Pocket Word implementation of DocumentDeserializer
* for use by {@link
* org.openoffice.xmerge.converter.xml.sxw.pocketword.PluginFactoryImpl
* PluginFactoryImpl}.
This converts an OpenOffice Writer XML files to a Pocket Word file<.
* * @author Mark Murnane * @version 1.1 */ public final class DocumentSerializerImpl implements DocumentSerializer, OfficeConstants { private PocketWordDocument pswDoc; private SxwDocument sxwDoc; private StyleCatalog styleCat = null; private boolean inList = false; /** *Initialises a new DocumentSerializerImpl using the.
* supplied Document
The supplied document should be an {@link * org.openoffice.xmerge.converter.xml.sxw.SxwDocument SxwDocument} * object.
* * @param doc TheDocument to convert.
*/
public DocumentSerializerImpl(Document doc) {
sxwDoc = (SxwDocument)doc;
pswDoc = new PocketWordDocument(sxwDoc.getName());
}
/**
* Convert the data passed into the DocumentSerializerImpl
* constructor into Pocket Word format.
This method may or may not be thread-safe. It is expected * that the user code does not call this method in more than one * thread. And for most cases, this method is only done once.
* * @returnConvertData object to pass back the
* converted data.
*
* @throws ConvertException If any conversion error occurs.
* @throws IOException If any I/O error occurs.
*/
public ConvertData serialize() throws IOException, ConvertException {
ConvertData cd = new ConvertData();
org.w3c.dom.Document doc = sxwDoc.getContentDOM();
// Load any style info before traversing the document content tree
loadStyles();
NodeList list = doc.getElementsByTagName(TAG_OFFICE_BODY);
int len = list.getLength();
if (len > 0) {
Node node = list.item(0);
traverseBody(node);
}
cd.addDocument(pswDoc);
return cd;
}
/*
* Handles the loading of defined styles from the style.xml file as well
* as automatic styles from the content.xml file.
*
* Any change to a defined style, such as a short bold section, falls into
* the latter category.
*/
private void loadStyles() {
org.w3c.dom.Document contentDom = sxwDoc.getContentDOM();
org.w3c.dom.Document styleDom = sxwDoc.getStyleDOM();
styleCat = new StyleCatalog(25);
NodeList nl = null;
String families[] = new String[] { PocketWordConstants.TEXT_STYLE_FAMILY,
PocketWordConstants.PARAGRAPH_STYLE_FAMILY,
PocketWordConstants.PARAGRAPH_STYLE_FAMILY };
Class classes[] = new Class[] { TextStyle.class,
ParaStyle.class,
TextStyle.class };
String[] styleTypes = new String[] { TAG_OFFICE_STYLES,
TAG_OFFICE_AUTOMATIC_STYLES,
TAG_OFFICE_MASTER_STYLES };
/*
* Documents converted from PSW -> SXW will not have a style.xml when
* being converted back to PSW. This would occur if a document was
* not modified within Writer between conversions.
*
* Any Writer modifications and saves create the style.xml and other
* portions of a complete Writer SXW file.
*/
if (styleDom != null) {
// Process the Style XML tree
for (int i = 0; i < styleTypes.length; i++ ) {
nl = styleDom.getElementsByTagName(styleTypes[i]);
if (nl.getLength() != 0) {
styleCat.add(nl.item(0), families, classes, null, false);
}
}
}
/*
* Process the content XML for any other style info.
* Should only be automatic types here.
*/
for (int i = 0; i < styleTypes.length; i++ ) {
nl = contentDom.getElementsByTagName(styleTypes[i]);
if (nl.getLength() != 0) {
styleCat.add(nl.item(0), families, classes, null, false);
}
}
}
/*
* Process the office:body tag.
*/
private void traverseBody(Node node) throws IOException, ConvertException {
if (node.hasChildNodes()) {
NodeList nList = node.getChildNodes();
int len = nList.getLength();
for (int i = 0; i < len; i++) {
Node child = nList.item(i);
if (child.getNodeType() == Node.ELEMENT_NODE) {
String nodeName = child.getNodeName();
if (nodeName.equals(TAG_PARAGRAPH)
|| nodeName.equals(TAG_HEADING)) {
traverseParagraph(child);
}
if (nodeName.equals(TAG_UNORDERED_LIST) ||
nodeName.equals(TAG_ORDERED_LIST)) {
traverseList(child);
}
}
}
}
}
/*
* Process a text:p tag
*/
private void traverseParagraph(Node node) throws IOException, ConvertException {
String styleName = getAttribute(node, ATTRIBUTE_TEXT_STYLE_NAME);
ParaStyle pstyle = (ParaStyle)styleCat.lookup(styleName,
PocketWordConstants.PARAGRAPH_STYLE_FAMILY, null,
ParaStyle.class);
if (pstyle != null) {
pstyle = (ParaStyle)pstyle.getResolved();
}
TextStyle tstyle = (TextStyle)styleCat.lookup(styleName,
PocketWordConstants.PARAGRAPH_STYLE_FAMILY, null,
TextStyle.class);
if (pstyle != null) {
tstyle = (TextStyle)tstyle.getResolved();
}
try {
pswDoc.addParagraph(pstyle, inList);
}
catch (Exception e) {
throw new ConvertException(
"Error adding paragraph to PocketWordDocument.\n"
+ e.toString());
}
traverseParagraphContents(node, tstyle);
}
/*
* Process the contents of a paragraph. This method handles situations
* where the paragraph contains multiple children, each representing a
* differently formatted piece of text.
*/
private void traverseParagraphContents (Node node, TextStyle defTextStyle)
throws IOException, ConvertException {
// First up, get the style of this little bit
String styleName = getAttribute(node, ATTRIBUTE_TEXT_STYLE_NAME);
TextStyle tStyle = (TextStyle)styleCat.lookup(styleName,
PocketWordConstants.TEXT_STYLE_FAMILY, null,
TextStyle.class);
if (tStyle == null) {
tStyle = defTextStyle;
}
if (node.hasChildNodes()) {
NodeList nList = node.getChildNodes();
int len = nList.getLength();
for (int i = 0; i < len; i++) {
Node child = nList.item(i);
short nodeType = child.getNodeType();
switch (nodeType) {
case Node.TEXT_NODE:
String s = child.getNodeValue();
if (s.length() > 0) {
try {
pswDoc.addParagraphData(s, tStyle);
}
catch (Exception e) {
throw new ConvertException(
"Error adding data to paragraph in " +
"PocketWordDocument.\n" + e.toString());
}
}
break;
case Node.ELEMENT_NODE:
if (child.getNodeName().equals(TAG_SPACE)) {
StringBuffer sb = new StringBuffer("");
int count = 1;
NamedNodeMap map = child.getAttributes();
if (map.getLength() > 0) {
Node attr = map.getNamedItem(ATTRIBUTE_SPACE_COUNT);
count = Integer.parseInt(attr.getNodeValue().trim());
}
for ( ; count > 0; count--) {
sb.append(" ");
}
/*
* May want to look at style info for spaces. Could
* be important when calculating font metrics.
*/
try {
pswDoc.addParagraphData(sb.toString(), tStyle);
}
catch (Exception e) {
throw new ConvertException(
"Error adding data to paragraph in " +
"PocketWordDocument.\n" + e.toString());
}
}
else if (child.getNodeName().equals(TAG_TAB_STOP)) {
try {
pswDoc.addParagraphData("\t", tStyle);
}
catch (Exception e) {
throw new ConvertException(
"Error adding data to paragraph in " +
"PocketWordDocument.\n" + e.toString());
}
}
else if (child.getNodeName().equals(TAG_LINE_BREAK)) {
/*
* Pocket Word does not support soft line breaks.
* They are just new paragraphs.
*/
}
else if (child.getNodeName().equals(TAG_SPAN)) {
/*
* This is where the interesting ones, i.e. format
* changes occur.
*/
traverseParagraphContents (child, defTextStyle);
}
else if (child.getNodeName().equals(TAG_HYPERLINK)) {
traverseParagraphContents (child, defTextStyle);
}
else {
// Should maybe have a default in here.
}
break;
default:
// Do nothing
}
}
}
else {
/*
* If the node has no children, then it is a blank paragraph, but
* they still require an entry in the Paragraph class to make sense.
*/
pswDoc.addParagraphData("", tStyle);
}
}
/*
* Process a text:ordered-list or text:unordered-list tag. Pocket Word has
* no concept of a list so there is no need to differentiate between the
* two.
*
* Each item on the list contains a text:p node.
*/
private void traverseList (Node node) throws IOException, ConvertException {
inList = true;
if (node.hasChildNodes()) {
NodeList nList = node.getChildNodes();
int len = nList.getLength();
for (int i = 0; i < len; i++) {
Node child = nList.item(i);
if (child.getNodeType() == Node.ELEMENT_NODE) {
String nodeName = child.getNodeName();
if (nodeName.equals(TAG_LIST_ITEM)) {
traverseListItem(child);
}
}
}
}
inList = false;
}
/*
* Process a text:list-item node. They usually contain have a single
* text:p child but can also have sections or other lists.
*
* For this case, only paragraphs are supported.
*/
private void traverseListItem (Node node) throws IOException, ConvertException {
if (node.hasChildNodes()) {
NodeList nList = node.getChildNodes();
int len = nList.getLength();
for (int i = 0; i < len; i++) {
Node child = nList.item(i);
if (child.getNodeType() == Node.ELEMENT_NODE) {
String nodeName = child.getNodeName();
if (nodeName.equals(TAG_PARAGRAPH)) {
traverseParagraph(child);
}
}
}
}
}
/*
* Utility method to retrieve a Node attribute.
*/
private String getAttribute (Node node, String attribute) {
NamedNodeMap attrNodes = node.getAttributes();
if (attrNodes != null) {
Node attr = attrNodes.getNamedItem(attribute);
if (attr != null) {
return attr.getNodeValue();
}
}
return null;
}
}