org.apache.poi.hwpf.usermodel.FieldsImpl Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of poi-scratchpad Show documentation
Show all versions of poi-scratchpad Show documentation
Apache POI - Java API To Access Microsoft Format Files (Scratchpad)
/* ====================================================================
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.poi.hwpf.usermodel;
import java.io.Serializable;
import java.util.ArrayList;
import java.util.Collection;
import java.util.Collections;
import java.util.Comparator;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import org.apache.poi.hwpf.model.FieldDescriptor;
import org.apache.poi.hwpf.model.FieldsDocumentPart;
import org.apache.poi.hwpf.model.FieldsTables;
import org.apache.poi.hwpf.model.PlexOfField;
import org.apache.poi.util.Internal;
/**
* Default implementation of {@link Field}
*
* @author Sergey Vladimirov (vlsergey {at} gmail {dot} com)
*/
@Internal
public class FieldsImpl implements Fields
{
/**
* This is port and adaptation of Arrays.binarySearch from Java 6 (Apache
* Harmony).
*/
private static int binarySearch( List list,
int startIndex, int endIndex, int requiredStartOffset )
{
checkIndexForBinarySearch( list.size(), startIndex, endIndex );
int low = startIndex, mid = -1, high = endIndex - 1, result = 0;
while ( low <= high )
{
mid = ( low + high ) >>> 1;
int midStart = list.get( mid ).getFcStart();
if ( midStart == requiredStartOffset )
{
return mid;
}
else if ( midStart < requiredStartOffset )
{
low = mid + 1;
}
else
{
high = mid - 1;
}
}
if ( mid < 0 )
{
int insertPoint = endIndex;
for ( int index = startIndex; index < endIndex; index++ )
{
if ( requiredStartOffset < list.get( index ).getFcStart() )
{
insertPoint = index;
}
}
return -insertPoint - 1;
}
return -mid - ( result >= 0 ? 1 : 2 );
}
private static void checkIndexForBinarySearch( int length, int start,
int end )
{
if ( start > end )
{
throw new IllegalArgumentException();
}
if ( length < end || 0 > start )
{
throw new ArrayIndexOutOfBoundsException();
}
}
private Map> _fieldsByOffset;
private PlexOfFieldComparator comparator = new PlexOfFieldComparator();
public FieldsImpl( FieldsTables fieldsTables )
{
_fieldsByOffset = new HashMap>(
FieldsDocumentPart.values().length );
for ( FieldsDocumentPart part : FieldsDocumentPart.values() )
{
List plexOfCps = fieldsTables.getFieldsPLCF( part );
_fieldsByOffset.put( part, parseFieldStructure( plexOfCps ) );
}
}
public Collection getFields( FieldsDocumentPart part )
{
Map map = _fieldsByOffset.get( part );
if ( map == null || map.isEmpty() )
return Collections.emptySet();
return Collections. unmodifiableCollection( map.values() );
}
public FieldImpl getFieldByStartOffset( FieldsDocumentPart documentPart,
int offset )
{
Map map = _fieldsByOffset.get( documentPart );
if ( map == null || map.isEmpty() )
return null;
return map.get( Integer.valueOf( offset ) );
}
private Map parseFieldStructure(
List plexOfFields )
{
if ( plexOfFields == null || plexOfFields.isEmpty() )
return new HashMap();
Collections.sort( plexOfFields, comparator );
List fields = new ArrayList(
plexOfFields.size() / 3 + 1 );
parseFieldStructureImpl( plexOfFields, 0, plexOfFields.size(), fields );
HashMap result = new HashMap(
fields.size() );
for ( FieldImpl field : fields )
{
result.put( Integer.valueOf( field.getFieldStartOffset() ), field );
}
return result;
}
private void parseFieldStructureImpl( List plexOfFields,
int startOffsetInclusive, int endOffsetExclusive,
List result )
{
int next = startOffsetInclusive;
while ( next < endOffsetExclusive )
{
PlexOfField startPlexOfField = plexOfFields.get( next );
if ( startPlexOfField.getFld().getBoundaryType() != FieldDescriptor.FIELD_BEGIN_MARK )
{
/* Start mark seems to be missing */
next++;
continue;
}
/*
* we have start node. end offset points to next node, separator or
* end
*/
int nextNodePositionInList = binarySearch( plexOfFields, next + 1,
endOffsetExclusive, startPlexOfField.getFcEnd() );
if ( nextNodePositionInList < 0 )
{
/*
* too bad, this start field mark doesn't have corresponding end
* field mark or separator field mark in fields table
*/
next++;
continue;
}
PlexOfField nextPlexOfField = plexOfFields
.get( nextNodePositionInList );
switch ( nextPlexOfField.getFld().getBoundaryType() )
{
case FieldDescriptor.FIELD_SEPARATOR_MARK:
{
PlexOfField separatorPlexOfField = nextPlexOfField;
int endNodePositionInList = binarySearch( plexOfFields,
nextNodePositionInList, endOffsetExclusive,
separatorPlexOfField.getFcEnd() );
if ( endNodePositionInList < 0 )
{
/*
* too bad, this separator field mark doesn't have
* corresponding end field mark in fields table
*/
next++;
continue;
}
PlexOfField endPlexOfField = plexOfFields
.get( endNodePositionInList );
if ( endPlexOfField.getFld().getBoundaryType() != FieldDescriptor.FIELD_END_MARK )
{
/* Not and ending mark */
next++;
continue;
}
FieldImpl field = new FieldImpl( startPlexOfField,
separatorPlexOfField, endPlexOfField );
result.add( field );
// adding included fields
if ( startPlexOfField.getFcStart() + 1 < separatorPlexOfField
.getFcStart() - 1 )
{
parseFieldStructureImpl( plexOfFields, next + 1,
nextNodePositionInList, result );
}
if ( separatorPlexOfField.getFcStart() + 1 < endPlexOfField
.getFcStart() - 1 )
{
parseFieldStructureImpl( plexOfFields,
nextNodePositionInList + 1, endNodePositionInList,
result );
}
next = endNodePositionInList + 1;
break;
}
case FieldDescriptor.FIELD_END_MARK:
{
// we have no separator
FieldImpl field = new FieldImpl( startPlexOfField, null,
nextPlexOfField );
result.add( field );
// adding included fields
if ( startPlexOfField.getFcStart() + 1 < nextPlexOfField
.getFcStart() - 1 )
{
parseFieldStructureImpl( plexOfFields, next + 1,
nextNodePositionInList, result );
}
next = nextNodePositionInList + 1;
break;
}
case FieldDescriptor.FIELD_BEGIN_MARK:
default:
{
/* something is wrong, ignoring this mark along with start mark */
next++;
continue;
}
}
}
}
private static final class PlexOfFieldComparator implements Comparator, Serializable {
public int compare( PlexOfField o1, PlexOfField o2 )
{
int thisVal = o1.getFcStart();
int anotherVal = o2.getFcStart();
return thisVal < anotherVal ? -1 : thisVal == anotherVal ? 0 : 1;
}
}
}
© 2015 - 2024 Weber Informatics LLC | Privacy Policy