I'm trying to convert html text to generate a word table. It works pretty well, and the created word file is correct, except the character styles.
This is my first try with Apache POI.
So far, I was able to detect new line (<br>) tags from text paragraph (see code below). But I'd like to also check a few other tags such as <b>, <li>, <font> and set the right run values for each part.
For example :
This is my text <i> which now is in italic<b> but also in bold</b> depending on its importance</i>
I gess I should parse the text, and apply different runs for each part, but I don't know how to do.
private static XWPFParagraph getTableParagraph(XWPFTableCell cell, String text)
{
int fontsize= 11;
XWPFParagraph paragraph = cell.addParagraph();
cell.removeParagraph(0);
paragraph.setSpacingAfterLines(0);
paragraph.setSpacingAfter(0);
XWPFRun myRun1 = paragraph.createRun();
if (text==null) text="";
else
{
while (true)
{
int x = text.indexOf("<br>");
if (x <0) break;
String work = text.substring(0,x );
text= text.substring(x+4);
myRun1.setText(work);
myRun1.addBreak();
}
}
myRun1.setText(text);
myRun1.setFontSize(fontsize);
return paragraph;
}
While converting HTML text one never should go on the HTML
using string methods only. XML
as well as HTML
are markup languages. Their content is markup and not only plain text. The markup needs to be traversed to get all the single nodes together with the meanings out of it. This traversing process never is trivial and so special libraries are there for. Deep inside those libraries also needs using string methods but those are wrapped into useful methods for traversing the markup.
For traversing HTML
jsoup may be used for example. Especially NodeTraversor using NodeVisitor is useful for traversing HTML
.
My example creates a ParagraphNodeVisitor
which implements NodeVisitor
. This interface requests method public void head(Node node, int depth)
which is called every time the NodeTraversor
is on head of a node and public void tail(Node node, int depth)
which is called every time the NodeTraversor
is on tail of a node. In those methods the process for handling the single nodes can be implemented. In our case main part of the process is whether we need a new XWPFRun
and what settings this run needs.
Example:
import java.io.FileOutputStream;
import org.apache.poi.xwpf.usermodel.*;
import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Node;
import org.jsoup.nodes.TextNode;
import org.jsoup.nodes.Element;
import org.jsoup.select.Elements;
import org.jsoup.select.NodeVisitor;
import org.jsoup.select.NodeTraversor;
public class HTMLtoDOCX {
private XWPFDocument document;
public HTMLtoDOCX(String html, String docxPath) throws Exception {
this.document = new XWPFDocument();
XWPFParagraph paragraph = null;
Document htmlDocument = Jsoup.parse(html);
Elements htmlParagraphs = htmlDocument.select("p");
for(Element htmlParagraph : htmlParagraphs) {
System.out.println(htmlParagraph);
paragraph = document.createParagraph();
createParagraphFromHTML(paragraph, htmlParagraph);
}
FileOutputStream out = new FileOutputStream(docxPath);
document.write(out);
out.close();
document.close();
}
void createParagraphFromHTML(XWPFParagraph paragraph, Element htmlParagraph) {
ParagraphNodeVisitor nodeVisitor = new ParagraphNodeVisitor(paragraph);
NodeTraversor.traverse(nodeVisitor, htmlParagraph);
}
private class ParagraphNodeVisitor implements NodeVisitor {
String nodeName;
boolean needNewRun;
boolean isItalic;
boolean isBold;
boolean isUnderlined;
int fontSize;
String fontColor;
XWPFParagraph paragraph;
XWPFRun run;
ParagraphNodeVisitor(XWPFParagraph paragraph) {
this.paragraph = paragraph;
this.run = paragraph.createRun();
this.nodeName = "";
this.needNewRun = false;
this.isItalic = false;
this.isBold = false;
this.isUnderlined = false;
this.fontSize = 11;
this.fontColor = "000000";
}
@Override
public void head(Node node, int depth) {
nodeName = node.nodeName();
System.out.println("Start "+nodeName+": " + node);
needNewRun = false;
if ("#text".equals(nodeName)) {
run.setText(((TextNode)node).text());
needNewRun = true; //after setting the text in the run a new run is needed
} else if ("i".equals(nodeName)) {
isItalic = true;
} else if ("b".equals(nodeName)) {
isBold = true;
} else if ("u".equals(nodeName)) {
isUnderlined = true;
} else if ("br".equals(nodeName)) {
run.addBreak();
} else if ("font".equals(nodeName)) {
fontColor = (!"".equals(node.attr("color")))?node.attr("color").substring(1):"000000";
fontSize = (!"".equals(node.attr("size")))?Integer.parseInt(node.attr("size")):11;
}
if (needNewRun) run = paragraph.createRun();
needNewRun = false;
run.setItalic(isItalic);
run.setBold(isBold);
if (isUnderlined) run.setUnderline(UnderlinePatterns.SINGLE); else run.setUnderline(UnderlinePatterns.NONE);
run.setColor(fontColor); run.setFontSize(fontSize);
}
@Override
public void tail(Node node, int depth) {
nodeName = node.nodeName();
System.out.println("End "+nodeName);
if ("i".equals(nodeName)) {
isItalic = false;
} else if ("b".equals(nodeName)) {
isBold = false;
} else if ("u".equals(nodeName)) {
isUnderlined = false;
} else if ("font".equals(nodeName)) {
fontColor = "000000";
fontSize = 11;
}
if (needNewRun) run = paragraph.createRun();
needNewRun = false;
run.setItalic(isItalic);
run.setBold(isBold);
if (isUnderlined) run.setUnderline(UnderlinePatterns.SINGLE); else run.setUnderline(UnderlinePatterns.NONE);
run.setColor(fontColor); run.setFontSize(fontSize);
}
}
public static void main(String[] args) throws Exception {
String html =
"<p><font size='32' color='#0000FF'><b>First paragraph.</font></b><br/>Just like a heading</p>"
+"<p>This is my text <i>which now is in italic <b>but also in bold</b> depending on its <u>importance</u></i>.<br/>Now a <b><i><u>new</u></i></b> line starts <i>within <b>the same</b> paragraph</i>.</p>"
+"<p><b>Last <u>paragraph <i>comes</u> here</b> finally</i>.</p>"
+"<p>But yet <u><i><b>another</i></u></b> paragraph having <i><font size='22' color='#FF0000'>special <u>font</u> settings</font></i>. Now default font again.</p>";
HTMLtoDOCX htmlToDOCX = new HTMLtoDOCX(html, "./CreateWordParagraphFromHTML.docx");
}
}
Result:
Disclaimer: This is a working draft showing the principle. Neither it is fully ready nor it is code ready for use in productive environments.