Examples with SectionSegBean net.heartsome.cat.converter.word2007.common.SectionSegBean used on opensource projects

Example 1 with SectionSegBean

use of net.heartsome.cat.converter.word2007.common.SectionSegBean in project translationstudio8 by heartsome.

the class PartOperate method getLinkText.

//		<w:hyperlink r:id="rId5" w:history="1">
//		<w:r w:rsidRPr="007267A9">
//			<w:rPr>
//				<w:rStyle w:val="a5" />
//				<w:noProof />
//			</w:rPr>
//			<w:t>www.baidu.</w:t>
//		</w:r>
//		<w:r w:rsidRPr="0077163A">
//			<w:rPr>
//				<w:rStyle w:val="a5" />
//				<w:noProof />
//				<w:color w:val="FF0000" />
//			</w:rPr>
//			<w:t>com</w:t>
//		</w:r>
//	   </w:hyperlink>
/**
	 * 获取链接的特殊文本<br/>
	 * 备注：上面是一个关于　w:hyperlink 的例子，这时，一个　w:hyperlink　就当作一个　g 标记进行处理。那么，它的特列文本　就是其应提取的所有内容，不单指 "www.baidu.com"，而是　3　个　w:r 节点。<br/>
	 * 若当前节点中只有一个　w:r 节点，那么不予考虑。返回　null,　若有多个　w:r 节点，那将它们拼接成　sub 与 ph 标记组合的文本，当作　特殊文本进行返回。
	 * @return
	 */
private String getLinkText() throws Exception {
    vn.push();
    String curLinkText = null;
    int rNodeCount = getNodeCount("./w:r");
    if (rNodeCount == 1) {
        return curLinkText;
    }
    List<SectionSegBean> segList = new LinkedList<SectionSegBean>();
    extendAP.selectXPath("./node()");
    while (extendAP.evalXPath() != -1) {
        String rStyle = getRStyle();
        String text = "";
        String extendNodesStr = getExtendNodes();
        int tNodeCount = getNodeCount("./w:t");
        // 有　t 节点的，就一定有数据
        if (tNodeCount == 1) {
            text = getText();
            // extendNodesAttr 这个属性是自定义的，专门针对于除样式与文本之外的其他节点，如 <w:tab/>
            String extendNodesAttr = ("".equals(extendNodesStr) || extendNodesStr == null) ? "" : " extendNodes='" + xlfOutput.cleanString(extendNodesStr) + "'";
            // rprAttr 这个是保存 rpr 属性的。rpr属性是用于保存 w:r 样式。
            String rprAttr = ("".equals(rStyle) || rStyle == null) ? "" : " rPr='" + xlfOutput.cleanString(rStyle) + "'";
            segList.add(new SectionSegBean(null, text, rprAttr, extendNodesAttr, null));
        } else {
            String phTagStr = vu.getElementFragment();
            segList.add(new SectionSegBean(null, null, null, null, xlfOutput.cleanString(phTagStr)));
        }
    }
    vn.pop();
    StringBuffer specialTextSB = new StringBuffer();
    SectionSegBean bean;
    for (int i = 0; i < segList.size(); i++) {
        bean = segList.get(i);
        String style = bean.getStyle() == null ? "" : bean.getStyle();
        String extendNodes = bean.getExtendNodesStr() == null ? "" : bean.getExtendNodesStr();
        if (("".equals(style)) && ("".equals(extendNodes)) && bean.getPhTagStr() == null) {
            specialTextSB.append(bean.getText());
        } else if (bean.getPhTagStr() == null) {
            specialTextSB.append("<sub id='" + xlfOutput.useTagId() + "'" + style + extendNodes + ">");
            specialTextSB.append(bean.getText());
            // 判断下一个是否样式与扩展节点的内容都相同，若相同，就组装成一个g标记
            while (i + 1 < segList.size()) {
                bean = segList.get(i + 1);
                String curStyle = bean.getStyle() == null ? "" : bean.getStyle();
                String curExtendNodes = bean.getExtendNodesStr() == null ? "" : bean.getExtendNodesStr();
                // 当两个的 ctype 都为空时，才能进行拼接，因为 ctype 多半为 <w:hyperlink .... >
                if (curStyle.equals(style) && curExtendNodes.equals(extendNodes)) {
                    specialTextSB.append(bean.getText());
                    i++;
                } else {
                    break;
                }
            }
            specialTextSB.append("</sub>");
        } else {
            String phTagStr = bean.getPhTagStr();
            if (!"".equals(phTagStr.trim())) {
                specialTextSB.append("<ph id='" + xlfOutput.useTagId() + "'>");
                specialTextSB.append(phTagStr);
                while (i + 1 < segList.size()) {
                    bean = segList.get(i + 1);
                    if (bean.getPhTagStr() != null) {
                        specialTextSB.append(bean.getPhTagStr());
                        i++;
                    } else {
                        break;
                    }
                }
                specialTextSB.append("</ph>");
            }
        }
    }
    curLinkText = specialTextSB.toString();
    if (curLinkText.length() == 0) {
        curLinkText = null;
    }
    return curLinkText;
}

Example 2 with SectionSegBean

use of net.heartsome.cat.converter.word2007.common.SectionSegBean in project translationstudio8 by heartsome.

the class XliffInputer method anaysisTgtOrSrcNode.

/**
	 * 分析 source 或 target 节点，获取其内容
	 * @throws Exception
	 */
private void anaysisTgtOrSrcNode(StringBuffer textSB, VTDNav vn, boolean isText) throws Exception {
    vn.push();
    AutoPilot otherAP = new AutoPilot(vn);
    String childXpath = "./text()|node()";
    otherAP.selectXPath(childXpath);
    int tokenId = -1;
    int index = -1;
    Map<Integer, SectionSegBean> targetMap = new TreeMap<Integer, SectionSegBean>();
    while (otherAP.evalXPath() != -1) {
        index = vn.getCurrentIndex();
        tokenId = vn.getTokenType(index);
        if (tokenId == 0) {
            //节点子节点
            ananysisTag(vn, targetMap);
        } else if (tokenId == 5) {
            // 文本子节点
            //				if ("+1 845-536-1416".equals(vn.toString(index))) {
            //					System.out.println("问题开始了。。。。");
            //				}
            targetMap.put(index, new SectionSegBean(null, vn.toRawString(index), null, null, null));
        }
    }
    vn.pop();
    SectionSegBean bean;
    for (Entry<Integer, SectionSegBean> entry : targetMap.entrySet()) {
        bean = entry.getValue();
        if (isText) {
            if (bean.getText() != null) {
                textSB.append(bean.getText());
            }
        } else {
            // 这个要组装 w:r 等节点
            String ctype = bean.getCtype() == null ? "" : bean.getCtype();
            String style = bean.getStyle() == null ? "" : bean.getStyle();
            String extendNodes = bean.getExtendNodesStr() == null ? "" : bean.getExtendNodesStr();
            if (bean.getPhTagStr() != null) {
                textSB.append(bean.getPhTagStr());
            } else {
                if ("".equals(ctype)) {
                    textSB.append("<w:r>" + style + extendNodes);
                    textSB.append("<w:t xml:space=\"preserve\">" + bean.getText() + "</w:t></w:r>");
                } else {
                    // <w:hyperlink r:id="rId8" w:history="1">
                    int endIdx = ctype.indexOf(" ") == -1 ? ctype.indexOf(">") : ctype.indexOf(" ");
                    String nodeName = ctype.substring(ctype.indexOf("<") + 1, endIdx);
                    textSB.append(ctype);
                    textSB.append("<w:r>" + style + extendNodes);
                    textSB.append("<w:t xml:space=\"preserve\">" + bean.getText() + "</w:t></w:r>");
                    textSB.append("</" + nodeName + ">");
                }
            }
        }
    }
}

Example 3 with SectionSegBean

use of net.heartsome.cat.converter.word2007.common.SectionSegBean in project translationstudio8 by heartsome.

the class PartOperate method analysisNodeP.

/**
	 * 分析每个 w:p 节点，将要翻译的东西提取出来，用占位符替代。
	 *  or name()='w:fldSimple'
	 * @throws Exception
	 */
protected void analysisNodeP() throws Exception {
    // 如果这个节点里面还有 p 节点，那么不进行处理
    int textCount = getTextCount();
    int index = -1;
    Map<Integer, SectionSegBean> sectionSegMap = new TreeMap<Integer, SectionSegBean>();
    // 占位符
    StringBuffer placeHolderSB = new StringBuffer();
    // 开始处理每个节点的文本
    if (textCount == 1) {
        // 如果一个节点里面只有一个 w:r//w:t ,那么直接获取出内容，如果这个节点里面还有 p 节点，那么不进行处理
        childAP.selectXPath("./node()[(name()='w:r' or name()='w:hyperlink') and not(@w:anchor) and not(descendant::node()[name()='w:p'])]/descendant::w:t/text()");
        vn.push();
        if (childAP.evalXPath() != -1) {
            index = vn.getCurrentIndex();
            String segment = vn.toRawString(index);
            if ("".equals(deleteBlank(segment))) {
                vn.pop();
                return;
            }
            String[] segs = segmenter.segment(segment);
            for (String seg : segs) {
                // 生成 trans-unit 节点
                placeHolderSB.append(xlfOutput.addTransUnit(seg));
            }
            xm.updateToken(index, placeHolderSB.toString());
        }
        vn.pop();
    } else if (textCount > 1) {
        // 没有 w:/t 节点的段落，不进行处理。如果这个节点里面还有 p 节点，那么不进行处理
        //			System.out.println("vn.getCurrentIndex() = " + vn.getCurrentIndex());
        //			if (185 == vn.getCurrentIndex()) {
        //				System.out.println("调试开始了。。。。");
        //			}
        vn.push();
        // 先获取出这个段落里所有要翻译的数据，再分段
        List<StringBuffer> segList = new ArrayList<StringBuffer>();
        StringBuffer segSB = new StringBuffer();
        childAP.selectXPath("./node()[(name()='w:r' or name()='w:hyperlink') and not(@w:anchor) and not(descendant::node()[name()='w:p'])]");
        while (childAP.evalXPath() != -1) {
            if (vu.getElementContent().indexOf("<w:br") != -1) {
                segList.add(segSB);
                segSB = new StringBuffer();
            }
            String curText = getText();
            curText = (curText == null ? "" : curText);
            segSB.append(curText);
        }
        segList.add(segSB);
        // 如果为空格，直接退出
        StringBuffer checkBlankSB = new StringBuffer();
        for (StringBuffer curSB : segList) {
            checkBlankSB.append(curSB);
        }
        if (deleteBlank(checkBlankSB.toString()).length() <= 0) {
            vn.pop();
            return;
        }
        //			System.out.println("checkBlankSB.toString() =" + checkBlankSB.toString());
        //			if (segSB.toString().indexOf("Definition des vorzuhaltenden") != -1) {
        //				System.out.println("错误信息开始了。。。。。");
        //			}
        vn.pop();
        // 开始分割文本段
        List<String> segArrayList = new ArrayList<String>();
        for (StringBuffer curSB : segList) {
            String[] segArray = segmenter.segment(curSB.toString());
            for (int i = 0; i < segArray.length; i++) {
                segArrayList.add(segArray[i]);
            }
        }
        String[] segArray = segArrayList.toArray(new String[] {});
        // 开始遍历每个节点 ./node()，进行处理
        vn.push();
        childAP.selectXPath("./node()");
        // 一个文本段的开始
        boolean segStart = false;
        // 一个文本段结束的标记
        boolean segOver = false;
        int segIdx = 0;
        String seg = segArray[segIdx];
        while (childAP.evalXPath() != -1) {
            index = vn.getCurrentIndex();
            isLink = false;
            linkText = null;
            String nodeName = vu.getCurrentElementName();
            if ("w:r".equals(nodeName) || "w:hyperlink".equals(nodeName) || "w:fldSimple".equals(nodeName)) {
                if ("w:hyperlink".equals(nodeName)) {
                    isLink = true;
                    linkText = getLinkText();
                }
                if ("w:r".equals(nodeName)) {
                    // 检查是否有软回车 <w: br/>
                    vn.push();
                    boolean hasBr = false;
                    extendAP.selectXPath("./w:br");
                    if (extendAP.evalXPath() != -1) {
                        hasBr = true;
                    }
                    vn.pop();
                    if (hasBr && (sectionSegMap.size() > 0)) {
                        // 遇到软回车就开始分段
                        segOver = true;
                        segStart = false;
                        String placeHoderStr = xlfOutput.addTransUnit(createSourceStr(sectionSegMap));
                        xm.insertBeforeElement(placeHoderStr);
                        segArray[segIdx] = seg;
                        continue;
                    }
                }
                String text = getText();
                if (text != null) {
                    segStart = true;
                    segOver = false;
                    String ctypeAttrStr = "";
                    String rStyle = null;
                    String extendNodesStr = null;
                    if (isLink) {
                        ctypeAttrStr = " ctype='" + xlfOutput.cleanString(vu.getElementHead()) + "'";
                        vn.push();
                        extendAP.selectXPath("./w:r");
                        // 如果　linkText 不为空的话，就不用获取　rStyle 与　extendNodesStr 属性了
                        if (linkText == null && extendAP.evalXPath() != -1) {
                            rStyle = getRStyle();
                            extendNodesStr = getExtendNodes();
                        }
                        vn.pop();
                    } else {
                        rStyle = getRStyle();
                        extendNodesStr = getExtendNodes();
                    }
                    // extendNodesAttr 这个属性是自定义的，专门针对于除样式与文本之外的其他节点，如 <w:tab/>
                    String extendNodesAttr = ("".equals(extendNodesStr) || extendNodesStr == null) ? "" : " extendNodes='" + xlfOutput.cleanString(extendNodesStr) + "'";
                    // rprAttr 这个是保存 rpr 属性的。rpr属性是用于保存 w:r 样式。
                    String rprAttr = ("".equals(rStyle) || rStyle == null) ? "" : " rPr='" + xlfOutput.cleanString(rStyle) + "'";
                    // 如果当前文本是一个独立的分段，则将其文本用占位符替换。
                    if (text.equals(segArray[segIdx])) {
                        String placeHoderStr = "";
                        if (isLink && linkText != null) {
                            placeHoderStr = xlfOutput.addTransUnit(linkText);
                        } else {
                            placeHoderStr = xlfOutput.addTransUnit(seg);
                        }
                        updateTextToPlaceHoder(placeHoderStr);
                        if (segIdx + 1 < segArray.length) {
                            seg = segArray[++segIdx];
                        }
                        segOver = true;
                        segStart = false;
                        continue;
                    }
                    // 链接不应支持分段，现开始分析
                    if (isLink) {
                        List<Object> resultList = modifySeg(segArray, segIdx, text, seg);
                        segArray = (String[]) resultList.get(0);
                        seg = (String) resultList.get(1);
                    }
                    // 分析分割后的文本段
                    if (text.equals(seg)) {
                        if ("".equals(rprAttr) && "".equals(extendNodesAttr)) {
                            // 只添加纯文本
                            sectionSegMap.put(index, newSectionSegBean(ctypeAttrStr, text, null, null, null));
                        } else {
                            sectionSegMap.put(index, newSectionSegBean(ctypeAttrStr, text, rprAttr, extendNodesAttr, null));
                        }
                        if (segIdx + 1 < segArray.length) {
                            seg = segArray[++segIdx];
                        }
                        segOver = true;
                        segStart = false;
                        xm.remove();
                        String placeHoderStr = xlfOutput.addTransUnit(createSourceStr(sectionSegMap));
                        xm.insertAfterElement(placeHoderStr);
                    } else if (text.length() < seg.length() && seg.indexOf(text) == 0) {
                        // 如果当前文本长度小于分段长度
                        if ("".equals(rprAttr) && "".equals(extendNodesAttr)) {
                            sectionSegMap.put(index, newSectionSegBean(ctypeAttrStr, text, null, null, null));
                        } else {
                            sectionSegMap.put(index, newSectionSegBean(ctypeAttrStr, text, rprAttr, extendNodesAttr, null));
                        }
                        xm.remove();
                        // 在替换有 "(" 的情况。必须加一个 \\( 或者 [(]，否则会报错
                        seg = seg.substring(text.length());
                    } else if (text.length() > seg.length() && text.indexOf(seg) == 0) {
                        if ("".equals(rprAttr) && "".equals(extendNodesAttr)) {
                            sectionSegMap.put(index, newSectionSegBean(ctypeAttrStr, seg, null, null, null));
                        } else {
                            sectionSegMap.put(index, newSectionSegBean(ctypeAttrStr, seg, rprAttr, extendNodesAttr, null));
                        }
                        text = text.substring(seg.length());
                        // 由于这里有可能要插入多个占位符，所以要把所有的占位符放到一起，一次性存入，因为 xmlModifial 不允许在同一个地方修改多次。
                        StringBuffer replaceHolderSB = new StringBuffer();
                        replaceHolderSB.append(xlfOutput.addTransUnit(createSourceStr(sectionSegMap)));
                        xm.remove();
                        if (segIdx + 1 < segArray.length) {
                            seg = segArray[++segIdx];
                        }
                        // 开始处理剩下的文本
                        while (text.length() != 0) {
                            if (text.equals(seg)) {
                                if ("".equals(rprAttr) && "".equals(extendNodesAttr)) {
                                    sectionSegMap.put(index, newSectionSegBean(ctypeAttrStr, text, null, null, null));
                                } else {
                                    sectionSegMap.put(index, newSectionSegBean(ctypeAttrStr, text, rprAttr, extendNodesAttr, null));
                                }
                                text = "";
                                segOver = true;
                                segStart = false;
                                replaceHolderSB.append(xlfOutput.addTransUnit(createSourceStr(sectionSegMap)));
                                if (segIdx + 1 < segArray.length) {
                                    seg = segArray[++segIdx];
                                }
                            } else if (text.length() < seg.length() && seg.indexOf(text) == 0) {
                                if ("".equals(rprAttr) && "".equals(extendNodesAttr)) {
                                    sectionSegMap.put(index, newSectionSegBean(ctypeAttrStr, text, null, null, null));
                                } else {
                                    sectionSegMap.put(index, newSectionSegBean(ctypeAttrStr, text, rprAttr, extendNodesAttr, null));
                                }
                                seg = seg.substring(text.length());
                                text = "";
                            } else if (text.length() > seg.length() && text.indexOf(seg) == 0) {
                                if ("".equals(rprAttr) && "".equals(extendNodesAttr)) {
                                    sectionSegMap.put(index, newSectionSegBean(ctypeAttrStr, seg, null, null, null));
                                } else {
                                    sectionSegMap.put(index, newSectionSegBean(ctypeAttrStr, seg, rprAttr, extendNodesAttr, null));
                                }
                                text = text.substring(seg.length());
                                replaceHolderSB.append(xlfOutput.addTransUnit(createSourceStr(sectionSegMap)));
                                if (segIdx + 1 < segArray.length) {
                                    seg = segArray[++segIdx];
                                }
                            }
                        }
                        xm.insertAfterElement(replaceHolderSB.toString());
                    }
                } else if (segStart && !segOver) {
                    sectionSegMap.put(index, new SectionSegBean(null, null, null, null, xlfOutput.cleanString(vu.getElementFragment())));
                    xm.remove();
                }
            } else if (segStart && !segOver) {
                sectionSegMap.put(index, new SectionSegBean(null, null, null, null, xlfOutput.cleanString(vu.getElementFragment())));
                xm.remove();
            }
        }
        vn.pop();
    }
}

Example 4 with SectionSegBean

use of net.heartsome.cat.converter.word2007.common.SectionSegBean in project translationstudio8 by heartsome.

the class PartOperate method createSourceStr.

/**
	 * 通过sectionSegMap生成要添加到 trans-unit 节点的源文本
	 * @param sectionSegMap
	 * @return
	 */
private String createSourceStr(Map<Integer, SectionSegBean> sectionSegMap) {
    List<SectionSegBean> segList = new LinkedList<SectionSegBean>();
    for (Entry<Integer, SectionSegBean> entry : sectionSegMap.entrySet()) {
        segList.add(entry.getValue());
    }
    StringBuffer srcTextSB = new StringBuffer();
    SectionSegBean bean = null;
    for (int i = 0; i < segList.size(); i++) {
        bean = segList.get(i);
        String ctype = bean.getCtype() == null ? "" : bean.getCtype();
        String style = bean.getStyle() == null ? "" : bean.getStyle();
        String extendNodes = bean.getExtendNodesStr() == null ? "" : bean.getExtendNodesStr();
        if (("".equals(ctype) && "".equals(style)) && ("".equals(extendNodes)) && bean.getPhTagStr() == null) {
            srcTextSB.append(bean.getText());
        } else if (bean.getPhTagStr() == null) {
            srcTextSB.append("<g id='" + xlfOutput.useTagId() + "'" + ctype + style + extendNodes + ">");
            srcTextSB.append(bean.getText());
            // 判断下一个是否样式与扩展节点的内容都相同，若相同，就组装成一个g标记
            while (i + 1 < segList.size()) {
                bean = segList.get(i + 1);
                String curCtype = bean.getCtype() == null ? "" : bean.getCtype();
                String curStyle = bean.getStyle() == null ? "" : bean.getStyle();
                String curExtendNodes = bean.getExtendNodesStr() == null ? "" : bean.getExtendNodesStr();
                // 当两个的 ctype 都为空时，才能进行拼接，因为 ctype 多半为 <w:hyperlink .... >
                if (curStyle.equals(style) && curExtendNodes.equals(extendNodes) && "".equals(ctype) && "".equals(curCtype)) {
                    srcTextSB.append(bean.getText());
                    i++;
                } else {
                    break;
                }
            }
            srcTextSB.append("</g>");
        } else {
            String phTagStr = bean.getPhTagStr();
            if (!"".equals(phTagStr.trim())) {
                srcTextSB.append("<ph id='" + xlfOutput.useTagId() + "'>");
                srcTextSB.append(phTagStr);
                while (i + 1 < segList.size()) {
                    bean = segList.get(i + 1);
                    if (bean.getPhTagStr() != null) {
                        srcTextSB.append(bean.getPhTagStr());
                        i++;
                    } else {
                        break;
                    }
                }
                srcTextSB.append("</ph>");
            }
        }
    }
    sectionSegMap.clear();
    return srcTextSB.toString();
}

Example 5 with SectionSegBean

use of net.heartsome.cat.converter.word2007.common.SectionSegBean in project translationstudio8 by heartsome.

the class XliffInputer method ananysisTag.

/**
	 * 分析标记
	 */
private void ananysisTag(VTDNav vn, Map<Integer, SectionSegBean> targetMap) throws Exception {
    vn.push();
    AutoPilot tagAP = new AutoPilot(vn);
    int index = vn.getCurrentIndex();
    String tagName = vn.toString(index);
    if ("g".equals(tagName)) {
        String style = "";
        int attrIdx = -1;
        if ((attrIdx = vn.getAttrVal("rPr")) != -1) {
            style = vn.toString(attrIdx);
        }
        String extendNodes = "";
        if ((attrIdx = vn.getAttrVal("extendNodes")) != -1) {
            extendNodes = vn.toString(attrIdx);
        }
        String ctype = "";
        if ((attrIdx = vn.getAttrVal("ctype")) != -1) {
            ctype = vn.toString(attrIdx);
        }
        // 首先检查　g 标记下是否有　sub　节点
        int subNodeCount = -1;
        tagAP.selectXPath("count(./descendant::sub)");
        subNodeCount = (int) tagAP.evalXPathToNumber();
        tagAP.selectXPath("./node()|text()");
        if (subNodeCount > 0) {
            int curIdx = vn.getCurrentIndex();
            StringBuffer gTextSB = new StringBuffer();
            Map<Integer, String> gTextMap = new TreeMap<Integer, String>();
            while (tagAP.evalXPath() != -1) {
                index = vn.getCurrentIndex();
                int tokenType = vn.getTokenType(index);
                if (tokenType == 0) {
                    //节点子节点
                    String nodeName = vn.toString(index);
                    if ("ph".equals(nodeName)) {
                        gTextMap.put(index, resetCleanStr(vu.getElementContent()));
                    } else if ("g".equals(nodeName)) {
                        ananysisTag(vn, targetMap);
                    } else if ("sub".equals(nodeName)) {
                        ananysisSubTag(vn, gTextMap, targetMap);
                    }
                } else if (tokenType == 5) {
                    //文本子节点
                    gTextMap.put(index, resetCleanStr(vn.toRawString(index)));
                }
            }
            for (Entry<Integer, String> entry : gTextMap.entrySet()) {
                gTextSB.append(entry.getValue());
            }
            targetMap.put(curIdx, new SectionSegBean(ctype, gTextSB.toString(), style, extendNodes, null));
        } else {
            while (tagAP.evalXPath() != -1) {
                index = vn.getCurrentIndex();
                int tokenType = vn.getTokenType(index);
                if (tokenType == 0) {
                    //节点子节点
                    String nodeName = vn.toString(index);
                    if ("ph".equals(nodeName)) {
                        targetMap.put(index, new SectionSegBean(null, null, null, null, resetCleanStr(vu.getElementContent())));
                    } else if ("g".equals(nodeName)) {
                        ananysisTag(vn, targetMap);
                    }
                } else if (tokenType == 5) {
                    //文本子节点
                    targetMap.put(index, new SectionSegBean(ctype, vn.toRawString(index), style, extendNodes, null));
                }
            }
        }
    } else if ("ph".equals(tagName)) {
        targetMap.put(index, new SectionSegBean(null, null, null, null, resetCleanStr(vu.getElementContent())));
    } else if ("sub".equals(tagName)) {
        String style = "";
        int attrIdx = -1;
        if ((attrIdx = vn.getAttrVal("rPr")) != -1) {
            style = vn.toString(attrIdx);
        }
        String extendNodes = "";
        if ((attrIdx = vn.getAttrVal("extendNodes")) != -1) {
            extendNodes = vn.toString(attrIdx);
        }
        tagAP.selectXPath("./node()|text()");
        while (tagAP.evalXPath() != -1) {
            index = vn.getCurrentIndex();
            int tokenType = vn.getTokenType(index);
            if (tokenType == 0) {
                //节点子节点
                String nodeName = vn.toString(index);
                if ("ph".equals(nodeName)) {
                    targetMap.put(index, new SectionSegBean(null, null, null, null, resetCleanStr(vu.getElementContent())));
                } else if ("g".equals(nodeName)) {
                    ananysisTag(vn, targetMap);
                }
            } else if (tokenType == 5) {
                //文本子节点
                targetMap.put(index, new SectionSegBean(null, vn.toRawString(index), style, extendNodes, null));
            }
        }
    } else {
        //其他节点，一律当做字符串处理
        targetMap.put(index, new SectionSegBean(null, null, null, null, resetCleanStr(vu.getElementFragment())));
    }
    vn.pop();
}