Apache
/
poi
mirror of https://github.com/apache/poi.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Rework EscherRecordHolder parsing 

Modify the parsing done by EscherRecordHolder to be more deterministic. The format of the OfficeArtContent structure, which the EscherRecordHolder represents, is well defined in the MS-DOC spec. A clear class structure makes it easier to reason about the availability of data.

git-svn-id: https://svn.apache.org/repos/asf/poi/trunk@1887008 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Marius Volkhart 2021-02-28 18:49:42 +00:00
parent fab0ec3e08
commit 9c88bb2201
1 changed files with 87 additions and 107 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

180
src/scratchpad/src/org/apache/poi/hwpf/model/EscherRecordHolder.java
View File

@ -18,15 +18,18 @@
package org.apache.poi.hwpf.model;
import java.util.ArrayList;
import java.util.Iterator;
import java.util.List;
import org.apache.logging.log4j.LogManager;
import org.apache.poi.ddf.DefaultEscherRecordFactory;
import org.apache.poi.ddf.EscherContainerRecord;
import org.apache.poi.ddf.EscherRecord;
import org.apache.poi.ddf.EscherRecordFactory;
import org.apache.poi.ddf.EscherRecordTypes;
import org.apache.poi.util.Internal;
import static org.apache.logging.log4j.util.Unbox.box;
/**
* Based on AbstractEscherRecordHolder from HSSF.
*
@ -34,7 +37,27 @@ import org.apache.poi.util.Internal;
*/
@Internal
public final class EscherRecordHolder {
private final ArrayList<EscherRecord> escherRecords = new ArrayList<>();
/**
* {@link EscherRecordTypes#DGG_CONTAINER} containing drawing group information for the document.
*/
private final EscherContainerRecord drawingGroupData = new EscherContainerRecord();
/**
* {@link EscherRecordTypes#DG_CONTAINER} for drawings in the Main Document.
* <p>
* {@code null} to indicate that the document does not have a {@link EscherRecordTypes#DG_CONTAINER} for the Main
* Document.
*/
private EscherContainerRecord mainDocumentDgContainer;
/**
* {@link EscherRecordTypes#DG_CONTAINER} for drawings in the Header Document.
* <p>
* {@code null} to indicate that the document does not have a {@link EscherRecordTypes#DG_CONTAINER} for the Header
* Document.
*/
private EscherContainerRecord headerDocumentDgContainer;
public EscherRecordHolder(byte[] data, int offset, int size) {
fillEscherRecords(data, offset, size);
@ -47,129 +70,77 @@ public final class EscherRecordHolder {
*
* @see FileInformationBlock#getLcbDggInfo()
*/
private void fillEscherRecords(byte[] data, int offset, int size)
{
private void fillEscherRecords(byte[] data, int offset, int size) {
if (size == 0) return;
EscherRecordFactory recordFactory = new DefaultEscherRecordFactory();
int pos = offset;
while ( pos < offset + size)
{
EscherRecord r = recordFactory.createRecord(data, pos);
escherRecords.add(r);
int bytesRead = r.fillFields(data, pos, recordFactory);
pos += bytesRead + 1; // There is an empty byte between each top-level record in a Word doc
pos += drawingGroupData.fillFields(data, pos, recordFactory);
assert drawingGroupData.getRecordId() == EscherRecordTypes.DGG_CONTAINER.typeID;
/*
* After the drawingGroupData there is an array (2 slots max) that has data about drawings. According to the
* spec, the first slot is for the Main Document, the second for the Header Document. Additionally, the
* OfficeArtWordDrawing structure has a byte (dgglbl) that indicates whether the structure is for the Main or
* Header Document. In practice we've seen documents such as 61911.doc where the order of array entries does not
* match the dgglbl byte. As the byte is more likely to be reliable, we base the parsing off of that rather than
* array order.
*/
// This should loop at most twice
while (pos < offset + size) {
// Named this way to match section 2.9.172 of [MS-DOC] - v20191119.
byte dgglbl = data[pos];
assert dgglbl == 0x00 || dgglbl == 0x01;
pos++;
EscherContainerRecord dgContainer = new EscherContainerRecord();
pos+= dgContainer.fillFields(data, pos, recordFactory);
assert dgContainer.getRecordId() == EscherRecordTypes.DG_CONTAINER.typeID;
switch (dgglbl) {
case 0x00:
mainDocumentDgContainer = dgContainer;
break;
case 0x01:
headerDocumentDgContainer = dgContainer;
break;
default:
LogManager.getLogger(EscherRecordHolder.class).atWarn()
.log("dgglbl {} for OfficeArtWordDrawing is out of bounds [0, 1]", box(dgglbl));
}
}
assert pos == offset + size;
}
public List<EscherRecord> getEscherRecords() {
return escherRecords;
return drawingGroupData.getChildRecords();
}
@Override
public String toString() {
StringBuilder buffer = new StringBuilder();
if (escherRecords.size() == 0) {
buffer.append("No Escher Records Decoded").append("\n");
}
Iterator<EscherRecord> iterator = escherRecords.iterator();
while (iterator.hasNext()) {
EscherRecord r = iterator.next();
buffer.append(r);
}
return buffer.toString();
}
/**
* If we have a EscherContainerRecord as one of our
* children (and most top level escher holders do),
* then return that.
*/
public EscherContainerRecord getEscherContainer() {
for(Iterator<EscherRecord> it = escherRecords.iterator(); it.hasNext();) {
Object er = it.next();
if(er instanceof EscherContainerRecord) {
return (EscherContainerRecord)er;
}
}
return null;
}
/**
* Descends into all our children, returning the
* first EscherRecord with the given id, or null
* if none found
*/
public EscherRecord findFirstWithId(short id) {
return findFirstWithId(id, getEscherRecords());
}
private static EscherRecord findFirstWithId(short id, List<EscherRecord> records) {
// Check at our level
for(Iterator<EscherRecord> it = records.iterator(); it.hasNext();) {
EscherRecord r = it.next();
if(r.getRecordId() == id) {
return r;
}
}
// Then check our children in turn
for(Iterator<EscherRecord> it = records.iterator(); it.hasNext();) {
EscherRecord r = it.next();
if(r.isContainerRecord()) {
EscherRecord found = findFirstWithId(id, r.getChildRecords());
if(found != null) {
return found;
}
}
}
// Not found in this lot
return null;
}
public List<? extends EscherContainerRecord> getDgContainers()
{
List<EscherContainerRecord> dgContainers = new ArrayList<>(
1);
for ( EscherRecord escherRecord : getEscherRecords() )
{
if ( escherRecord.getRecordId() == (short) 0xF002 )
{
dgContainers.add( (EscherContainerRecord) escherRecord );
public List<? extends EscherContainerRecord> getDgContainers() {
List<EscherContainerRecord> dgContainers = new ArrayList<>(2);
if (mainDocumentDgContainer != null) {
dgContainers.add(mainDocumentDgContainer);
}
if (headerDocumentDgContainer != null) {
dgContainers.add(headerDocumentDgContainer);
}
return dgContainers;
}
public List<? extends EscherContainerRecord> getDggContainers()
{
List<EscherContainerRecord> dggContainers = new ArrayList<>(
1);
for ( EscherRecord escherRecord : getEscherRecords() )
{
if ( escherRecord.getRecordId() == (short) 0xF000 )
{
dggContainers.add( (EscherContainerRecord) escherRecord );
}
}
return dggContainers;
}
public List<? extends EscherContainerRecord> getBStoreContainers()
{
List<EscherContainerRecord> bStoreContainers = new ArrayList<>(
1);
for ( EscherContainerRecord dggContainer : getDggContainers() )
{
for ( EscherRecord escherRecord : dggContainer.getChildRecords() )
for ( EscherRecord escherRecord : drawingGroupData.getChildRecords() )
{
if ( escherRecord.getRecordId() == (short) 0xF001 )
{
bStoreContainers.add( (EscherContainerRecord) escherRecord );
}
}
}
return bStoreContainers;
}
@ -206,4 +177,13 @@ public final class EscherRecordHolder {
}
return spContainers;
}
@Override
public String toString() {
return "EscherRecordHolder{" +
"drawingGroupData=" + drawingGroupData +
", mainDocumentDgContainer=" + mainDocumentDgContainer +
", headerDocumentDgContainer=" + headerDocumentDgContainer +
'}';
}
}