/**
* The dummy section wrappers only contains one section with all the document. Therefore any
* sections within the document are ignored in the conversion process. As it doesn't need to check
* for sections it is faster and the html-Output only uses one section.<br>
* It will use the Header/Footer of the body sectPr. This isn't correct, if there are several
* Sections in the document, but to find the correct SectPr it would need to check the document
* content - and the aim of this method is a low overhead.
*
* @param wmlPackage
* @param dummyPageNumbering
* @return
*/
protected static List<ConversionSectionWrapper> processDummy(
WordprocessingMLPackage wmlPackage,
Document document,
RelationshipsPart rels,
BooleanDefaultTrue evenAndOddHeaders,
boolean dummyPageNumbering) {
List<ConversionSectionWrapper> conversionSections = new ArrayList<ConversionSectionWrapper>();
ConversionSectionWrapper currentSectionWrapper = null;
HeaderFooterPolicy previousHF =
new HeaderFooterPolicy(document.getBody().getSectPr(), null, rels, evenAndOddHeaders);
// SectionWrapper does work where sectPr is null (ie document has no body level sectPr),
// so document.getBody().getSectPr() is ok
currentSectionWrapper =
createSectionWrapper(
document.getBody().getSectPr(),
previousHF,
rels,
evenAndOddHeaders,
1,
document.getBody().getContent(),
dummyPageNumbering);
conversionSections.add(currentSectionWrapper);
return conversionSections;
}
/**
* Traverse the document, and return a map of all styles which are used directly in the document.
* (IE this does not include styles on which others are just BasedOn).
*
* @return
*/
public Set<String> getStylesInUse() {
org.docx4j.wml.Document wmlDocumentEl = (org.docx4j.wml.Document) this.getJaxbElement();
Body body = wmlDocumentEl.getBody();
List<Object> bodyChildren = body.getContent();
Set<String> stylesInUse = new HashSet<String>();
FontAndStyleFinder finder = new FontAndStyleFinder(null, null, stylesInUse);
finder.defaultCharacterStyle = this.getStyleDefinitionsPart().getDefaultCharacterStyle();
finder.defaultParagraphStyle = this.getStyleDefinitionsPart().getDefaultParagraphStyle();
new TraversalUtil(bodyChildren, finder);
finder.finish();
// Styles in headers, footers?
RelationshipsPart rp = this.getRelationshipsPart();
if (rp != null) {
for (Relationship r : rp.getRelationships().getRelationship()) {
Part part = rp.getPart(r);
if (part instanceof FooterPart) {
Ftr ftr = ((FooterPart) part).getJaxbElement();
finder.walkJAXBElements(ftr);
} else if (part instanceof HeaderPart) {
Hdr hdr = ((HeaderPart) part).getJaxbElement();
finder.walkJAXBElements(hdr);
}
}
}
// Styles in endnotes, footnotes?
if (this.getEndNotesPart() != null) {
log.debug("Looking at endnotes");
CTEndnotes endnotes = this.getEndNotesPart().getJaxbElement();
finder.walkJAXBElements(endnotes);
}
if (this.getFootnotesPart() != null) {
log.debug("Looking at footnotes");
CTFootnotes footnotes = this.getFootnotesPart().getJaxbElement();
finder.walkJAXBElements(footnotes);
}
// Comments
if (this.getCommentsPart() != null) {
log.debug("Looking at comments");
Comments comments = this.getCommentsPart().getJaxbElement();
finder.walkJAXBElements(comments);
}
return stylesInUse;
}
public static void main(String[] args) throws Exception {
String inputfilepath = System.getProperty("user.dir") + "/sample-docs/sample-docx.xml";
WordprocessingMLPackage wordMLPackage =
WordprocessingMLPackage.load(new java.io.File(inputfilepath));
MainDocumentPart documentPart = wordMLPackage.getMainDocumentPart();
org.docx4j.wml.Document wmlDocumentEl = (org.docx4j.wml.Document) documentPart.getJaxbElement();
Body body = wmlDocumentEl.getBody();
new TraversalUtil(
body,
new Callback() {
String indent = "";
@Override
public List<Object> apply(Object o) {
String text = "";
if (o instanceof org.docx4j.wml.Text) text = ((org.docx4j.wml.Text) o).getValue();
System.out.println(indent + o.getClass().getName() + " \"" + text + "\"");
return null;
}
@Override
public boolean shouldTraverse(Object o) {
return true;
}
// Depth first
@Override
public void walkJAXBElements(Object parent) {
indent += " ";
List children = getChildren(parent);
if (children != null) {
for (Object o : children) {
// if its wrapped in javax.xml.bind.JAXBElement, get its
// value; this is ok, provided the results of the Callback
// won't be marshalled
o = XmlUtils.unwrap(o);
this.apply(o);
if (this.shouldTraverse(o)) {
walkJAXBElements(o);
}
}
}
indent = indent.substring(0, indent.length() - 4);
}
@Override
public List<Object> getChildren(Object o) {
return TraversalUtil.getChildrenImpl(o);
}
});
}
/**
* Traverse the document, looking for fonts which have been applied, either directly, or via a
* style.
*
* @return
*/
public Set<String> fontsInUse() {
log.info("fontsInUse..");
getPropertyResolver(); // this inits our virtual DocDefaults style
// Setup
Set<String> fontsDiscovered = new java.util.HashSet<String>();
// // Keep track of styles we encounter, so we can
// // inspect these for fonts
// Set<String> stylesInUse = new java.util.HashSet<String>();
//
// org.docx4j.wml.Styles styles = null;
// if (this.getStyleDefinitionsPart()!=null) {
// styles = (org.docx4j.wml.Styles)this.getStyleDefinitionsPart().getJaxbElement();
// }
// // It is convenient to have a HashMap of styles
// Map<String, Style> stylesDefined = new java.util.HashMap<String, Style>();
// if (styles!=null) {
// for (Iterator<Style> iter = styles.getStyle().iterator(); iter.hasNext();) {
// Style s = iter.next();
// stylesDefined.put(s.getStyleId(), s);
// }
// }
// // We need to know what fonts and styles are used in the document
org.docx4j.wml.Document wmlDocumentEl = (org.docx4j.wml.Document) this.getJaxbElement();
Body body = wmlDocumentEl.getBody();
List<Object> bodyChildren = body.getContent();
FontDiscoveryCharacterVisitor visitor = new FontDiscoveryCharacterVisitor(fontsDiscovered);
RunFontSelector runFontSelector =
new RunFontSelector(
(WordprocessingMLPackage) this.pack, visitor, RunFontActionType.DISCOVERY);
FontAndStyleFinder finder = new FontAndStyleFinder(runFontSelector, fontsDiscovered, null);
finder.defaultCharacterStyle = this.getStyleDefinitionsPart().getDefaultCharacterStyle();
finder.defaultParagraphStyle = this.getStyleDefinitionsPart().getDefaultParagraphStyle();
new TraversalUtil(bodyChildren, finder);
// finder.finish();
fontsDiscovered.add(runFontSelector.getDefaultFont());
// fonts in headers, footers?
RelationshipsPart rp = this.getRelationshipsPart();
if (rp != null) {
for (Relationship r : rp.getRelationships().getRelationship()) {
Part part = rp.getPart(r);
if (part instanceof FooterPart) {
Ftr ftr = ((FooterPart) part).getJaxbElement();
finder.walkJAXBElements(ftr);
} else if (part instanceof HeaderPart) {
Hdr hdr = ((HeaderPart) part).getJaxbElement();
finder.walkJAXBElements(hdr);
}
}
}
// Styles in endnotes, footnotes?
if (this.getEndNotesPart() != null) {
log.debug("Looking at endnotes");
CTEndnotes endnotes = this.getEndNotesPart().getJaxbElement();
finder.walkJAXBElements(endnotes);
}
if (this.getFootnotesPart() != null) {
log.debug("Looking at footnotes");
CTFootnotes footnotes = this.getFootnotesPart().getJaxbElement();
finder.walkJAXBElements(footnotes);
}
// Comments
if (this.getCommentsPart() != null) {
log.debug("Looking at comments");
Comments comments = this.getCommentsPart().getJaxbElement();
finder.walkJAXBElements(comments);
}
// Add fonts used in the styles we discovered
// .. 2013 03 10: no longer necessary
// Fonts can also be used in the numbering part
// For now, treat any font mentioned in that part as in use.
// Ideally, we'd only register fonts used in numbering levels
// that were actually used in the document
if (getNumberingDefinitionsPart() != null) {
Numbering numbering = getNumberingDefinitionsPart().getJaxbElement();
for (Numbering.AbstractNum abstractNumNode : numbering.getAbstractNum()) {
for (Lvl lvl : abstractNumNode.getLvl()) {
if (lvl.getRPr() != null && lvl.getRPr().getRFonts() != null) {
String fontName = lvl.getRPr().getRFonts().getAscii();
if (fontName != null) {
fontsDiscovered.add(fontName);
log.debug(
"Registered "
+ fontName
+ " for abstract list "
+ abstractNumNode.getAbstractNumId()
+ " lvl "
+ lvl.getIlvl());
}
}
}
}
}
return fontsDiscovered;
}
public static void main(String[] args) throws Exception {
WordprocessingMLPackage wordMLPackage = WordprocessingMLPackage.createPackage();
MainDocumentPart documentPart = wordMLPackage.getMainDocumentPart();
CTSettings ct = new CTSettings();
DocumentSettingsPart dsp = documentPart.getDocumentSettingsPart();
if (dsp == null) {
dsp = new DocumentSettingsPart();
CTView ctView = Context.getWmlObjectFactory().createCTView();
ctView.setVal(STView.PRINT);
ct.setView(ctView);
BooleanDefaultTrue b = new BooleanDefaultTrue();
b.setVal(true);
ct.setUpdateFields(b);
dsp.setJaxbElement(ct);
documentPart.addTargetPart(dsp);
}
org.docx4j.wml.Document wmlDocumentEl = (org.docx4j.wml.Document) documentPart.getJaxbElement();
Body body = wmlDocumentEl.getBody();
ObjectFactory factory = Context.getWmlObjectFactory();
/*
* Create the following:
*
* <w:p> <w:r> <w:fldChar w:dirty="true" w:fldCharType="begin"/>
* <w:instrText xml:space="preserve">TOC \o "1-3" \h \z \ u
* \h</w:instrText> </w:r> <w:r/> <w:r> <w:fldChar w:fldCharType="end"/>
* </w:r> </w:p>
*/
P paragraphForTOC = factory.createP();
R r = factory.createR();
FldChar fldchar = factory.createFldChar();
fldchar.setFldCharType(STFldCharType.BEGIN);
fldchar.setDirty(true);
r.getContent().add(getWrappedFldChar(fldchar));
paragraphForTOC.getContent().add(r);
R r1 = factory.createR();
Text txt = new Text();
txt.setSpace("preserve");
txt.setValue("TOC \\o \"1-3\" \\h \\z \\u ");
r.getContent().add(factory.createRInstrText(txt));
paragraphForTOC.getContent().add(r1);
FldChar fldcharend = factory.createFldChar();
fldcharend.setFldCharType(STFldCharType.END);
R r2 = factory.createR();
r2.getContent().add(getWrappedFldChar(fldcharend));
paragraphForTOC.getContent().add(r2);
body.getContent().add(paragraphForTOC);
documentPart.addStyledParagraphOfText("Heading1", "Hello 1");
documentPart.addStyledParagraphOfText("Heading2", "Hello 2");
documentPart.addStyledParagraphOfText("Heading3", "Hello 3");
documentPart.addStyledParagraphOfText("Heading1", "Hello 1");
wordMLPackage.save(
new java.io.File(System.getProperty("user.dir") + "/OUT_TableOfContentsAdd.docx"));
}
protected static List<ConversionSectionWrapper> processComplete(
WordprocessingMLPackage wmlPackage,
Document document,
RelationshipsPart rels,
BooleanDefaultTrue evenAndOddHeaders,
boolean dummyPageNumbering) {
List<ConversionSectionWrapper> conversionSections = new ArrayList<ConversionSectionWrapper>();
List<Object> sectionContent = new ArrayList<Object>();
ConversionSectionWrapper currentSectionWrapper = null;
HeaderFooterPolicy previousHF = null;
int conversionSectionIndex = 0;
// According to the ECMA-376 2ed, if type is not specified, read it as next page
// However Word 2007 sometimes treats it as continuous, and sometimes doesn't??
// 20130216 Review above comment: ! In the Word UI, the Word "continuous" is shown where it is
// effective.
// In the XML, it is stored in the next following sectPr.
// First, remove content controls,
// since the P could be in a content control.
// (It is easier to remove content controls, than
// to make the code below TraversalUtil based)
// RemovalHandler is an XSLT-based way of doing this,
// but here we avoid introducing a dependency on
// XSLT (Xalan) for PDF output.
SdtBlockFinder sbr = new SdtBlockFinder();
new TraversalUtil(document.getContent(), sbr);
for (int i = sbr.sdtBlocks.size() - 1; i >= 0; i--) {
// Have to process in reverse order
// so that parentList is correct for nested sdt
SdtBlock sdtBlock = sbr.sdtBlocks.get(i);
List<Object> parentList = null;
if (sdtBlock.getParent() instanceof ArrayList) {
parentList = (ArrayList) sdtBlock.getParent();
} else {
log.error("Handle " + sdtBlock.getParent().getClass().getName());
}
int index = parentList.indexOf(sdtBlock);
parentList.remove(index);
parentList.addAll(index, sdtBlock.getSdtContent().getContent());
}
// if (log.isDebugEnabled()) {
// log.debug(XmlUtils.marshaltoString(document, true, true));
// }
// Make a list, so it is easy to look at the following sectPr,
// which we need to do to handle continuous sections properly
List<SectPr> sectPrs = new ArrayList<SectPr>();
for (Object o : document.getBody().getContent()) {
if (o instanceof org.docx4j.wml.P) {
if (((org.docx4j.wml.P) o).getPPr() != null) {
org.docx4j.wml.PPr ppr = ((org.docx4j.wml.P) o).getPPr();
if (ppr.getSectPr() != null) {
sectPrs.add(ppr.getSectPr());
}
}
}
}
if (document.getBody().getSectPr() != null) {
// usual case
sectPrs.add(document.getBody().getSectPr());
} else {
log.debug("No body level sectPr in document");
// OK if the last object is w:p and it contains a sectPr.
List<Object> all = document.getBody().getContent();
Object last = all.get(all.size() - 1);
if (last instanceof P
&& ((P) last).getPPr() != null
&& ((P) last).getPPr().getSectPr() != null) {
// ok
log.debug(
".. but last p contains sectPr .. move it"); // so our assumption later about there
// being a following section is correct
SectPr thisSectPr = ((P) last).getPPr().getSectPr();
document.getBody().setSectPr(thisSectPr);
((P) last).getPPr().setSectPr(null);
sectPrs.remove(thisSectPr);
} else {
document.getBody().setSectPr(Context.getWmlObjectFactory().createSectPr());
sectPrs.add(document.getBody().getSectPr());
}
}
int sectPrIndex = 0; // includes continuous ones
for (Object o : document.getBody().getContent()) {
if (o instanceof org.docx4j.wml.P) {
if (((org.docx4j.wml.P) o).getPPr() != null) {
org.docx4j.wml.PPr ppr = ((org.docx4j.wml.P) o).getPPr();
if (ppr.getSectPr() != null) {
// If the *following* section is continuous, don't add *this* section
boolean ignoreThisSection = false;
SectPr followingSectPr = sectPrs.get(++sectPrIndex);
if (followingSectPr.getType() != null
&& followingSectPr.getType().getVal().equals("continuous")) {
ignoreThisSection = true;
// If the w:pgSz on the two sections differs,
// then Word inserts a page break (ie doesn't treat it as continuous).
// If no w:pgSz element is present, then Word defaults
// (presumably to Legal? TODO CHECK. There is no default setting in the docx).
// Word always inserts a w:pgSz element?
PgSz pgSzThis = ppr.getSectPr().getPgSz();
PgSz pgSzNext = followingSectPr.getPgSz();
if (pgSzThis != null && pgSzNext != null) {
if (pgSzThis.getH().compareTo(pgSzNext.getH()) != 0) {
ignoreThisSection = false;
}
if (pgSzThis.getW().compareTo(pgSzNext.getW()) != 0) {
ignoreThisSection = false;
}
// Orientation:default is portrait
boolean portraitThis = true;
if (pgSzThis.getOrient() != null) {
portraitThis = pgSzThis.getOrient().equals(STPageOrientation.PORTRAIT);
}
boolean portraitNext = true;
if (pgSzNext.getOrient() != null) {
portraitNext = pgSzNext.getOrient().equals(STPageOrientation.PORTRAIT);
}
if (portraitThis != portraitNext) {
ignoreThisSection = false;
}
}
// TODO: handle cases where one or both pgSz elements are missing,
// or H or W is missing.
// Treat pgSz element missing as Legal size?
}
if (ignoreThisSection) {
// In case there are some headers/footers that get inherited by the next section
previousHF =
new HeaderFooterPolicy(ppr.getSectPr(), previousHF, rels, evenAndOddHeaders);
} else {
currentSectionWrapper =
createSectionWrapper(
ppr.getSectPr(),
previousHF,
rels,
evenAndOddHeaders,
++conversionSectionIndex,
sectionContent,
dummyPageNumbering);
conversionSections.add(currentSectionWrapper);
previousHF = currentSectionWrapper.getHeaderFooterPolicy();
sectionContent = new ArrayList<Object>();
}
}
}
}
sectionContent.add(o);
// System.out.println(XmlUtils.marshaltoString(o, true));
}
currentSectionWrapper =
createSectionWrapper(
document.getBody().getSectPr(),
previousHF,
rels,
evenAndOddHeaders,
++conversionSectionIndex,
sectionContent,
dummyPageNumbering);
conversionSections.add(currentSectionWrapper);
return conversionSections;
}