[bug-66312] partial fix for insertNewParagraph(XmlCursor cursor)

git-svn-id: https://svn.apache.org/repos/asf/poi/trunk@1904563 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
PJ Fanning 2022-10-13 10:38:31 +00:00
parent ce7c1c92da
commit ba1e8de446
7 changed files with 125 additions and 111 deletions

View File

@ -222,12 +222,12 @@ Office Open XML schemas (poi-ooxml-full-*.jar)
Furthermore, both Microsoft and Adobe have granted patent licenses
to this work [3,4,5].
[1] http://www.ecma-international.org/publications/standards/Ecma-376.htm
[2] http://www.ecma-international.org/memento/Ecmabylaws.htm
[1] https://www.ecma-international.org/publications/standards/Ecma-376.htm
[2] https://www.ecma-international.org/memento/Ecmabylaws.htm
[3] http://www.microsoft.com/openspecifications/en/us/programs/osp/default.aspx
[4] http://www.ecma-international.org/publications/files/ECMA-ST/Ecma%20PATENT/
[4] https://www.ecma-international.org/publications/files/ECMA-ST/Ecma%20PATENT/
Patent%20statements%20ok/ECMA-376%20Edition%202%20Microsoft%20Patent%20Declaration.pdf
[5] http://www.ecma-international.org/publications/files/ECMA-ST/Ecma%20PATENT/
[5] https://www.ecma-international.org/publications/files/ECMA-ST/Ecma%20PATENT/
Patent%20statements%20ok/ECMA-376%20Adobe%20Patent%20Declaration.pdf

View File

@ -56,7 +56,7 @@ import org.openxmlformats.schemas.presentationml.x2006.main.SldMasterDocument;
* If you are using these low level classes, then you
* will almost certainly need to refer to the OOXML
* specifications from
* http://www.ecma-international.org/publications/standards/Ecma-376.htm
* https://www.ecma-international.org/publications/standards/Ecma-376.htm
*
* WARNING - APIs expected to change rapidly
*/

View File

@ -96,7 +96,7 @@ import org.openxmlformats.schemas.wordprocessingml.x2006.main.StylesDocument;
* XML structure come through. You'll therefore almost
* certainly need to refer to the OOXML specifications
* from
* http://www.ecma-international.org/publications/standards/Ecma-376.htm
* https://www.ecma-international.org/publications/standards/Ecma-376.htm
* at some point in your use.
*/
@SuppressWarnings("unused")
@ -678,118 +678,102 @@ public class XWPFDocument extends POIXMLDocument implements Document, IBody {
*
* @param cursor The cursor-position where the new paragraph should be added.
* @return the {@link XWPFParagraph} object representing the newly inserted
* CTP object
* CTP object.
*/
@Override
public XWPFParagraph insertNewParagraph(XmlCursor cursor) {
if (isCursorInBody(cursor)) {
String uri = CTP.type.getName().getNamespaceURI();
/*
* TODO DO not use a coded constant, find the constant in the OOXML
* classes instead, as the child of type CT_Paragraph is defined in the
* OOXML schema as 'p'
*/
String localPart = "p";
// creates a new Paragraph, cursor is positioned inside the new
// element
cursor.beginElement(localPart, uri);
// move the cursor to the START token to the paragraph just created
cursor.toParent();
CTP p = (CTP) cursor.getObject();
XWPFParagraph newP = new XWPFParagraph(p, this);
XmlObject o = null;
/*
* move the cursor to the previous element until a) the next
* paragraph is found or b) all elements have been passed
*/
while (!(o instanceof CTP) && (cursor.toPrevSibling())) {
o = cursor.getObject();
}
/*
* if the object that has been found is a) not a paragraph or b) is
* the paragraph that has just been inserted, as the cursor in the
* while loop above was not moved as there were no other siblings,
* then the paragraph that was just inserted is the first paragraph
* in the body. Otherwise, take the previous paragraph and calculate
* the new index for the new paragraph.
*/
if ((!(o instanceof CTP)) || o == p) {
paragraphs.add(0, newP);
} else {
int pos = paragraphs.indexOf(getParagraph((CTP) o)) + 1;
paragraphs.add(pos, newP);
}
/*
* create a new cursor, that points to the START token of the just
* inserted paragraph
*/
try (XmlCursor newParaPos = p.newCursor()) {
/*
* Calculate the paragraphs index in the list of all body
* elements
*/
int i = 0;
cursor.toCursor(newParaPos);
while (cursor.toPrevSibling()) {
o = cursor.getObject();
if (o instanceof CTP || o instanceof CTTbl) {
i++;
}
}
bodyElements.add(i, newP);
cursor.toCursor(newParaPos);
cursor.toEndToken();
return newP;
}
String uri = CTP.type.getName().getNamespaceURI();
/*
* TODO DO not use a coded constant, find the constant in the OOXML
* classes instead, as the child of type CT_Paragraph is defined in the
* OOXML schema as 'p'
*/
String localPart = "p";
// creates a new Paragraph, cursor is positioned inside the new
// element
cursor.beginElement(localPart, uri);
// move the cursor to the START token to the paragraph just created
cursor.toParent();
CTP p = (CTP) cursor.getObject();
XWPFParagraph newP = new XWPFParagraph(p, this);
XmlObject o = null;
/*
* move the cursor to the previous element until a) the next
* paragraph is found or b) all elements have been passed
*/
while (!(o instanceof CTP) && (cursor.toPrevSibling())) {
o = cursor.getObject();
}
/*
* if the object that has been found is a) not a paragraph or b) is
* the paragraph that has just been inserted, as the cursor in the
* while loop above was not moved as there were no other siblings,
* then the paragraph that was just inserted is the first paragraph
* in the body. Otherwise, take the previous paragraph and calculate
* the new index for the new paragraph.
*/
if ((!(o instanceof CTP)) || o == p) {
paragraphs.add(0, newP);
} else {
int pos = paragraphs.indexOf(getParagraph((CTP) o)) + 1;
paragraphs.add(pos, newP);
}
/*
* create a new cursor, that points to the START token of the just
* inserted paragraph
*/
try (XmlCursor newParaPos = p.newCursor()) {
/*
* Calculate the paragraphs index in the list of all body
* elements
*/
int i = 0;
cursor.toCursor(newParaPos);
while (cursor.toPrevSibling()) {
o = cursor.getObject();
if (o instanceof CTP || o instanceof CTTbl) {
i++;
}
}
bodyElements.add(i, newP);
cursor.toCursor(newParaPos);
cursor.toEndToken();
return newP;
}
return null;
}
@Override
public XWPFTable insertNewTbl(XmlCursor cursor) {
if (isCursorInBody(cursor)) {
String uri = CTTbl.type.getName().getNamespaceURI();
String localPart = "tbl";
cursor.beginElement(localPart, uri);
cursor.toParent();
CTTbl t = (CTTbl) cursor.getObject();
XWPFTable newT = new XWPFTable(t, this);
XmlObject o = null;
while (!(o instanceof CTTbl) && (cursor.toPrevSibling())) {
o = cursor.getObject();
}
if (!(o instanceof CTTbl)) {
tables.add(0, newT);
} else {
int pos = tables.indexOf(getTable((CTTbl) o)) + 1;
tables.add(pos, newT);
}
int i = 0;
try (XmlCursor tableCursor = t.newCursor()) {
cursor.toCursor(tableCursor);
while (cursor.toPrevSibling()) {
o = cursor.getObject();
if (o instanceof CTP || o instanceof CTTbl) {
i++;
}
}
bodyElements.add(i, newT);
cursor.toCursor(tableCursor);
cursor.toEndToken();
return newT;
}
String uri = CTTbl.type.getName().getNamespaceURI();
String localPart = "tbl";
cursor.beginElement(localPart, uri);
cursor.toParent();
CTTbl t = (CTTbl) cursor.getObject();
XWPFTable newT = new XWPFTable(t, this);
XmlObject o = null;
while (!(o instanceof CTTbl) && (cursor.toPrevSibling())) {
o = cursor.getObject();
}
return null;
}
/**
* verifies that cursor is on the right position
*/
private boolean isCursorInBody(XmlCursor cursor) {
try (XmlCursor verify = cursor.newCursor()) {
verify.toParent();
return (verify.getObject() == this.ctDocument.getBody());
if (!(o instanceof CTTbl)) {
tables.add(0, newT);
} else {
int pos = tables.indexOf(getTable((CTTbl) o)) + 1;
tables.add(pos, newT);
}
int i = 0;
try (XmlCursor tableCursor = t.newCursor()) {
cursor.toCursor(tableCursor);
while (cursor.toPrevSibling()) {
o = cursor.getObject();
if (o instanceof CTP || o instanceof CTTbl) {
i++;
}
}
bodyElements.add(i, newT);
cursor.toCursor(tableCursor);
cursor.toEndToken();
return newT;
}
}

View File

@ -33,7 +33,7 @@ import org.openxmlformats.schemas.spreadsheetml.x2006.main.CTCols;
/**
* Test asserts the POI produces <cols> element that could be read and properly interpreted by the MS Excel.
* For specification of the "cols" element see the chapter 3.3.1.16 of the "Office Open XML Part 4 - Markup Language Reference.pdf".
* The specification can be downloaded at http://www.ecma-international.org/publications/files/ECMA-ST/Office%20Open%20XML%201st%20edition%20Part%204%20(PDF).zip.
* The specification can be downloaded at https://www.ecma-international.org/publications/files/ECMA-ST/Office%20Open%20XML%201st%20edition%20Part%204%20(PDF).zip.
*
* <p><em>
* The test saves xlsx file on a disk if the system property is set:

View File

@ -43,6 +43,9 @@ import org.apache.poi.poifs.filesystem.Ole10Native;
import org.apache.poi.poifs.filesystem.POIFSFileSystem;
import org.apache.poi.xwpf.extractor.XWPFWordExtractor;
import org.apache.poi.xwpf.usermodel.XWPFDocument;
import org.apache.poi.xwpf.usermodel.XWPFParagraph;
import org.apache.poi.xwpf.usermodel.XWPFTableCell;
import org.apache.xmlbeans.XmlCursor;
import org.apache.xmlbeans.XmlException;
import org.junit.jupiter.api.Test;
import org.openxmlformats.schemas.wordprocessingml.x2006.main.DocumentDocument;
@ -182,4 +185,31 @@ class TestXWPFBugs {
}
}
}
@Test
void insertParagraphDirectlyIntoBody() throws IOException {
try (XWPFDocument document = new XWPFDocument(samples.openResourceAsStream("bug66312.docx"))) {
XWPFParagraph paragraph = document.getParagraphArray(0);
insertParagraph(paragraph, document);
assertEquals("Hello", document.getParagraphArray(0).getText());
assertEquals("World", document.getParagraphArray(1).getText());
}
}
@Test
void insertParagraphIntoTable() throws IOException {
try (XWPFDocument document = new XWPFDocument(samples.openResourceAsStream("bug66312.docx"))) {
XWPFTableCell cell = document.getTableArray(0).getRow(0).getCell(0);
XWPFParagraph paragraph = cell.getParagraphArray(0);
insertParagraph(paragraph, document);
//TODO the issue reporter thinks that there should be 2 paragraphs (with 'Hello' and 'World' repectively).
assertEquals("World", cell.getParagraphArray(0).getText());
}
}
public static void insertParagraph(XWPFParagraph xwpfParagraph, XWPFDocument document) {
XmlCursor xmlCursor = xwpfParagraph.getCTP().newCursor();
XWPFParagraph xwpfParagraph2 = document.insertNewParagraph(xmlCursor);
xwpfParagraph2.createRun().setText("Hello");
}
}

View File

@ -560,7 +560,7 @@ public final class CryptoFunctions {
* @param password the password
* @return the ansi bytes
*
* @see <a href="http://www.ecma-international.org/news/TC45_current_work/Office%20Open%20XML%20Part%204%20-%20Markup%20Language%20Reference.pdf">Part 4 - Markup Language Reference - Ecma International - section 3.2.29 (workbookProtection)</a>
* @see <a href="https://www.ecma-international.org/news/TC45_current_work/Office%20Open%20XML%20Part%204%20-%20Markup%20Language%20Reference.pdf">Part 4 - Markup Language Reference - Ecma International - section 3.2.29 (workbookProtection)</a>
*/
private static byte[] toAnsiPassword(String password) {
// TODO: charset conversion (see ecma spec)

Binary file not shown.