org.docx4j.TraversalUtil Java Examples

public static void replaceRichText(WordprocessingMLPackage wordMLPackage, Map<String, String> richTextMap) throws Docx4JException, JAXBException {
	MainDocumentPart document = wordMLPackage.getMainDocumentPart();
	Map<String, List<Object>> textNodeMap = new HashMap<String, List<Object>>();
	findRichTextNode(textNodeMap, document.getContents().getBody(), null);
	Iterator<String> iterator = richTextMap.keySet().iterator();
	while (iterator.hasNext()) {
		String textTag = iterator.next();
		List<Object> textNodeList = textNodeMap.get(textTag);
		if (textNodeList != null && richTextMap.containsKey(textTag)) {
			List<Object> textObjList = convertToWmlObject(wordMLPackage, richTextMap.get(textTag));
			for (int i = 0, iSize = textNodeList.size(); i < iSize; i++) {
				Object nodeObject = textNodeList.get(i);
				if (nodeObject != null) {
					//setWmlPprSetting(textNodeList.get(i), textObjList);
					TraversalUtil.replaceChildren(nodeObject , textObjList);
				}
			}
		}
	}
}
 

protected void findNameForCurrentTable(final TableManager currentTable) {
    new TraversalUtil(currentTable.firstRow,
            new RegexpFinder<P>(docxFormatter, AbstractFormatter.BAND_NAME_DECLARATION_PATTERN, P.class) {
                @Override
                protected void onFind(P paragraph, Matcher matcher) {
                    if (currentTable.bandName == null) {
                        super.onFind(paragraph, matcher);
                        currentTable.bandName = matcher.group(1);
                        String bandNameDeclaration = matcher.group();
                        Set<Text> mergedTexts = new TextMerger(paragraph, bandNameDeclaration).mergeMatchedTexts();
                        for (Text text : mergedTexts) {
                            text.setValue(text.getValue().replace(bandNameDeclaration, ""));
                        }
                    }
                }
            });
}
 

protected Part resolveTextPartForDOCX(Text text, WordprocessingMLPackage wordPackage) {
    java.util.List<SectionWrapper> sectionWrappers = wordPackage.getDocumentModel().getSections();
    for (SectionWrapper sw : sectionWrappers) {
        HeaderFooterPolicy hfp = sw.getHeaderFooterPolicy();
        List<Part> parts = Arrays.asList(hfp.getFirstHeader(), hfp.getDefaultHeader(), hfp.getEvenHeader(),
                hfp.getFirstFooter(), hfp.getDefaultFooter(), hfp.getEvenFooter());
        for (Part part : parts) {
            TextMatchCallback callback = new TextMatchCallback(text);
            new TraversalUtil(part, callback);
            if (callback.matched) {
                return part;
            }
        }
    }
    return wordPackage.getMainDocumentPart();
}
 

public void removeAllComment(String filePath, String savePath)  
        throws Exception {  
    WordprocessingMLPackage wordMLPackage = WordprocessingMLPackage  
            .load(new java.io.File(filePath));  
    //清空comments.xml内容   
    Parts parts = wordMLPackage.getParts();  
    HashMap<PartName, Part> partMap = parts.getParts();  
    CommentsPart commentPart = (CommentsPart) partMap.get(new PartName(  
            "/word/comments.xml"));  
    Comments comments = commentPart.getContents();  
    List<Comment> commentList = comments.getComment();  
    for (int i = 0, len = commentList.size(); i < len; i++) {  
        commentList.remove(0);  
    }  
      
    //清空document.xml文件中批注  
    MainDocumentPart documentPart = wordMLPackage.getMainDocumentPart();  
    org.docx4j.wml.Document wmlDocumentEl = (org.docx4j.wml.Document) documentPart  
            .getContents();  
    Body body = wmlDocumentEl.getBody();  
    CommentFinder cf = new CommentFinder();  
    new TraversalUtil(body, cf);  
    for (Child commentElement : cf.commentElements) {  
        System.out.println(commentElement.getClass().getName());  
        Object parent = commentElement.getParent();  
        List<Object> theList = ((ContentAccessor) parent).getContent();  
        boolean removeResult = remove(theList, commentElement);  
        System.out.println(removeResult);  
    }  
    wordMLPackage.save(new FileOutputStream(savePath));  
}
 

/**
 * Replace text
 *
 * See also http://www.docx4java.org/forums/docx-java-f6/best-approach-to-search-replace-in-a-template-merge-t1040.html
 */
public static void replaceText(InputStream input, HashMap<String, String> model, OutputStream output) throws Docx4JException,
		JAXBException, IOException {
	log.info("replaceText({}, {}, {})", new Object[]{input, model, output});
	WordprocessingMLPackage wordMLPackage = WordprocessingMLPackage.load(input);
	MainDocumentPart documentPart = wordMLPackage.getMainDocumentPart();

	for (final Map.Entry<String, String> entry : model.entrySet()) {
		new TraversalUtil(documentPart, new TraversalUtil.CallbackImpl() {
			@Override
			public List<Object> apply(Object child) {
				if (child instanceof org.docx4j.wml.Text) {
					org.docx4j.wml.Text t = (org.docx4j.wml.Text) child;

					if (t.getValue().contains(entry.getKey())) {
						t.setValue(t.getValue().replaceAll(entry.getKey(), entry.getValue()));
					}
				}

				return null;
			}
		});
	}

	// Save it
	wordMLPackage.save(output);
	log.info("replaceText: void");
}
 

protected void collectDataFromObjects(Object... objects) {
    for (Object object : objects) {
        if (object != null) {
            TextVisitor collectAliasesCallback = new TextVisitor(docxFormatter);
            new TraversalUtil(object, collectAliasesCallback);
            texts.addAll(collectAliasesCallback.textWrappers);
        }
    }
}
 

public Tr copyRow(Tr row) {
    Tr copiedRow = XmlUtils.deepCopy(row);
    new TraversalUtil(copiedRow, INVARIANTS_SETTER);//set parent for each sub-element of copied row (otherwise parent would be JaxbElement)
    int index = table.getContent().indexOf(row);
    table.getContent().add(index, copiedRow);
    return copiedRow;
}
 

/**
 * 替换书签为图片
 *
 * @param wordMLPackage
 * @param documentPart
 * @param imageParameters
 * @throws Exception
 */
private static void replaceBookMarkWithImage(WordprocessingMLPackage wordMLPackage,
                                             MainDocumentPart documentPart,
                                             Map<String, String> imageParameters)
        throws Exception {
    Document wmlDoc = documentPart.getContents();
    Body body = wmlDoc.getBody();
    // 提取正文中所有段落
    List<Object> paragraphs = body.getContent();
    // 提取书签并创建书签的游标
    RangeFinder rt = new RangeFinder("CTBookmark", "CTMarkupRange");
    new TraversalUtil(paragraphs, rt);

    // 遍历书签
    for (CTBookmark bm : rt.getStarts()) {
        String bookmarkName = bm.getName();
        String imagePath = imageParameters.get(bookmarkName);
        if (imagePath != null) {
            File imageFile = new File(imagePath);
            InputStream imageStream = new FileInputStream(imageFile);
            // 读入图片并转化为字节数组，因为docx4j只能字节数组的方式插入图片
            byte[] bytes = IOUtils.toByteArray(imageStream);
            // 创建一个行内图片
            BinaryPartAbstractImage imagePart = BinaryPartAbstractImage.createImagePart(wordMLPackage, bytes);
            // createImageInline函数的前四个参数我都没有找到具体啥意思
            // 最后一个是限制图片的宽度，缩放的依据
            Inline inline = imagePart.createImageInline(null, null, 0, 1, false, 800);
            // 获取该书签的父级段落
            P p = (P) (bm.getParent());
            ObjectFactory factory = new ObjectFactory();
            // 新建一个Run
            R run = factory.createR();
            // drawing 画布
            Drawing drawing = factory.createDrawing();
            drawing.getAnchorOrInline()
                    .add(inline);
            run.getContent()
                    .add(drawing);
            p.getContent()
                    .add(run);
        }
    }
}
 

/**
   * Since we start with headers/footers which each take up approx half the page,
   * there is little room for the body content (which would result in many pages,
   * and unnecessary processing).
   * 
   * At the same time, we need enough body content to produce first page, odd page,
   * and even page for each section.
   * 
   * So this method replaces the existing body content with content which is sufficient
   * for our needs.  This method isn't essential, but it should make things faster.
   * 
   * It leaves the headers/footers untouched, since it is those which we're 
   * most interested in at this point.
   *  
   * @param hfPkg
   */
  static void trimContent(WordprocessingMLPackage hfPkg)   {
  	
  	// Find the sectPrs
  	SectPrFinder sf = new SectPrFinder(hfPkg.getMainDocumentPart());
try {
	new TraversalUtil(hfPkg.getMainDocumentPart().getContents(), sf);
} catch (Docx4JException e) {
	// TODO Auto-generated catch block
	log.error(e.getMessage(), e);
}  

List<SectPr> sectPrList = sf.getSectPrList();

// Was there a body level one?
if (hfPkg.getMainDocumentPart().getJaxbElement().getBody().getSectPr()!=null) {
	//then delete the first entry (which is where SectPrFinder put it)
	sectPrList.remove(0);  
}

// Now generate content; let's use
P filler = createFillerP();
List<Object> contents = hfPkg.getMainDocumentPart().getContent();
contents.clear();

for (SectPr sectPr : sectPrList) {
	
	contents.add(filler);
	contents.add(filler);
	contents.add(filler);
	contents.add(filler);
	
	// We expect to cause, in due course, something like:
	// WARN org.apache.fop.apps.FOUserAgent .processEvent line 97 - 
	//          The contents of fo:region-body on page 6 exceed its viewport 
	//          by 29068 millipoints. (See position 1:1038)


	// now add the sectPr
   	P p = Context.getWmlObjectFactory().createP(); 
  	    PPr ppr = Context.getWmlObjectFactory().createPPr(); 
  	    p.setPPr(ppr);
  	    ppr.setSectPr(sectPr);
  	    
	contents.add(p);
	
}

// Add content before the body level sectPr
if (hfPkg.getMainDocumentPart().getJaxbElement().getBody().getSectPr()!=null) {

	contents.add(filler);
	contents.add(filler);
	contents.add(filler);
	contents.add(filler);
	
}
  }
 

protected void collectTexts() {
    TextVisitor collectAliasesCallback = new TextVisitor(docxFormatter);
    new TraversalUtil(mainDocumentPart, collectAliasesCallback);
    texts = collectAliasesCallback.textWrappers;
}
 

protected void collectTables() {
    TableCollector collectTablesCallback = new TableCollector(docxFormatter);
    new TraversalUtil(mainDocumentPart, collectTablesCallback);
    tables = collectTablesCallback.tableManagers;
}
 

public List<Object> apply(Object object) {
    final TableManager currentTable = !currentTables.isEmpty() ? currentTables.peek() : null;
    if (currentTable == null || currentTable.isSkipIt()) {
        return null;
    }

    if (object instanceof Tr) {
        Tr currentRow = (Tr) object;

        if (currentTable.firstRow == null) {
            currentTable.firstRow = currentRow;

            findNameForCurrentTable(currentTable);

            if (currentTable.bandName == null) {
                currentTable.setSkipIt(true);
            } else {
                tableManagers.add(currentTable);
            }
        }

        if (currentTable.rowWithAliases == null) {
            RegexpCollectionFinder<P> aliasFinder = new RegexpCollectionFinder<P>(docxFormatter, AbstractFormatter.UNIVERSAL_ALIAS_PATTERN, P.class);
            new TraversalUtil(currentRow, aliasFinder);
            List<String> foundAliases = aliasFinder.getValues();
            if (!foundAliases.isEmpty()) {
                boolean fromCurrentBand = false;
                for (String foundAlias : foundAliases) {
                    String parameterName = docxFormatter.unwrapParameterName(foundAlias);
                    if (parameterName != null) {
                        String[] parts = parameterName.split("\\.");
                        if (parts.length == 1) {
                            fromCurrentBand = true;
                            break;
                        } else if (docxFormatter.findBandByPath(parts[0]) == null) {
                            fromCurrentBand = true;
                            break;
                        }
                    }
                }
                if (fromCurrentBand) {
                    currentTable.rowWithAliases = currentRow;
                }
            }
        }
    }

    return null;
}
 

protected void handleUrls() {
    UrlVisitor urlVisitor = new UrlVisitor(new DocxFormatterDelegate(this), wordprocessingMLPackage.getMainDocumentPart());
    new TraversalUtil(wordprocessingMLPackage.getMainDocumentPart(), urlVisitor);
}