(alias, exprColumns[i]);
}
}
return aliasMap;
}
/**
* Method declaration
*
*
* @throws HsqlException
*/
void resolve(Session session) throws HsqlException {
resolveTables();
resolveTypes(session);
setFilterConditions(session);
}
/**
* Method declaration
*
*
* @throws HsqlException
*/
private void resolveTables() throws HsqlException {
// replace the aliases with expressions
for (int i = iResultLen; i < exprColumns.length; i++) {
if (exprColumns[i].getType() == Expression.COLUMN) {
if (exprColumns[i].joinedTableColumnIndex == -1) {
boolean descending = exprColumns[i].isDescending();
exprColumns[i] =
exprColumns[i].getExpressionForAlias(exprColumns,
iResultLen);
if (descending) {
exprColumns[i].setDescending();
}
}
} else {
exprColumns[i].replaceAliases(exprColumns, iResultLen);
}
}
if (queryCondition != null) {
queryCondition.replaceAliases(exprColumns, iResultLen);
}
int len = tFilter.length;
for (int i = 0; i < len; i++) {
resolveTables(tFilter[i]);
}
}
/**
* Converts the types of the columns in set operations to those in the first
* Select.
*/
void resolveUnionColumnTypes() throws HsqlException {
if (unionSelect != null) {
if (unionSelect.iResultLen != iResultLen) {
throw Trace.error(Trace.COLUMN_COUNT_DOES_NOT_MATCH);
}
for (int i = 0; i < iResultLen; i++) {
Expression e = exprColumns[i];
if (!e.isTypeEqual(unionSelect.exprColumns[i])) {
unionSelect.exprColumns[i] =
new Expression(unionSelect.exprColumns[i],
e.getDataType(), e.getColumnSize(),
e.getColumnScale());
}
}
}
}
/**
* Sets the types of all the expressions that have so far resolved.
*
* @throws HsqlException
*/
void resolveTypes(Session session) throws HsqlException {
int len = exprColumns.length;
for (int i = 0; i < len; i++) {
exprColumns[i].resolveTypes(session);
}
if (queryCondition != null) {
queryCondition.resolveTypes(session);
}
}
void resolveTablesUnion(TableFilter f) throws HsqlException {
if (unionArray == null) {
resolveTables(f);
} else {
for (int i = 0; i < unionArray.length; i++) {
unionArray[i].resolveTables(f);
}
}
}
/**
* Resolves the tables for all the Expression in the Select object
* if it is possible to do so with the given TableFilter.
*
* @param f
*
* @throws HsqlException
*/
void resolveTables(TableFilter f) throws HsqlException {
int len = exprColumns.length;
for (int i = 0; i < len; i++) {
exprColumns[i].resolveTables(f);
}
if (queryCondition != null) {
queryCondition.resolveTables(f);
}
}
private void setFilterConditions(Session session) throws HsqlException {
if (queryCondition == null) {
return;
}
for (int i = 0; i < tFilter.length; i++) {
tFilter[i].setConditions(session, queryCondition);
}
}
/**
* Check all Expression have resolved. Return true or false as a result.
* Throw if false and check parameter is true.
*
* @throws HsqlException
*/
boolean checkResolved(boolean check) throws HsqlException {
boolean result = true;
int len = exprColumns.length;
for (int i = 0; i < len; i++) {
result = result && exprColumns[i].checkResolved(check);
}
if (queryCondition != null) {
result = result && queryCondition.checkResolved(check);
}
if (havingCondition != null) {
result = result && havingCondition.checkResolved(check);
}
for (int i = 0; i < tFilter.length; i++) {
if (tFilter[i].filterIndex == null) {
tFilter[i].filterIndex =
tFilter[i].filterTable.getPrimaryIndex();
}
}
return result;
}
/**
* Removes all the TableFilters from the Expressions.
*
* @throws HsqlException
*/
/*
void removeFilters() throws HsqlException {
int len = eColumn.length;
for (int i = 0; i < len; i++) {
eColumn[i].removeFilters();
}
if (eCondition != null) {
eCondition.removeFilters();
}
}
*/
/**
* Returns a single value result or throws if the result has more than
* one row with one value.
*
* @param type data type
* @param session context
* @return the single valued result
* @throws HsqlException
*/
Object getValue(Session session, int type) throws HsqlException {
resolve(session);
Result r = getResult(session, 2); // 2 records are required for test
int size = r.getSize();
int len = r.getColumnCount();
if (len == 1) {
if (size == 0) {
return null;
} else if (size == 1) {
Object o = r.rRoot.data[0];
return r.metaData.colTypes[0] == type ? o
: Column.convertObject(
o, type);
} else {
throw Trace.error(Trace.CARDINALITY_VIOLATION_NO_SUBCLASS);
}
}
HsqlException e =
Trace.error(Trace.CARDINALITY_VIOLATION_NO_SUBCLASS);
throw new HsqlInternalException(e);
}
/**
* Resolves expressions and pepares thre metadata for the result.
*/
void prepareResult(Session session) throws HsqlException {
resolveAll(session, true);
if (iGroupLen > 0) { // has been set in Parser
isGrouped = true;
groupColumnNames = new HashSet();
for (int i = iResultLen; i < iResultLen + iGroupLen; i++) {
// MarcH: this is wrong for a CASE WHEN statement in a SELECT CASE WHEN ...,<something aggregate> statement
// collectColumnName collects no columns if exprColumns[i]'s expressiontype is Expression.CASEWHEN
// collectAllColumnNames collects all columns used in the CASE WHEN statement
// exprColumns[i].collectColumnName(groupColumnNames);
exprColumns[i].collectAllColumnNames(groupColumnNames);
}
}
int len = exprColumns.length;
resultMetaData = new Result.ResultMetaData(len);
Result.ResultMetaData rmd = resultMetaData;
// tony_lai@users having
int groupByStart = iResultLen;
int groupByEnd = groupByStart + iGroupLen;
int orderByStart = groupByEnd + iHavingLen;
int orderByEnd = orderByStart + iOrderLen;
for (int i = 0; i < len; i++) {
Expression e = exprColumns[i];
rmd.colTypes[i] = e.getDataType();
rmd.colSizes[i] = e.getColumnSize();
rmd.colScales[i] = e.getColumnScale();
if (e.isAggregate()) {
isAggregated = true;
}
if (i >= groupByStart && i < groupByEnd
&&!exprColumns[i].canBeInGroupBy()) {
Trace.error(Trace.INVALID_GROUP_BY, exprColumns[i]);
}
if (i >= groupByEnd && i < groupByEnd + iHavingLen
&&!exprColumns[i].isConditional()) {
Trace.error(Trace.INVALID_HAVING, exprColumns[i]);
}
if (i >= orderByStart && i < orderByEnd
&&!exprColumns[i].canBeInOrderBy()) {
Trace.error(Trace.INVALID_ORDER_BY, exprColumns[i]);
}
if (i < iResultLen) {
rmd.colLabels[i] = e.getAlias();
rmd.isLabelQuoted[i] = e.isAliasQuoted();
rmd.schemaNames[i] = e.getTableSchemaName();
rmd.tableNames[i] = e.getTableName();
rmd.colNames[i] = e.getColumnName();
if (rmd.isTableColumn(i)) {
rmd.colNullable[i] = e.nullability;
rmd.isIdentity[i] = e.isIdentity;
rmd.isWritable[i] = e.isWritable;
}
rmd.classNames[i] = e.getValueClassName();
}
}
// selected columns
checkAggregateOrGroupByColumns(0, iResultLen);
// having columns
checkAggregateOrGroupByColumns(groupByEnd, orderByStart);
// order by columns
checkAggregateOrGroupByOrderColumns(orderByStart, orderByEnd);
prepareSort();
simpleLimit = (isDistinctSelect == false && isGrouped == false
&& unionSelect == null && iOrderLen == 0);
}
/**
* This is called externally only on the first Select in a UNION chain.
*/
void prepareUnions() throws HsqlException {
int count = 0;
for (Select current = this; current != null;
current = current.unionSelect, count++) {}
if (count == 1) {
if (unionDepth != 0) {
throw Trace.error(Trace.MISSING_CLOSEBRACKET);
}
return;
}
unionArray = new Select[count];
count = 0;
for (Select current = this; current != null;
current = current.unionSelect, count++) {
unionArray[count] = current;
unionMaxDepth = current.unionDepth > unionMaxDepth
? current.unionDepth
: unionMaxDepth;
}
if (unionArray[unionArray.length - 1].unionDepth != 0) {
throw Trace.error(Trace.MISSING_CLOSEBRACKET);
}
}
/**
* Returns the result of executing this Select.
*
* @param maxrows may be 0 to indicate no limit on the number of rows.
* Positive values limit the size of the result set.
* @return the result of executing this Select
* @throws HsqlException if a database access error occurs
*/
Result getResult(Session session, int maxrows) throws HsqlException {
Result r;
if (unionArray == null) {
r = getSingleResult(session, maxrows);
} else {
r = getResultMain(session);
if (sortUnion) {
sortResult(session, r);
r.trimResult(getLimitStart(session),
getLimitCount(session, maxrows));
}
}
// fredt - now there is no need for the sort and group columns
r.setColumnCount(iResultLen);
return r;
}
private Result getResultMain(Session session) throws HsqlException {
Result[] unionResults = new Result[unionArray.length];
for (int i = 0; i < unionArray.length; i++) {
unionResults[i] = unionArray[i].getSingleResult(session,
Integer.MAX_VALUE);
}
for (int depth = unionMaxDepth; depth >= 0; depth--) {
for (int pass = 0; pass < 2; pass++) {
for (int i = 0; i < unionArray.length - 1; i++) {
if (unionResults[i] != null
&& unionArray[i].unionDepth >= depth) {
if (pass == 0
&& unionArray[i].unionType
!= Select.INTERSECT) {
continue;
}
if (pass == 1
&& unionArray[i].unionType
== Select.INTERSECT) {
continue;
}
int nextIndex = i + 1;
for (; nextIndex < unionArray.length; nextIndex++) {
if (unionResults[nextIndex] != null) {
break;
}
}
if (nextIndex == unionArray.length) {
break;
}
unionArray[i].mergeResults(session, unionResults[i],
unionResults[nextIndex]);
unionResults[nextIndex] = unionResults[i];
unionResults[i] = null;
}
}
}
}
return unionResults[unionResults.length - 1];
}
/**
* Merges the second result into the first using the unionMode
* set operation.
*/
private void mergeResults(Session session, Result first,
Result second) throws HsqlException {
switch (unionType) {
case UNION :
first.append(second);
first.removeDuplicates(session, iResultLen);
break;
case UNIONALL :
first.append(second);
break;
case INTERSECT :
first.removeDifferent(session, second, iResultLen);
break;
case EXCEPT :
first.removeSecond(session, second, iResultLen);
break;
}
}
int getLimitStart(Session session) throws HsqlException {
if (limitCondition != null) {
Integer limit =
(Integer) limitCondition.getArg().getValue(session);
if (limit != null) {
return limit.intValue();
}
}
return 0;
}
/**
* For SELECT LIMIT n m ....
* finds cases where the result does not have to be fully built and
* returns an adjusted rowCount with LIMIT params.
*/
int getLimitCount(Session session, int rowCount) throws HsqlException {
int limitCount = 0;
if (limitCondition != null) {
Integer limit =
(Integer) limitCondition.getArg2().getValue(session);
if (limit != null) {
limitCount = limit.intValue();
}
}
if (rowCount != 0 && (limitCount == 0 || rowCount < limitCount)) {
limitCount = rowCount;
}
return limitCount;
}
/**
* translate the rowCount into total number of rows needed from query,
* including any rows skipped at the beginning
*/
int getMaxRowCount(Session session, int rowCount) throws HsqlException {
int limitStart = getLimitStart(session);
int limitCount = getLimitCount(session, rowCount);
if (!simpleLimit) {
rowCount = Integer.MAX_VALUE;
} else {
if (rowCount == 0) {
rowCount = limitCount;
}
if (rowCount == 0 || rowCount > Integer.MAX_VALUE - limitStart) {
rowCount = Integer.MAX_VALUE;
} else {
rowCount += limitStart;
}
}
return rowCount;
}
private Result getSingleResult(Session session,
int rowCount) throws HsqlException {
if (resultMetaData == null) {
prepareResult(session);
}
Result r = buildResult(session, getMaxRowCount(session, rowCount));
// the result is perhaps wider (due to group and order by)
// so use the visible columns to remove duplicates
if (isDistinctSelect) {
r.removeDuplicates(session, iResultLen);
}
if (!sortUnion) {
sortResult(session, r);
r.trimResult(getLimitStart(session),
getLimitCount(session, rowCount));
}
return r;
}
private void prepareSort() {
if (iOrderLen == 0) {
return;
}
sortOrder = new int[iOrderLen];
sortDirection = new int[iOrderLen];
int startCol = iResultLen + iGroupLen + iHavingLen;
for (int i = startCol, j = 0; j < iOrderLen; i++, j++) {
int colindex = i;
// fredt - when a union, use the visible select columns for sort comparison
// also whenever a column alias is used
if (exprColumns[i].joinedTableColumnIndex != -1) {
colindex = exprColumns[i].joinedTableColumnIndex;
}
sortOrder[j] = colindex;
sortDirection[j] = exprColumns[i].isDescending() ? -1
: 1;
}
}
private void sortResult(Session session, Result r) throws HsqlException {
if (iOrderLen == 0) {
return;
}
r.sortResult(session, sortOrder, sortDirection);
}
/**
* Check result columns for aggregate or group by violation.
* If any result column is aggregated, then all result columns need to be
* aggregated, unless it is included in the group by clause.
*/
private void checkAggregateOrGroupByColumns(int start,
int end) throws HsqlException {
if (start < end) {
HsqlArrayList colExps = new HsqlArrayList();
for (int i = start; i < end; i++) {
exprColumns[i].collectInGroupByExpressions(colExps);
}
for (int i = 0, size = colExps.size(); i < size; i++) {
Expression exp = (Expression) colExps.get(i);
if (inAggregateOrGroupByClause(exp)) {
continue;
}
throw Trace.error(Trace.NOT_IN_AGGREGATE_OR_GROUP_BY, exp);
}
}
}
private void checkAggregateOrGroupByOrderColumns(int start,
int end) throws HsqlException {
checkAggregateOrGroupByColumns(start, end);
if (start < end && isDistinctSelect) {
HsqlArrayList colExps = new HsqlArrayList();
for (int i = start; i < end; i++) {
exprColumns[i].collectInGroupByExpressions(colExps);
}
for (int i = 0, size = colExps.size(); i < size; i++) {
Expression exp = (Expression) colExps.get(i);
if (isSimilarIn(exp, 0, iResultLen)) {
continue;
}
throw Trace.error(Trace.INVALID_ORDER_BY_IN_DISTINCT_SELECT,
exp);
}
}
}
/**
* Check if the given expression is acceptable in a select that may
* include aggregate function and/or group by clause.
* <p>
* The expression is acceptable if:
* <UL>
* <LI>The select does not containt any aggregate function;
* <LI>The expression itself can be included in an aggregate select;
* <LI>The expression is defined in the group by clause;
* <LI>All the columns in the expression are defined in the group by clause;
* </UL)
*/
private boolean inAggregateOrGroupByClause(Expression exp) {
if (isGrouped) {
return isSimilarIn(exp, iResultLen, iResultLen + iGroupLen)
|| allColumnsAreDefinedIn(exp, groupColumnNames);
} else if (isAggregated) {
return exp.canBeInAggregate();
} else {
return true;
}
}
/**
* Check if the given expression is similar to any of the eColumn
* expressions within the given range.
*/
private boolean isSimilarIn(Expression exp, int start, int end) {
for (int i = start; i < end; i++) {
if (exp.similarTo(exprColumns[i])) {
return true;
}
}
return false;
}
/**
* Check if all the column names used in the given expression are defined
* in the given defined column names.
*/
static boolean allColumnsAreDefinedIn(Expression exp,
HashSet definedColumns) {
HashSet colNames = new HashSet();
exp.collectAllColumnNames(colNames);
if ((colNames.size() > 0) && (definedColumns == null)) {
return false;
}
Iterator i = colNames.iterator();
while (i.hasNext()) {
if (!definedColumns.contains(i.next())) {
return false;
}
}
return true;
}
// fredt@users 20030810 - patch 1.7.2 - OUTER JOIN rewrite
private Result buildResult(Session session,
int limitcount) throws HsqlException {
GroupedResult gResult = new GroupedResult(this, resultMetaData);
final int len = exprColumns.length;
final int filter = tFilter.length;
boolean[] first = new boolean[filter];
boolean[] outerused = new boolean[filter];
int level = 0;
// fredt - shortcut needed by OpenOffice to speed up empty query processing for metadata
boolean notempty = !(queryCondition != null
&& queryCondition.isFixedConditional()
&&!queryCondition.testCondition(session));
while (notempty && level >= 0) {
// perform a join
TableFilter t = tFilter[level];
boolean found;
boolean outerfound;
if (!first[level]) {
found = t.findFirst(session);
// if outer join, and no inner result, get next outer row
// nonJoinIsNull disallows getting the next outer row in some circumstances
outerused[level] = outerfound = t.isOuterJoin &&!found
&&!outerused[level]
&&!t.nonJoinIsNull
&& t.nextOuter(session);
first[level] = found;
} else {
found = t.next(session);
outerused[level] = outerfound = t.isOuterJoin &&!found
&&!first[level]
&&!outerused[level]
&&!t.nonJoinIsNull
&& t.nextOuter(session);
first[level] = found;
}
if (!found &&!outerfound) {
level--;
continue;
}
if (level < filter - 1) {
level++;
continue;
} else {
while (outerused[level]) {
outerused[level--] = false;
}
}
// apply condition
if (queryCondition == null
|| queryCondition.testCondition(session)) {
try {
Object[] row = new Object[len];
// gets the group by column values first.
for (int i = gResult.groupBegin; i < gResult.groupEnd;
i++) {
row[i] = exprColumns[i].getValue(session);
}
row = gResult.getRow(row);
// Get all other values
for (int i = 0; i < gResult.groupBegin; i++) {
row[i] =
isAggregated && exprColumns[i].isAggregate()
? exprColumns[i].updateAggregatingValue(session,
row[i])
: exprColumns[i].getValue(session);
}
for (int i = gResult.groupEnd; i < len; i++) {
row[i] =
isAggregated && exprColumns[i].isAggregate()
? exprColumns[i].updateAggregatingValue(session,
row[i])
: exprColumns[i].getValue(session);
}
gResult.addRow(row);
if (gResult.size() >= limitcount) {
break;
}
} catch (HsqlInternalException e) {
continue;
}
}
}
if (isAggregated &&!isGrouped && gResult.size() == 0) {
Object[] row = new Object[len];
for (int i = 0; i < len; i++) {
row[i] = exprColumns[i].isAggregate() ? null
: exprColumns[i]
.getValue(session);
}
gResult.addRow(row);
}
Iterator it = gResult.iterator();
while (it.hasNext()) {
Object[] row = (Object[]) it.next();
if (isAggregated) {
for (int i = 0; i < len; i++) {
if (exprColumns[i].isAggregate()) {
row[i] = exprColumns[i].getAggregatedValue(session,
row[i]);
}
}
}
if (iHavingLen > 0) {
// The test value, either aggregate or not, is set already.
// Removes the row that does not satisfy the HAVING
// condition.
if (!Boolean.TRUE.equals(row[iResultLen + iGroupLen])) {
it.remove();
}
}
}
return gResult.getResult();
}
/**
* Skeleton under development. Needs a lot of work.
*/
public StringBuffer getDDL() throws HsqlException {
StringBuffer sb = new StringBuffer();
sb.append(Token.T_SELECT).append(' ');
//limitStart;
//limitCount;
for (int i = 0; i < iResultLen; i++) {
sb.append(exprColumns[i].getDDL());
if (i < iResultLen - 1) {
sb.append(',');
}
}
sb.append(Token.T_FROM);
for (int i = 0; i < tFilter.length; i++) {
// find out if any expression in any of the filters isInJoin then use this form
TableFilter filter = tFilter[i];
// if any expression isInJoin
if (i != 0) {
if (filter.isOuterJoin) {
sb.append(Token.T_FROM).append(' ');
sb.append(Token.T_JOIN).append(' ');
}
// eStart and eEnd expressions
}
// otherwise use a comma delimited table list
sb.append(',');
}
// if there are any expressions that are not isInJoin
sb.append(' ').append(Token.T_WHERE).append(' ');
for (int i = 0; i < tFilter.length; i++) {
TableFilter filter = tFilter[i];
// eStart and eEnd expressions that are not isInJoin
}
// if has GROUP BY
sb.append(' ').append(Token.T_GROUP).append(' ');
for (int i = iResultLen; i < iResultLen + iGroupLen; i++) {
sb.append(exprColumns[i].getDDL());
if (i < iResultLen + iGroupLen - 1) {
sb.append(',');
}
}
// if has HAVING
sb.append(' ').append(Token.T_HAVING).append(' ');
for (int i = iResultLen + iGroupLen;
i < iResultLen + iGroupLen + iHavingLen; i++) {
sb.append(exprColumns[i].getDDL());
if (i < iResultLen + iGroupLen - 1) {
sb.append(',');
}
}
if (unionSelect != null) {
switch (unionType) {
case EXCEPT :
sb.append(' ').append(Token.T_EXCEPT).append(' ');
break;
case INTERSECT :
sb.append(' ').append(Token.T_INTERSECT).append(' ');
break;
case UNION :
sb.append(' ').append(Token.T_UNION).append(' ');
break;
case UNIONALL :
sb.append(' ').append(Token.T_UNION).append(' ').append(
Token.T_ALL).append(' ');
break;
}
}
// if has ORDER BY
int groupByEnd = iResultLen + iGroupLen;
int orderByStart = groupByEnd + iHavingLen;
int orderByEnd = orderByStart + iOrderLen;
sb.append(' ').append(Token.T_ORDER).append(Token.T_BY).append(' ');
for (int i = orderByStart; i < orderByEnd; i++) {
sb.append(exprColumns[i].getDDL());
if (i < iResultLen + iGroupLen - 1) {
sb.append(',');
}
}
return sb;
}
boolean isResolved = false;
/**
* @todo - post 1.8.0 - review resolve and check resolve -
* determine if isResolved is specific to main query or the full set including UNION
*
*/
boolean resolveAll(Session session, boolean check) throws HsqlException {
if (isResolved) {
return true;
}
resolve(session);
isResolved = checkResolved(check);
if (unionSelect != null) {
if (unionSelect.iResultLen != iResultLen) {
throw Trace.error(Trace.COLUMN_COUNT_DOES_NOT_MATCH);
}
for (int i = 0; i < iResultLen; i++) {
Expression e = exprColumns[i];
if (!e.isTypeEqual(unionSelect.exprColumns[i])) {
unionSelect.exprColumns[i] =
new Expression(unionSelect.exprColumns[i],
e.getDataType(), e.getColumnSize(),
e.getColumnScale());
}
}
isResolved &= unionSelect.resolveAll(session, check);
}
return isResolved;
}
boolean isResolved() {
return isResolved;
}
public String describe(Session session) {
StringBuffer sb;
String temp;
// temporary : it is currently unclear whether this may affect
// later attempts to retrieve an actual result (calls getResult(1)
// in preProcess mode). Thus, toString() probably should not be called
// on Select objects that will actually be used to retrieve results,
// only on Select objects used by EXPLAIN PLAN FOR
try {
getResult(session, 1);
} catch (HsqlException e) {}
sb = new StringBuffer();
sb.append(super.toString()).append("[\n");
if (sIntoTable != null) {
sb.append("into table=[").append(sIntoTable.name).append("]\n");
}
if (limitCondition != null) {
sb.append("offset=[").append(
limitCondition.getArg().describe(session)).append("]\n");
sb.append("limit=[").append(
limitCondition.getArg2().describe(session)).append("]\n");
}
sb.append("isDistinctSelect=[").append(isDistinctSelect).append(
"]\n");
sb.append("isGrouped=[").append(isGrouped).append("]\n");
sb.append("isAggregated=[").append(isAggregated).append("]\n");
sb.append("columns=[");
int columns = exprColumns.length - iOrderLen;
for (int i = 0; i < columns; i++) {
sb.append(exprColumns[i].describe(session));
}
sb.append("\n]\n");
sb.append("tableFilters=[\n");
for (int i = 0; i < tFilter.length; i++) {
sb.append("[\n");
sb.append(tFilter[i].describe(session));
sb.append("\n]");
}
sb.append("]\n");
temp = queryCondition == null ? "null"
: queryCondition.describe(session);
sb.append("eCondition=[").append(temp).append("]\n");
temp = havingCondition == null ? "null"
: havingCondition.describe(session);
sb.append("havingCondition=[").append(temp).append("]\n");
sb.append("groupColumns=[").append(groupColumnNames).append("]\n");
if (unionSelect != null) {
switch (unionType) {
case EXCEPT :
sb.append(" EXCEPT ");
break;
case INTERSECT :
sb.append(" INTERSECT ");
break;
case UNION :
sb.append(" UNION ");
break;
case UNIONALL :
sb.append(" UNION ALL ");
break;
default :
sb.append(" UNKNOWN SET OPERATION ");
}
sb.append("[\n").append(unionSelect.describe(session)).append(
"]\n");
}
return sb.toString();
}
Result describeResult() {
Result r;
Result.ResultMetaData rmd;
Expression e;
r = new Result(ResultConstants.DATA, iResultLen);
rmd = r.metaData;
for (int i = 0; i < iResultLen; i++) {
e = exprColumns[i];
rmd.colTypes[i] = e.getDataType();
rmd.classNames[i] = e.getValueClassName();
rmd.colSizes[i] = e.getColumnSize();
rmd.colScales[i] = e.getColumnScale();
rmd.colLabels[i] = e.getAlias();
rmd.isLabelQuoted[i] = e.isAliasQuoted();
rmd.tableNames[i] = e.getTableName();
rmd.colNames[i] = e.getColumnName();
if (rmd.isTableColumn(i)) {
rmd.schemaNames[i] = e.getTableSchemaName();
rmd.colNullable[i] = e.nullability;
rmd.isIdentity[i] = e.isIdentity;
rmd.isWritable[i] = e.isWritable;
}
}
return r;
}
}
Other HSQLDB examples (source code examples)Here is a short list of links related to this HSQLDB Select.java source code file: |
The search page
Other HSQLDB source code examples at this package level
Click here to learn more about this project
