Example usage for org.apache.poi.xwpf.usermodel XWPFPictureData getFileName

List of usage examples for org.apache.poi.xwpf.usermodel XWPFPictureData getFileName

  1. HOME
  2. Java
  3. org.apache.poi
  4. org.apache.poi.xwpf.usermodel.*
  5. XWPFPictureData
  6. getFileName

Introduction

In this page you can find the example usage for org.apache.poi.xwpf.usermodel XWPFPictureData getFileName.

Prototype

public String getFileName()

Source Link

Document

Returns the file name of the image, eg image7.jpg .

Usage

From source file:demo.poi.image.SimpleImages.java

License:Apache License

public static List<byte[]> extractImagesFromWord(InputStream file) {
    try {/* ww w. j a  va2  s .co  m*/
        List<byte[]> result = new ArrayList<byte[]>();
        XWPFDocument doc = new XWPFDocument(file);
        log.debug("{}", doc);

        for (XWPFPictureData picture : doc.getAllPictures()) {
            result.add(picture.getData());
            log.debug("{}", picture.getFileName());
        }

        return result;
    } catch (Exception e) {
        throw new RuntimeException(e);
    }
}

From source file:fr.opensagres.poi.xwpf.converter.core.XWPFDocumentVisitor.java

License:Open Source License

private void visitGraphicalObject(T parentContainer, CTGraphicalObject graphic, Float offsetX,
        STRelFromH.Enum relativeFromH, Float offsetY, STRelFromV.Enum relativeFromV, STWrapText.Enum wrapText)
        throws Exception {
    if (graphic != null) {
        CTGraphicalObjectData graphicData = graphic.getGraphicData();
        if (graphicData != null) {
            XmlCursor c = graphicData.newCursor();
            c.selectPath("./*");
            while (c.toNextSelection()) {
                XmlObject o = c.getObject();
                if (o instanceof CTPicture) {
                    CTPicture picture = (CTPicture) o;
                    // extract the picture if needed
                    IImageExtractor extractor = getImageExtractor();
                    if (extractor != null) {
                        XWPFPictureData pictureData = getPictureData(picture);
                        if (pictureData != null) {
                            try {
                                extractor.extract(WORD_MEDIA + pictureData.getFileName(),
                                        pictureData.getData());
                            } catch (Throwable e) {
                                LOGGER.log(Level.SEVERE,
                                        "Error while extracting the image " + pictureData.getFileName(), e);
                            }/*from  w ww  .  j  a  va  2 s.co  m*/
                        }
                    }
                    // visit the picture.
                    visitPicture(picture, offsetX, relativeFromH, offsetY, relativeFromV, wrapText,
                            parentContainer);
                }
            }
            c.dispose();
        }
    }
}

From source file:fr.opensagres.poi.xwpf.converter.xhtml.internal.XHTMLMapper.java

License:Open Source License

@Override
protected void visitPicture(CTPicture picture, Float offsetX, Enum relativeFromH, Float offsetY,
        org.openxmlformats.schemas.drawingml.x2006.wordprocessingDrawing.STRelFromV.Enum relativeFromV,
        org.openxmlformats.schemas.drawingml.x2006.wordprocessingDrawing.STWrapText.Enum wrapText,
        Object parentContainer) throws Exception {

    AttributesImpl attributes = null;
    // Src attribute
    XWPFPictureData pictureData = super.getPictureData(picture);
    if (pictureData != null) {
        // img/@src
        String src = pictureData.getFileName();
        if (StringUtils.isNotEmpty(src)) {
            src = resolver.resolve(WORD_MEDIA + src);
            attributes = SAXHelper.addAttrValue(attributes, SRC_ATTR, src);
        }/*www  .j  a  v a 2 s. c om*/

        CTPositiveSize2D ext = picture.getSpPr().getXfrm().getExt();
        CSSStyle style = new CSSStyle(IMG_ELEMENT, null);
        // img/@width
        float width = emu2points(ext.getCx());
        // img/@height
        float height = emu2points(ext.getCy());
        style.addProperty(WIDTH, getStylesDocument().getValueAsPoint(width));
        style.addProperty(HEIGHT, getStylesDocument().getValueAsPoint(height));
        attributes = SAXHelper.addAttrValue(attributes, STYLE_ATTR, style.getInlineStyles());
    } else {
        // external link images inserted
        String link = picture.getBlipFill().getBlip().getLink();
        String src = document.getPackagePart().getRelationships().getRelationshipByID(link).getTargetURI()
                .toString();
        attributes = SAXHelper.addAttrValue(null, SRC_ATTR, src);

        CTPositiveSize2D ext = picture.getSpPr().getXfrm().getExt();
        CSSStyle style = new CSSStyle(IMG_ELEMENT, null);
        // img/@width
        float width = emu2points(ext.getCx());
        // img/@height
        float height = emu2points(ext.getCy());
        style.addProperty(WIDTH, getStylesDocument().getValueAsPoint(width));
        style.addProperty(HEIGHT, getStylesDocument().getValueAsPoint(height));
        attributes = SAXHelper.addAttrValue(attributes, STYLE_ATTR, style.getInlineStyles());
    }
    if (attributes != null) {
        startElement(IMG_ELEMENT, attributes);
        endElement(IMG_ELEMENT);
    }
}

From source file:kz.service.DocumentReader.java

public static String readDocxFile(String fileName) {

    try {/*from  ww w.j ava  2 s.  c  om*/
        File file = new File(fileName);
        FileInputStream fis = new FileInputStream(file.getAbsolutePath());
        StringBuffer content = new StringBuffer();

        XWPFDocument document = new XWPFDocument(fis);
        XWPFStyles styles = document.getStyles();

        List<XWPFParagraph> paragraphs = document.getParagraphs();
        List<XWPFTable> tables = document.getTables();
        List<XWPFPictureData> pictures = document.getAllPictures();

        //int Picture_ID = 0;
        for (XWPFPictureData picture : pictures) {
            //XWPFPictureData picture = pictures.get(Picture_ID);
            System.out.println("Picture: " + picture.getFileName());
            byte[] pictureData = picture.getData();
            BufferedImage image = ImageIO.read(new ByteArrayInputStream(pictureData));
            ImageIO.write(image, picture.getFileName(), file);
            content.append("<p>");
            content.append("Here must be image");
            content.append("</p>");
            //Picture_ID++;
        }

        Iterator<IBodyElement> bodyElementIterator = document.getBodyElementsIterator();
        int Table_ID = 0;
        int Paragraph_ID = 0;
        while (bodyElementIterator.hasNext()) {

            IBodyElement element = bodyElementIterator.next();
            System.out.println(element.getElementType().name());//prints Element type name

            if ("TABLE".equalsIgnoreCase(element.getElementType().name())) {

                content.append("<table>");
                XWPFTable table = tables.get(Table_ID);
                CTTbl cttbl = table.getCTTbl();
                CTTblPr cttblPr = cttbl.getTblPr();

                List<XWPFTableRow> tblRows = table.getRows();
                for (XWPFTableRow tblRow : tblRows) {
                    content.append("<tr>");
                    List<XWPFTableCell> tblCells = tblRow.getTableCells();
                    for (XWPFTableCell tblCell : tblCells) {
                        content.append("<td>");
                        content.append(tblCell.getText());
                        content.append("</td>");
                    }
                    content.append("</tr>");
                }
                content.append("</table>");
                Table_ID++;

            } else if ("PARAGRAPH".equalsIgnoreCase(element.getElementType().name())) {

                XWPFParagraph paragraph = paragraphs.get(Paragraph_ID);

                String styleClass = null;
                if (paragraph.getStyleID() != null) {
                    content.append("<p class=''>");
                    XWPFStyle style = styles.getStyle(paragraph.getStyleID());
                    if (style != null && style.getName() != null) {
                        //here will be code creation of tag with class style
                    }
                } else {
                    content.append("<p>");
                }
                content.append(paragraph.getText());
                content.append("</p>");
                Paragraph_ID++;

            }
        }

        fis.close();
        return content.toString();
    } catch (Exception e) {
        return e.toString();
    }

}

From source file:mj.ocraptor.extraction.tika.parser.microsoft.ooxml.XWPFWordExtractorDecorator.java

License:Apache License

private TmpFormatting processRun(XWPFRun run, XWPFParagraph paragraph, XHTMLContentHandler xhtml,
        TmpFormatting tfmtg) throws SAXException, XmlException, IOException {
    // True if we are currently in the named style tag:
    if (run.isBold() != tfmtg.isBold()) {
        if (tfmtg.isItalic()) {
            xhtml.endElement("i");
            tfmtg.setItalic(false);//from   w ww  . j  a va2  s.  co  m
        }
        if (run.isBold()) {
            xhtml.startElement("b");
        } else {
            xhtml.endElement("b");
        }
        tfmtg.setBold(run.isBold());
    }

    if (run.isItalic() != tfmtg.isItalic()) {
        if (run.isItalic()) {
            xhtml.startElement("i");
        } else {
            xhtml.endElement("i");
        }
        tfmtg.setItalic(run.isItalic());
    }

    boolean addedHREF = false;
    if (run instanceof XWPFHyperlinkRun) {
        XWPFHyperlinkRun linkRun = (XWPFHyperlinkRun) run;
        XWPFHyperlink link = linkRun.getHyperlink(document);
        if (link != null && link.getURL() != null) {
            xhtml.startElement("a", "href", link.getURL());
            addedHREF = true;
        } else if (linkRun.getAnchor() != null && linkRun.getAnchor().length() > 0) {
            xhtml.startElement("a", "href", "#" + linkRun.getAnchor());
            addedHREF = true;
        }
    }

    xhtml.characters(run.toString());

    // If we have any pictures, output them
    for (XWPFPicture picture : run.getEmbeddedPictures()) {
        if (paragraph.getDocument() != null) {
            XWPFPictureData data = picture.getPictureData();
            if (data != null) {
                AttributesImpl attr = new AttributesImpl();

                attr.addAttribute("", "src", "src", "CDATA", "embedded:" + data.getFileName());
                attr.addAttribute("", "alt", "alt", "CDATA", picture.getDescription());

                xhtml.startElement("img", attr);
                xhtml.endElement("img");
            }
        }
    }

    if (addedHREF) {
        xhtml.endElement("a");
    }

    return tfmtg;
}

From source file:org.sleuthkit.autopsy.imageExtractor.ImageExtractor.java

private List<ExtractedImage> extractImagesFromDocx(AbstractFile af) {
    // check for BBArtifact ENCRYPTION_DETECTED? Might be detected elsewhere...?
    // TODO check for BBArtifact ENCRYPTION_DETECTED? Might be detected elsewhere...?
    List<ExtractedImage> listOfExtractedImages = new ArrayList<ExtractedImage>();
    String parentFileName = getUniqueName(af);
    XWPFDocument docxA = null;/*from  w ww. j  a v  a2 s.  co  m*/
    try {
        docxA = new XWPFDocument(new ReadContentInputStream(af));
    } catch (IOException ex) {
        logger.log(Level.WARNING,
                "XWPFDocument container could not be instantiated while reading " + af.getName(), ex);
        return null;
    }
    List<XWPFPictureData> listOfAllPictures = docxA.getAllPictures();

    // if no images are extracted from the ppt, return null, else initialize
    // the output folder for image extraction.
    String outputFolderPath;
    if (listOfAllPictures.isEmpty()) {
        return null;
    } else {
        outputFolderPath = getOutputFolderPath(parentFileName);
    }
    if (outputFolderPath == null) {
        logger.log(Level.WARNING, "Could not get path for image extraction from AbstractFile: {0}",
                af.getName());
        return null;
    }
    for (XWPFPictureData xwpfPicture : listOfAllPictures) {
        String fileName = xwpfPicture.getFileName();
        FileOutputStream fos = null;
        try {
            fos = new FileOutputStream(outputFolderPath + File.separator + fileName);
        } catch (FileNotFoundException ex) {
            logger.log(Level.WARNING, "Invalid path provided for image extraction", ex);
            continue;
        }
        try {
            fos.write(xwpfPicture.getData());
            fos.close();
        } catch (IOException ex) {
            logger.log(Level.WARNING, "Could not write to the provided location", ex);
            continue;
        }
        String fileRelativePath = File.separator + moduleDirRelative + File.separator + parentFileName
                + File.separator + fileName;
        long size = xwpfPicture.getData().length;
        ExtractedImage extractedimage = new ExtractedImage(fileName, fileRelativePath, size, af);
        listOfExtractedImages.add(extractedimage);
    }
    return listOfExtractedImages;
}

From source file:org.sleuthkit.autopsy.modules.embeddedfileextractor.ImageExtractor.java

License:Open Source License

/**
 * Extract images from docx format files.
 *
 * @param af the file from which images are to be extracted.
 *
 * @return list of extracted images. Returns null in case no images were
 *         extracted./*w  w w . j  a  v  a  2s.c  om*/
 */
private List<ExtractedImage> extractImagesFromDocx(AbstractFile af) {
    List<ExtractedImage> listOfExtractedImages;
    XWPFDocument docx = null;
    try {
        docx = new XWPFDocument(new ReadContentInputStream(af));
    } catch (Throwable ex) {
        // instantiating POI containers throw RuntimeExceptions
        logger.log(Level.WARNING, NbBundle.getMessage(this.getClass(),
                "EmbeddedFileExtractorIngestModule.ImageExtractor.docxContainer.init.err", af.getName()), ex); //NON-NLS
        return null;
    }
    List<XWPFPictureData> listOfAllPictures = null;
    try {
        listOfAllPictures = docx.getAllPictures();
    } catch (Exception ex) {
        // log internal Java and Apache errors as WARNING
        logger.log(Level.WARNING, NbBundle.getMessage(this.getClass(),
                "EmbeddedFileExtractorIngestModule.ImageExtractor.processing.err", af.getName()), ex); //NON-NLS
        return null;
    }

    // if no images are extracted from the PPT, return null, else initialize
    // the output folder for image extraction.
    String outputFolderPath;
    if (listOfAllPictures.isEmpty()) {
        return null;
    } else {
        outputFolderPath = getOutputFolderPath(this.parentFileName);
    }
    if (outputFolderPath == null) {
        logger.log(Level.WARNING, NbBundle.getMessage(this.getClass(),
                "EmbeddedFileExtractorIngestModule.ImageExtractor.extractImageFrom.outputPath.exception.msg",
                af.getName())); //NON-NLS
        return null;
    }
    listOfExtractedImages = new ArrayList<>();
    byte[] data = null;
    for (XWPFPictureData xwpfPicture : listOfAllPictures) {
        String fileName = xwpfPicture.getFileName();
        try {
            data = xwpfPicture.getData();
        } catch (Exception ex) {
            // log internal Java and Apache errors as WARNING
            logger.log(Level.WARNING,
                    NbBundle.getMessage(this.getClass(),
                            "EmbeddedFileExtractorIngestModule.ImageExtractor.processing.err", af.getName()),
                    ex); //NON-NLS
            return null;
        }
        writeExtractedImage(Paths.get(outputFolderPath, fileName).toString(), data);
        listOfExtractedImages.add(
                new ExtractedImage(fileName, getFileRelativePath(fileName), xwpfPicture.getData().length, af));
    }
    return listOfExtractedImages;
}