moparisthebest/poi
1
0
Fork 0
Code Issues Pull Requests Releases 1 Wiki Activity

HWPF: Improve reading of auto-saved ("complex") document

Browse Source 
git-svn-id: https://svn.apache.org/repos/asf/poi/trunk@960587 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Maxim Valyanskiy 2010-07-05 12:56:02 +00:00
parent 8c42604abf
commit 78b0c18ade
3 changed files with 31 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
src/documentation/content/xdocs/status.xml
View File

@ -39,6 +39,7 @@
<action dev="POI-DEVELOPERS" type="add">49508 - Allow the addition of paragraphs to XWPF Table Cells</action>
<action dev="POI-DEVELOPERS" type="fix">49446 - Don't consider 17.16.23 field codes as properly part of the paragraph's text</action>
<action dev="POI-DEVELOPERS" type="fix">XSLFSlideShow shouldn't break on .thmx (theme) files. Support for them is still very limited though</action>
<action dev="POI-DEVELOPERS" type="fix">HWPF: Improve reading of auto-saved ("complex") documents</action>
</release>
<release version="3.7-beta1" date="2010-06-20">
<action dev="POI-DEVELOPERS" type="fix">49432 - Lazy caching of XSSFComment CTComment objects by reference, to make repeated comment searching faster</action>

2
src/scratchpad/src/org/apache/poi/hwpf/HWPFOldDocument.java
View File

@ -86,7 +86,7 @@ public class HWPFOldDocument extends HWPFDocumentCore {
TextPiece tp = new TextPiece(
0, textData.length, textData, pd, 0
);
tpt.getTextPieces().add(tp);
tpt.add(tp);
text.append(tp.getStringBuffer());
}

40
src/scratchpad/src/org/apache/poi/hwpf/model/TextPieceTable.java
View File

@ -17,14 +17,15 @@
package org.apache.poi.hwpf.model;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.List;
import org.apache.poi.hwpf.model.io.HWPFOutputStream;
import org.apache.poi.poifs.common.POIFSConstants;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Collections;
import java.util.Comparator;
import java.util.List;
/**
* The piece table for matching up character positions to bits of text. This
* mostly works in bytes, but the TextPieces themselves work in characters. This
@ -34,6 +35,7 @@ import org.apache.poi.poifs.common.POIFSConstants;
*/
public final class TextPieceTable implements CharIndexTranslator {
protected ArrayList<TextPiece> _textPieces = new ArrayList<TextPiece>();
protected ArrayList<TextPiece> _textPiecesFCOrder = new ArrayList<TextPiece>();
// int _multiple;
int _cpMin;
@ -96,11 +98,9 @@ public final class TextPieceTable implements CharIndexTranslator {
// In the interest of our sanity, now sort the text pieces
// into order, if they're not already
TextPiece[] tp = _textPieces.toArray(new TextPiece[_textPieces.size()]);
Arrays.sort(tp);
for (int i = 0; i < tp.length; i++) {
_textPieces.set(i, tp[i]);
}
Collections.sort(_textPieces);
_textPiecesFCOrder = new ArrayList<TextPiece>(_textPieces);
Collections.sort(_textPiecesFCOrder, new FCComparator());
}
public int getCpMin() {
@ -111,6 +111,13 @@ public final class TextPieceTable implements CharIndexTranslator {
return _textPieces;
}
public void add(TextPiece piece) {
_textPieces.add(piece);
_textPiecesFCOrder.add(piece);
Collections.sort(_textPieces);
Collections.sort(_textPiecesFCOrder, new FCComparator());
}
/**
* Is the text at the given Character offset unicode, or plain old ascii? In
* a very evil fashion, you have to actually know this to make sense of
@ -238,7 +245,7 @@ public final class TextPieceTable implements CharIndexTranslator {
public int getCharIndex(int bytePos) {
int charCount = 0;
for(TextPiece tp : _textPieces) {
for(TextPiece tp : _textPiecesFCOrder) {
int pieceStart = tp.getPieceDescriptor().getFilePosition();
if (pieceStart >= bytePos) {
break;
@ -259,4 +266,15 @@ public final class TextPieceTable implements CharIndexTranslator {
return charCount;
}
private static class FCComparator implements Comparator<TextPiece> {
public int compare(TextPiece textPiece, TextPiece textPiece1) {
if (textPiece.getPieceDescriptor().fc>textPiece1.getPieceDescriptor().fc) {
return 1;
} else if (textPiece.getPieceDescriptor().fc<textPiece1.getPieceDescriptor().fc) {
return -1;
} else {
return 0;
}
}
}
}