/** Example of how to find an object in document.xml via traversal (as opposed to XPath) */
public static void main(String[] args) throws Exception {
String inputfilepath = System.getProperty("user.dir") + "/checkbox.docx";
WordprocessingMLPackage wordMLPackage =
WordprocessingMLPackage.load(new java.io.File(inputfilepath));
MainDocumentPart documentPart = wordMLPackage.getMainDocumentPart();
Finder finder = new Finder(FldChar.class); // <----- change this to suit
new TraversalUtil(documentPart.getContent(), finder);
System.out.println("got " + finder.results.size() + " of type " + finder.typeToFind.getName());
for (Object o : finder.results) {
Object o2 = XmlUtils.unwrap(o);
// this is ok, provided the results of the Callback
// won't be marshalled
if (o2 instanceof org.docx4j.wml.Text) {
org.docx4j.wml.Text txt = (org.docx4j.wml.Text) o2;
System.out.println(txt.getValue());
} else {
System.out.println(XmlUtils.marshaltoString(o, true, true));
}
}
}
@Override
public String getCssProperty() {
// Note regarding numbering case; handling of tab after the number:-
// We get this right in the PDF case, via setXslFOListBlock below.
// We don't attempt to get the tab right in the HTML case,
// since without some research, I don't know what markup would be required.
String prop = "position: relative; ";
BigInteger left = ((Ind) this.getObject()).getLeft();
if (left != null) {
prop = prop + composeCss(CSS_NAME, UnitsOfMeasurement.twipToBest(left.intValue()));
}
// SPEC: The firstLine and hanging attributes are mutually exclusive, if both are specified,
// then
// the firstLine value is ignored.
BigInteger firstline = ((Ind) this.getObject()).getFirstLine();
BigInteger hanging = ((Ind) this.getObject()).getHanging();
if (hanging != null) {
prop =
prop + composeCss("text-indent", "-" + UnitsOfMeasurement.twipToBest(hanging.intValue()));
} else if (firstline != null) {
prop = prop + composeCss("text-indent", UnitsOfMeasurement.twipToBest(firstline.intValue()));
}
if (left == null && firstline == null && hanging == null) {
log.debug("What to do with " + XmlUtils.marshaltoString(this.getObject(), true, true));
prop = CSS_NULL;
}
return prop;
}
private void processDescendantCondition(Object sdt, String xpathBase, int index, Tag tag) {
Condition c = null;
HashMap<String, String> map = QueryString.parseQueryString(tag.getVal(), true);
String conditionId = map.get(BINDING_ROLE_CONDITIONAL);
if (conditionId != null) {
c = ConditionsPart.getConditionById(conditions, conditionId);
if (c == null) {
log.error("Missing condition " + conditionId);
throw new InputIntegrityException("Required condition '" + conditionId + "' is missing");
}
// TODO: this code assumes the condition contains
// a simple xpath
log.debug("Using condition" + XmlUtils.marshaltoString(c, true, true));
Condition newCondition = c.repeat(xpathBase, index, conditions, xPaths);
// set sdt to use it
map.put(BINDING_ROLE_CONDITIONAL, newCondition.getId());
tag.setVal(QueryString.create(map));
}
}
public OpenDoPEHandler(WordprocessingMLPackage wordMLPackage) throws Docx4JException {
this.wordMLPackage = wordMLPackage;
if (wordMLPackage.getMainDocumentPart().getXPathsPart() == null) {
throw new Docx4JException("OpenDoPE XPaths part missing");
} else {
xPaths = wordMLPackage.getMainDocumentPart().getXPathsPart().getJaxbElement();
log.debug(XmlUtils.marshaltoString(xPaths, true, true));
}
if (wordMLPackage.getMainDocumentPart().getConditionsPart() != null) {
conditions = wordMLPackage.getMainDocumentPart().getConditionsPart().getJaxbElement();
log.debug(XmlUtils.marshaltoString(conditions, true, true));
}
if (wordMLPackage.getMainDocumentPart().getComponentsPart() != null) {
components = wordMLPackage.getMainDocumentPart().getComponentsPart().getJaxbElement();
log.debug(XmlUtils.marshaltoString(components, true, true));
}
shallowTraversor = new ShallowTraversor();
shallowTraversor.wordMLPackage = wordMLPackage;
}
protected void applyPPr(PPr pPrToApply, PPr effectivePPr) {
log.debug(
"apply "
+ XmlUtils.marshaltoString(pPrToApply, true, true)
+ "\n\r to "
+ XmlUtils.marshaltoString(effectivePPr, true, true));
if (pPrToApply == null) {
return;
}
List<Property> properties = PropertyFactory.createProperties(wordMLPackage, pPrToApply);
for (Property p : properties) {
if (p != null) {
// log.debug("applying pPr " + p.getClass().getName() );
((AbstractParagraphProperty) p)
.set(effectivePPr); // NB, this new method does not copy. TODO?
}
}
log.debug("result " + XmlUtils.marshaltoString(effectivePPr, true, true));
}
protected void init(Object docxObject) {
org.docx4j.wml.P para = null;
JAXBIntrospector inspector = Context.jc.createJAXBIntrospector();
if (docxObject == null) {; // implied ParagraphML
} else if (inspector.isElement(docxObject)) {
Object value = JAXBIntrospector.getValue(docxObject);
if (value instanceof org.docx4j.wml.P) {
para = (org.docx4j.wml.P) value;
this.isDummy = false;
} else {
// Create a dummy ParagraphML for this unsupported element
// TODO: A more informative text content in dummy ParagraphML
QName name = inspector.getElementName(docxObject);
String renderedText;
if (name != null) {
renderedText = XmlUtil.getEnclosingTagPair(name);
} else {
// Should not happen but it could.
renderedText = "<w:unknownTag></w:unknownTag>";
log.warn(
"init(): Unknown tag was detected for a JAXBElement = "
+ XmlUtils.marshaltoString(docxObject, true));
}
para = ObjectFactory.createP(renderedText);
this.isDummy = true;
}
} else {
throw new IllegalArgumentException("Unsupported Docx Object = " + docxObject);
}
initParagraphProperties(para);
initChildren(para);
}
/** @param args */
public static void main(String[] args) throws Exception {
try {
getInputFilePath(args);
} catch (IllegalArgumentException e) {
// inputfilepath = System.getProperty("user.dir") + "/sample-docs/glox/Hier2Level.glox";
inputfilepath =
System.getProperty("user.dir") + "/sample-docs/glox/Picture Organization Chart.glox";
}
GloxPackage gloxPackage = GloxPackage.load(new java.io.File(inputfilepath));
String title0 =
gloxPackage.getDiagramLayoutHeaderPart().getJaxbElement().getTitle().get(0).getVal();
System.out.println("Title: " + title0);
String desc0 =
gloxPackage.getDiagramLayoutHeaderPart().getJaxbElement().getDesc().get(0).getVal();
System.out.println("Description: " + desc0);
String layoutXml =
XmlUtils.marshaltoString(gloxPackage.getDiagramLayoutPart().getJaxbElement(), true, true);
System.out.println(layoutXml);
}
private List<Object> cloneRepeatSdt(Object sdt, String xpathBase, int numRepeats) {
List<Object> newContent = new ArrayList<Object>();
SdtPr sdtPr = getSdtPr(sdt);
log.debug(XmlUtils.marshaltoString(sdtPr, true, true));
// CTDataBinding binding =
// (CTDataBinding)XmlUtils.unwrap(sdtPr.getDataBinding());
CTDataBinding binding = sdtPr.getDataBinding();
if (binding != null) { // Shouldn't be a binding anyway
sdtPr.getRPrOrAliasOrLock().remove(binding);
}
emptyRepeatTagValue(sdtPr.getTag()); // 2012 07 15: do it to the first one
for (int i = 0; i < numRepeats; i++) {
// 2012 07 13: for "od:RptPosCon" processing to
// work (conditional inclusion dependant on position
// in repeat), we need each entry (ie including the
// original) to have the same tag (which I've changed
// to od:rptd).
if (i > 0) {
// Change ID
sdtPr.setId();
} // preserve ID on index 0, important for OpenDoPEReverter!
// Clone
newContent.add(XmlUtils.deepCopy(sdt));
}
return newContent;
}
@Override
public List<Object> apply(Object o) {
if (o instanceof P) {
currentP = createNode(document, NODE_BLOCK);
currentSpan = null;
if (tc.peek() != null) {
tc.peek().appendChild(currentP);
} else {
parentNode.appendChild(currentP);
}
pPr = ((P) o).getPPr();
currentP = handlePPr(conversionContext, pPr, false, currentP);
} else if (o instanceof org.docx4j.wml.R) {
if (!conversionContext.isInComplexFieldDefinition()) {
// Convert run to span
Element spanEl = createNode(document, NODE_INLINE);
currentSpan = spanEl;
rPr = ((R) o).getRPr();
if (rPr != null) {
handleRPr(conversionContext, pPr, rPr, currentSpan);
}
if (currentP == null) {
// Hyperlink special case
parentNode.appendChild(spanEl);
} else {
rtlAwareAppendChildToCurrentP(spanEl);
}
// To merge nested span (which we could do if there is a single child span),
// TraversalUtil Callback would need an after walk children
}
} else if (o instanceof org.docx4j.wml.FldChar) {
conversionContext.updateComplexFieldDefinition(((org.docx4j.wml.FldChar) o).getFldCharType());
} else if (o instanceof org.docx4j.wml.Text) {
if (!conversionContext.isInComplexFieldDefinition()) {
if (currentSpan == null) {
// eg after <br/>
log.error("null currentSpan! " + ((Text) o).getValue());
Element spanEl = createNode(document, NODE_INLINE);
if (currentP == null) {
// Hyperlink special case
parentNode.appendChild(spanEl);
} else {
currentP.appendChild(spanEl);
}
currentSpan = spanEl;
}
log.debug(((Text) o).getValue());
DocumentFragment df =
(DocumentFragment)
conversionContext.getRunFontSelector().fontSelector(pPr, rPr, ((Text) o));
XmlUtils.treeCopy(df, currentSpan);
// TODO would be more efficient without the treeCopy
// but fontSelector would need to be refactored a bit
}
} else if (o instanceof org.docx4j.wml.R.Tab) {
convertTabToNode(conversionContext, document);
} else if (o instanceof org.docx4j.wml.CTSimpleField) {
convertToNode(
conversionContext, o, AbstractFldSimpleWriter.WRITER_ID, document, getCurrentParent());
} else if (o instanceof org.docx4j.wml.P.Hyperlink) {
convertToNode(
conversionContext, o, AbstractHyperlinkWriter.WRITER_ID, document, getCurrentParent());
} else if (o instanceof org.docx4j.wml.CTBookmark) {
convertToNode(
conversionContext,
o,
AbstractBookmarkStartWriter.WRITER_ID,
document,
getCurrentParent());
} else if (o instanceof org.docx4j.wml.Tbl) {
convertToNode(
conversionContext,
o,
AbstractTableWriter.WRITER_ID,
document,
(currentP != null ? currentP : parentNode));
currentP = null;
currentSpan = null;
} else if (o instanceof org.docx4j.wml.Tr) {
// done in walkJAXBElements
// tr = document.createElementNS(Namespaces.NS_WORD12, "tr");
// //parentNode is in this case the DocumentFragment, that get's passed
// //to the TableModel/TableModelWriter
// parentNode.appendChild(tr);
} else if (o instanceof org.docx4j.wml.Tc) {
// done in walkJAXBElements
// tc = document.createElementNS(Namespaces.NS_WORD12, "tc");
// tr.appendChild(tc);
// // now the html p content will go temporarily go in w:tc,
// // which is what we need for our existing table model.
// System.out.println("#wrapped in w:tc OK");
} else if (o instanceof org.docx4j.dml.wordprocessingDrawing.Inline
|| o instanceof org.docx4j.dml.wordprocessingDrawing.Anchor) {
anchorOrInline = o; // keep this until we handle CTBlip
} else if (o instanceof org.docx4j.dml.CTBlip) {
/*<w:drawing>
<wp:inline distT="0" distB="0" distL="0" distR="0">
<a:graphic xmlns:a="http://schemas.openxmlformats.org/drawingml/2006/main">
<a:graphicData uri="http://schemas.openxmlformats.org/drawingml/2006/picture">
<pic:pic xmlns:pic="http://schemas.openxmlformats.org/drawingml/2006/picture">
<pic:blipFill>
<a:blip r:embed="rId10" cstate="print"/> */
DocumentFragment foreignFragment = createImage(IMAGE_E20, conversionContext, anchorOrInline);
anchorOrInline = null;
currentP.appendChild(document.importNode(foreignFragment, true));
} else if (o instanceof org.docx4j.wml.Pict) {
/*<w:pict>
<v:shape id="_x0000_i1025" type="#_x0000_t75" style="width:428.25pt;height:321pt">
<v:imagedata r:id="rId4" o:title=""/>
</v:shape> */
org.docx4j.vml.CTTextbox textBox = getTextBox((org.docx4j.wml.Pict) o);
if (textBox == null) {
// Assume it contains an image!
DocumentFragment foreignFragment = createImage(IMAGE_E10, conversionContext, o);
currentP.appendChild(document.importNode(foreignFragment, true));
} else {
convertToNode(
conversionContext, o, AbstractPictWriter.WRITER_ID, document, getCurrentParent());
}
} else if (o instanceof Br) {
handleBr((Br) o);
} else if (o instanceof org.docx4j.wml.R.Sym) {
convertToNode(
conversionContext, o, AbstractSymbolWriter.WRITER_ID, document, getCurrentParent());
} else if ((o instanceof org.docx4j.wml.ProofErr)
|| (o instanceof org.docx4j.wml.R.LastRenderedPageBreak)
|| (o instanceof org.docx4j.wml.CTMarkupRange)) {
// Ignore theese types, they don't need to be outputed/handled
// CTMarkupRange is the w:bookmarkEnd
} else {
getLog().warn("Need to handle " + o.getClass().getName());
log.debug(XmlUtils.marshaltoString(o));
}
return null;
}
@Override
public void apply(SdtElement element, Object parent, List<Object> siblings) {
System.out.println();
SdtPr sdtPr = element.getSdtPr();
if (sdtPr == null) {
System.out.println(
callback.indent
+ element.getClass().getSimpleName()
+ " [no sdtPr!]"
+ " (having parent "
+ parent.getClass().getSimpleName()
+ ")");
} else {
System.out.println(
callback.indent
+ element.getClass().getSimpleName()
+ " (having parent "
+ parent.getClass().getSimpleName()
+ ")");
CTDataBinding binding = (CTDataBinding) XmlUtils.unwrap(sdtPr.getDataBinding());
if (binding != null) {
System.out.println(callback.indent + " binding: " + binding.getXpath());
}
Tag tag = sdtPr.getTag();
if (tag == null) return;
System.out.println(callback.indent + " " + tag.getVal());
HashMap<String, String> map = QueryString.parseQueryString(tag.getVal(), true);
String conditionId = map.get(OpenDoPEHandler.BINDING_ROLE_CONDITIONAL);
String repeatId = map.get(OpenDoPEHandler.BINDING_ROLE_REPEAT);
String xp = map.get(OpenDoPEHandler.BINDING_ROLE_XPATH);
if (conditionId != null) {
Condition c = ConditionsPart.getConditionById(conditions, conditionId);
if (c == null) {
System.out.println(callback.indent + " " + "Missing condition " + conditionId);
}
if (c.getParticle() instanceof org.opendope.conditions.Xpathref) {
org.opendope.conditions.Xpathref xpathRef = (Xpathref) c.getParticle();
if (xpathRef == null) {
System.out.println(
callback.indent
+ " "
+ "Condition "
+ c.getId()
+ " references a missing xpath!");
}
org.opendope.xpaths.Xpaths.Xpath xpath =
XPathsPart.getXPathById(xPaths, xpathRef.getId());
if (xpath == null) {
System.out.println(
callback.indent
+ " "
+ "XPath specified in condition '"
+ c.getId()
+ "' is missing!");
} else {
System.out.println(
callback.indent
+ " "
+ xpath.getId()
+ ": "
+ xpath.getDataBinding().getXpath());
}
} else {
System.out.println("Complex condition: " + XmlUtils.marshaltoString(c, true, true));
}
} else if (repeatId != null) {
org.opendope.xpaths.Xpaths.Xpath xpath = XPathsPart.getXPathById(xPaths, repeatId);
if (xpath == null) {
System.out.println(
callback.indent
+ " "
+ "XPath specified in repeat '"
+ repeatId
+ "' is missing!");
} else {
System.out.println(
callback.indent + " " + xpath.getId() + ": " + xpath.getDataBinding().getXpath());
}
} else if (xp != null) {
org.opendope.xpaths.Xpaths.Xpath xpath = XPathsPart.getXPathById(xPaths, xp);
if (xpath == null) {
System.out.println(
callback.indent + " " + "XPath specified with id '" + xp + "' is missing!");
} else {
System.out.println(
callback.indent + " " + xpath.getId() + ": " + xpath.getDataBinding().getXpath());
}
}
}
}
/** @param args */
public static void main(String[] args) throws Exception {
String inputfilepath =
System.getProperty("user.dir") + "/sample-docs/word/databinding/invoice2.docx";
WordprocessingMLPackage wordMLPackage =
WordprocessingMLPackage.load(new java.io.File(inputfilepath));
filepathprefix = inputfilepath.substring(0, inputfilepath.lastIndexOf("."));
System.out.println(filepathprefix);
StringBuilder timingSummary = new StringBuilder();
// Process conditionals and repeats
long startTime = System.currentTimeMillis();
OpenDoPEHandler odh = new OpenDoPEHandler(wordMLPackage);
odh.preprocess();
long endTime = System.currentTimeMillis();
timingSummary.append("OpenDoPEHandler: " + (endTime - startTime));
System.out.println(
XmlUtils.marshaltoString(wordMLPackage.getMainDocumentPart().getJaxbElement(), true, true));
SaveToZipFile saver = new SaveToZipFile(wordMLPackage);
saver.save(filepathprefix + "_1_preprocessed.docx");
System.out.println("Saved: " + filepathprefix + "_1_preprocessed.docx");
startTime = System.currentTimeMillis();
OpenDoPEIntegrity odi = new OpenDoPEIntegrity();
odi.process(wordMLPackage);
endTime = System.currentTimeMillis();
timingSummary.append("\nOpenDoPEIntegrity: " + (endTime - startTime));
System.out.println(
XmlUtils.marshaltoString(wordMLPackage.getMainDocumentPart().getJaxbElement(), true, true));
saver = new SaveToZipFile(wordMLPackage);
saver.save(filepathprefix + "_2_integrity.docx");
System.out.println("Saved: " + filepathprefix + "_2_integrity.docx");
// Apply the bindings
BindingHandler.setHyperlinkStyle("Hyperlink");
startTime = System.currentTimeMillis();
BindingHandler.applyBindings(wordMLPackage.getMainDocumentPart());
endTime = System.currentTimeMillis();
timingSummary.append("\nBindingHandler.applyBindings: " + (endTime - startTime));
System.out.println(
XmlUtils.marshaltoString(wordMLPackage.getMainDocumentPart().getJaxbElement(), true, true));
saver.save(filepathprefix + "_3_bound.docx");
System.out.println("Saved: " + filepathprefix + "_3_bound.docx");
// Either demonstrate reverter, or stripping of controls;
// you can't do both. So comment out one or the other.
// reverter(inputfilepath, filepathprefix + "_bound.docx");
//
// Strip content controls
startTime = System.currentTimeMillis();
RemovalHandler rh = new RemovalHandler();
rh.removeSDTs(wordMLPackage, Quantifier.ALL);
endTime = System.currentTimeMillis();
timingSummary.append("\nRemovalHandler: " + (endTime - startTime));
saver.save(filepathprefix + "_4_stripped.docx");
System.out.println("Saved: " + filepathprefix + "_4_stripped.docx");
System.out.println(timingSummary);
}
public static DocumentFragment createBlockForP(
SvgConversionContext context,
String lvl,
String cNvPrName,
String phType,
NodeIterator childResults,
NodeIterator lvlNpPr) {
StyleTree styleTree = null;
try {
styleTree = context.getPmlPackage().getStyleTree();
} catch (InvalidFormatException e1) {
// TODO Auto-generated catch block
e1.printStackTrace();
}
log.debug("lvl:" + lvl);
int level;
if (lvl.equals("NaN")) {
level = 1;
} else {
level = Integer.parseInt(lvl);
}
String pStyleVal;
System.out.println("cNvPrName: " + cNvPrName + "; " + "phType: " + phType);
if (cNvPrName.toLowerCase().indexOf("subtitle") > -1
|| phType.toLowerCase().indexOf("subtitle") > -1) {
// Subtitle on first page in default layout is styled as a Body.
pStyleVal = "Lvl" + level + "Master" + context.getResolvedLayout().getMasterNumber() + "Body";
} else if (cNvPrName.toLowerCase().indexOf("title") > -1
|| phType.toLowerCase().indexOf("title") > -1) {
pStyleVal =
"Lvl" + level + "Master" + context.getResolvedLayout().getMasterNumber() + "Title";
} else {
// eg cNvPrName: TextBox 2; phType:
pStyleVal =
"Lvl" + level + "Master" + context.getResolvedLayout().getMasterNumber() + "Other";
}
System.out.println("--> " + pStyleVal);
try {
// Create a DOM builder and parse the fragment
Document document = XmlUtils.getNewDocumentBuilder().newDocument();
// log.info("Document: " + document.getClass().getName() );
Node xhtmlP = document.createElement("p");
document.appendChild(xhtmlP);
// Set @class
log.debug(pStyleVal);
Tree<AugmentedStyle> pTree = styleTree.getParagraphStylesTree();
org.docx4j.model.styles.Node<AugmentedStyle> asn = pTree.get(pStyleVal);
((Element) xhtmlP).setAttribute("class", StyleTree.getHtmlClassAttributeValue(pTree, asn));
StringBuilder inlineStyle = new StringBuilder();
// Do we have CTTextParagraphProperties
// <a:lvl?pPr>
// Convert it to a WordML pPr
CTTextParagraphProperties lvlPPr = unmarshalFormatting(lvlNpPr);
if (lvlPPr != null) {
log.debug("We have lvlPPr");
log.debug(
XmlUtils.marshaltoString(
lvlPPr,
true,
true,
Context.jcPML,
"FIXME",
"lvl1pPr",
CTTextParagraphProperties.class));
PPr pPr = TextStyles.getWmlPPr(lvlPPr);
if (pPr != null) {
HtmlCssHelper.createCss(context.getPmlPackage(), pPr, inlineStyle, false, false);
}
// TODO RPR
}
// Without this, top-margin is too large in Webkit (Midor).
// Not tested elsewhere...
inlineStyle.append("margin-left:3px; margin-top:3px;");
if (!inlineStyle.toString().equals("")) {
((Element) xhtmlP).setAttribute("style", inlineStyle.toString());
}
// Our fo:block wraps whatever result tree fragment
// our style sheet produced when it applied-templates
// to the child nodes
// init
Node n = childResults.nextNode();
do {
// getNumberXmlNode creates a span node, which is empty
// if there is no numbering.
// Let's get rid of any such <span/>.
// What we actually get is a document node
if (n.getNodeType() == Node.DOCUMENT_NODE) {
log.debug("handling DOCUMENT_NODE");
// Do just enough of the handling here
NodeList nodes = n.getChildNodes();
if (nodes != null) {
for (int i = 0; i < nodes.getLength(); i++) {
if (((Node) nodes.item(i)).getLocalName().equals("span")
&& !((Node) nodes.item(i)).hasChildNodes()) {
// ignore
log.debug(".. ignoring <span/> ");
} else {
XmlUtils.treeCopy((Node) nodes.item(i), xhtmlP);
}
}
}
} else {
// log.info("Node we are importing: " + n.getClass().getName() );
// foBlockElement.appendChild(
// document.importNode(n, true) );
/*
* Node we'd like to import is of type org.apache.xml.dtm.ref.DTMNodeProxy
* which causes
* org.w3c.dom.DOMException: NOT_SUPPORTED_ERR: The implementation does not support the requested type of object or operation.
*
* See http://osdir.com/ml/text.xml.xerces-j.devel/2004-04/msg00066.html
*
* So instead of importNode, use
*/
XmlUtils.treeCopy(n, xhtmlP);
}
// next
n = childResults.nextNode();
} while (n != null);
DocumentFragment docfrag = document.createDocumentFragment();
docfrag.appendChild(document.getDocumentElement());
return docfrag;
} catch (Exception e) {
log.error(e.getMessage(), e);
}
return null;
}