Review

16

com/itextpdf/text/pdf/parser/TaggedPdfReaderTool.java

parseTag(String, PdfObject, PdfDictionary)

class TaggedPdfReaderTool {
	/**
	 * Searches for a tag in a page.
	 * 
	 * @param tag
	 *            the name of the tag
	 * @param object
	 *            an identifier to find the marked content
	 * @param page
	 *            a page dictionary
	 * @throws IOException
	 */
	public void parseTag(String tag, PdfObject object, PdfDictionary page)
			throws IOException {
		// if the identifier is a number, we can extract the content right away
		if (object instanceof PdfNumber) {
			PdfNumber mcid = (PdfNumber) object;
			RenderFilter filter = new MarkedContentRenderFilter(mcid.intValue());
			TextExtractionStrategy strategy = new SimpleTextExtractionStrategy();
			FilteredTextRenderListener listener = new FilteredTextRenderListener(
					strategy, filter);
			PdfContentStreamProcessor processor = new PdfContentStreamProcessor(
					listener);
			processor.processContent(PdfReader.getPageContent(page), page
					.getAsDict(PdfName.RESOURCES));
			out.print(XMLUtil.escapeXML(listener.getResultantText(), true));
		}
		// if the identifier is an array, we call the parseTag method
		// recursively
		else if (object instanceof PdfArray) {
			PdfArray arr = (PdfArray) object;
			int n = arr.size();
			for (int i = 0; i < n; i++) {
				parseTag(tag, arr.getPdfObject(i), page);
				if (i < n - 1)
					out.println();
			}
		}
		// if the identifier is a dictionary, we get the resources from the
		// dictionary
		else if (object instanceof PdfDictionary) {
			PdfDictionary mcr = (PdfDictionary) object;
			parseTag(tag, mcr.getDirectObject(PdfName.MCID), mcr
					.getAsDict(PdfName.PG));
		}
	}

}