Package net.sf.jpluck.xml

Examples of net.sf.jpluck.xml.TextRecordResult

62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
        TextRecord textRecord = new TextRecord(((redirectionURI != null) ? redirectionURI : uri),
                                               jxlDocument.getOutputEncoding(), jxlDocument.isUseHiresMargins());
        if (redirectionURI != null) {
            textRecord.setAlternateURI(uri);
        }
        TextRecordResult result = new TextRecordResult(textRecord, jxlDocument.getURIRewriter(),
                                                       jxlDocument.getTextColorBrightness(),
                                                       jxlDocument.isParseTables(), pluckerDocument);
        Transformer transformer = TransformerFactory.newInstance().newTransformer();
        transformer.transform(new DOMSource(node), result);

        pluckerDocument.addRecord(textRecord);

        String[] linkURIs = result.getLinkURIs();
        for (int i = 0; i < linkURIs.length; i++) {
            String linkURI = linkURIs[i];
            addPageLink(linkURI);
        }

        String[] imageURIs = result.getEmbeddedImageURIs();
        for (int i = 0; i < imageURIs.length; i++) {
            String imageURI = imageURIs[i];
            addImageLink(imageURI);
        }
       
        // Add bookmarks defined by the page.
        Bookmark[] bookmarks = result.getBookmarks();
        for (int i = 0; i < bookmarks.length; i++) {
      Bookmark bookmark = bookmarks[i];
      pluckerDocument.addBookmark(bookmark);
    }
       
    // Process AutoBookmarks
        BookmarkProcessor processor = jxlDocument.getBookmarkProcessor();
        String title = result.getTitle();
        processor.process(pluckerDocument, uri, title);
    }
View Full Code Here
99
100
101
102
103
104
105
106
107
108
109
        if (mimeType.equals("text/html")) {
      Resource resource = new Resource("http://127.0.0.1/", null, new ContentType(mimeType), data, 0, false);
      org.w3c.dom.Document dom = resource.parseHTML();
      TextRecord textRecord = new TextRecord("http://127.0.0.1/");
      Transformer transformer = TransformerFactory.newInstance().newTransformer();
      TextRecordResult result = new TextRecordResult(textRecord, new URIRewriter(), 100, false, document);
      transformer.transform(new DOMSource(dom), result);
      document.addRecord(textRecord);
        } else if (mimeType.startsWith("image/")) {
      BufferedImage image = ImageIO.read(new ByteArrayInputStream(data));
      CompositeBitmap compositeBitmap = CompositeBitmap.create(image, bpp);
View Full Code Here
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
      }

      Transformer transformer = TransformerFactory.newInstance().newTransformer();
      TextRecord textRecord = resource.createTextRecord(jxlDocument.getOutputEncoding(),
                                jxlDocument.isUseHiresMargins());
      TextRecordResult result = new TextRecordResult(textRecord, jxlDocument.getURIRewriter(),
                               settings.getTextColorBrightness(), settings.isParseTables(),
                               pluckerDocument);
      transformer.transform(new DOMSource(formatter.getIndexDocument()), result);
      pluckerDocument.addRecord(textRecord);
      if (settings.isUseLinkStructure() || settings.isUseListStructure()) {
        String[] pageURIs = result.getLinkURIs();
        for (int i = 0; i < pageURIs.length; i++) {
          addPageLink(pageURIs[i]);
        }
      }

      String[] imageURIs = result.getEmbeddedImageURIs();
      for (int i = 0; i < imageURIs.length; i++) {
        addImageLink(imageURIs[i]);
      }
      if (ClientConfiguration.getDefault().isUseFeedDate() && (feed.getItems().length > 0)) {
        Date date = feed.getItems()[0].getDateIssued();
View Full Code Here
TOP

Related Classes of net.sf.jpluck.xml.TextRecordResult

  • net.sf.jpluck.apps.cmdline.Convert
  • net.sf.jpluck.handlers.DOMHandler
  • net.sf.jpluck.handlers.FeedHandler
  • net.sf.jpluck.plucker.parsing.html.HTMLSerializer
    • 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.