In need of a clear example on how to get the word count of DOC and DOCX files

Question

I am able to read a DOC file and get its word count, BUT it is wrong.

My code:

 public class WordCounter {
    public static void main(String[] args) throws Throwable {
        processDOC();
    }

    private static void processDOC() throws Throwable {
        File file = new File("/Users/yjiang/Desktop/whatever.doc");
        File file2 = new File("/Users/yjiang/Desktop/Test.docx");
        File file3 = new File("/Users/yjiang/Desktop/QB Tests 4-14-2014.xls");
        File file4 = new File("/Users/yjiang/Desktop/QB Tests 4-14-2014.xlsx");

        try {
            FileInputStream fs = new FileInputStream(file);
            POIFSFileSystem poifsFileSystem = new POIFSFileSystem(fs);
            DirectoryEntry directoryEntry = poifsFileSystem.getRoot();
            DocumentEntry documentEntry = (DocumentEntry) directoryEntry.getEntry(SummaryInformation.DEFAULT_STREAM_NAME);
            DocumentInputStream dis = new DocumentInputStream(documentEntry);
            PropertySet ps = new PropertySet(dis);
            SummaryInformation si = new SummaryInformation(ps);

            System.out.println(si.getWordCount());
        } catch (Exception e) {
            e.printStackTrace();
        }


        try {
            HWPFDocument hwpfDocument = new HWPFDocument(new FileInputStream(file));
            System.out.println(hwpfDocument.getDocProperties().getCWords()); // actually 71 words using word count in MSWord, returned 57.
            System.out.println(hwpfDocument.getDocProperties().getCWordsFtnEnd());
            XWPFDocument xwpfDocument = new XWPFDocument(new FileInputStream(file2)); // actually 71 words using word count in MSWord, returned 57.
            System.out.println(xwpfDocument.getProperties().getExtendedProperties().getUnderlyingProperties().getWords());



            System.out.println();
        } catch (Exception e) {
            e.printStackTrace();
        }
    }
}

"whatever.doc" has 71 words, when I run this, it returns only 57.

enter image description here

Seems I cannot use the same method to read DOCX files, when I run it I get the following:

org.apache.poi.poifs.filesystem.OfficeXmlFileException: The supplied data appears to be in the Office 2007+ XML. You are calling the part of POI that deals with OLE2 Office Documents. You need to call a different part of POI to process this data (eg XSSF instead of HSSF)

Could provide an example?

If you ask Word to update the document statistics, does POI then see the correct values? — Gagravarr, May 06 '14 at 05:12

score 1 · Answer 1 · answered Apr 13 '18 at 22:21

I've also found that the built-in word counters give strange counts, but text extraction seems to be more reliable, so I use this solution:

public long getWordCount(File file) throws IOException {
    POITextExtractor textExtractor;
    if (file.getName().endsWith(".docx")) {
        XWPFDocument doc = new XWPFDocument(new FileInputStream(file));
        textExtractor = new XWPFWordExtractor(doc);
    }
    else if (file.getName().endsWith(".doc")) {
        textExtractor = new WordExtractor(new FileInputStream(file));
    }
    else {
        throw new IllegalArgumentException("Not a MS Word file.");
    }

    return Arrays.stream(textExtractor.getText().split("\\s+"))
     .filter(s -> s.matches("^.*[\\p{L}\\p{N}].*$"))
     .count();
}

The regex at the bottom can be adjusted if needed, but overall this one has proved fairly resilient.

In need of a clear example on how to get the word count of DOC and DOCX files

1 Answers1