Package org.apache.poi.hdf.extractor

Examples of org.apache.poi.hdf.extractor.WordDocument

38
39
40
41
42
43
44
45
46
47
48
49
50
public void getDocument(String ifile, IndexableDoc doc)
{
  String bodyText = null;
  try
  { logger.info("Using POI classes to extract text from Word document: " + ifile);
  WordDocument wd = new WordDocument( new FileInputStream( new File(ifile) ) );
  StringWriter docTextWriter = new StringWriter();
  wd.writeAllText(new PrintWriter(docTextWriter));   
  docTextWriter.close();

  //*-- if no text was extracted, try antiword
  bodyText = docTextWriter.toString(); bodyText = StringTools.filterChars(bodyText);
  if (bodyText.length() == 0) bodyText = tryAntiword(ifile);
View Full Code Here
27
28
29
30
31
32
33
34
35
36
37
* @author Thierry Templier
*/
public class PoiWordDocumentHandler extends AbstractTypeFileDocumentHandler {

  protected String extractText(InputStream inputStream) throws IOException {
    WordDocument wordDocument = new WordDocument(inputStream);
    StringWriter textWriter = new StringWriter();
    wordDocument.writeAllText(new PrintWriter(textWriter));
    textWriter.close();
    return textWriter.toString();
  }
View Full Code Here
32
33
34
35
36
37
38
39
40
41
42
43
44
  public void buildDocument(InputStream is, Document doc) throws DocumentHandlerException {
    String bodyText = null;

    try {
      WordDocument wordDoc = new WordDocument(is);
      StringWriter sw = new StringWriter();
      wordDoc.writeAllText(sw);
      sw.close();
      bodyText = sw.toString();
    } catch (Exception e) {
      throw new DocumentHandlerException(
          "Cannot extract text from a Word document", e);
View Full Code Here
87
88
89
90
91
92
93
94
95
96
97
98
99
    throws IOException {
        FileInputStream finp = new FileInputStream(doc);
        FileOutputStream fout = new FileOutputStream(output);

        try {
            WordDocument wd = new WordDocument(finp);       
            Writer writer = new OutputStreamWriter(fout, "UTF-16BE");
            wd.writeAllText(writer);
        } finally {
            close(finp);
            close(fout);
        }
    }
View Full Code Here
TOP

Related Classes of org.apache.poi.hdf.extractor.WordDocument

  • net.sf.clairv.index.builder.impl.POIDocBuilder
  • org.apache.poi.hdf.extractor.data.ListTables
  • org.apache.poi.hdf.extractor.data.LVL
  • org.apache.poi.hdf.extractor.util.ChpxNode
  • org.apache.poi.hdf.extractor.util.PapxNode
  • org.apache.poi.hdf.extractor.util.PropertyNode
  • org.apache.poi.hdf.extractor.util.SepxNode
  • org.apache.poi.poifs.filesystem.DocumentEntry
  • org.apache.poi.poifs.filesystem.POIFSFileSystem
  • org.archive.util.ms.DocTest
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.