Package org.htmlparser.tags

Examples of org.htmlparser.tags.LinkTag

594
595
596
597
598
599
600
601
602
603
604
605
            "<A HREF=\"mailto:this@is.real\">this@is.real</A>",
            "http://www.cj.com/");
        parser.addScanner(new LinkScanner("-l"));
        parseAndAssertNodeCount(1);
        assertTrue("Node should be a HTMLLinkTag", node[0] instanceof LinkTag);
        LinkTag linkTag = (LinkTag) node[0];
        assertEquals(
            "Link Plain Text",
            "this@is.real",
            linkTag.toPlainTextString());
        assertTrue("Link is a mail link", linkTag.isMailLink());
    }
View Full Code Here
610
611
612
613
614
615
616
617
618
619
620
621
622
            "<A HREF=\"javascript:alert('hello');\">say hello</A>",
            "http://www.cj.com/");
        parser.addScanner(new LinkScanner("-l"));
        parseAndAssertNodeCount(1);
        assertTrue("Node should be a HTMLLinkTag", node[0] instanceof LinkTag);
        LinkTag linkTag = (LinkTag) node[0];

        assertEquals(
            "Link Plain Text",
            "say hello",
            linkTag.toPlainTextString());
        assertTrue("Link is a Javascript command", linkTag.isJavascriptLink());
    }
View Full Code Here
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
            "<A HREF=\"javascript_not.html\">say hello</A>",
            "http://www.cj.com/");
        parser.addScanner(new LinkScanner("-l"));
        parseAndAssertNodeCount(1);
        assertTrue("Node should be a HTMLLinkTag", node[0] instanceof LinkTag);
        LinkTag linkTag = (LinkTag) node[0];

        assertEquals(
            "Link Plain Text",
            "say hello",
            linkTag.toPlainTextString());
        assertTrue(
            "Link is not a Javascript command",
            !linkTag.isJavascriptLink());
    }
View Full Code Here
646
647
648
649
650
651
652
653
654
655
            "<A HREF=\"ftp://some.where.it\">my ftp</A>",
            "http://www.cj.com/");
        parser.addScanner(new LinkScanner("-l"));
        parseAndAssertNodeCount(1);
        assertTrue("Node should be a HTMLLinkTag", node[0] instanceof LinkTag);
        LinkTag linkTag = (LinkTag) node[0];

        assertEquals("Link Plain Text", "my ftp", linkTag.toPlainTextString());
        assertTrue("Link is a FTP site", linkTag.isFTPLink());
    }
View Full Code Here
658
659
660
661
662
663
664
665
666
667
    {
        createParser("<A HREF=\"ftp.html\">my ftp</A>", "http://www.cj.com/");
        parser.addScanner(new LinkScanner("-l"));
        parseAndAssertNodeCount(1);
        assertTrue("Node should be a HTMLLinkTag", node[0] instanceof LinkTag);
        LinkTag linkTag = (LinkTag) node[0];

        assertEquals("Link Plain Text", "my ftp", linkTag.toPlainTextString());
        assertTrue("Link is not a FTP site", !linkTag.isFTPLink());
    }
View Full Code Here
672
673
674
675
676
677
678
679
680
681
682
            "<A HREF=\"newpage.html\">New Page</A>",
            "http://www.mysite.com/books/some.asp");
        parser.addScanner(new LinkScanner("-l"));
        parseAndAssertNodeCount(1);
        assertTrue("Node should be a HTMLLinkTag", node[0] instanceof LinkTag);
        LinkTag linkTag = (LinkTag) node[0];
        assertEquals(
            "Link",
            "http://www.mysite.com/books/newpage.html",
            linkTag.getLink());
    }
View Full Code Here
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
            "<a href=\"registration.asp?EventID=1272\"><img border=\"0\" src=\"\\images\\register.gif\"</a>",
            "http://www.fedpage.com/Event.asp?EventID=1272");
        parser.registerScanners();
        parseAndAssertNodeCount(1);
        assertTrue("Node should be a HTMLLinkTag", node[0] instanceof LinkTag);
        LinkTag linkTag = (LinkTag) node[0];
        // Get the image tag from the link

        Node insideNodes[] = new Node[10];
        int j = 0;
        for (SimpleNodeIterator e = linkTag.children(); e.hasMoreNodes();)
        {
            insideNodes[j++] = (Node) e.nextNode();
        }
        assertEquals("Number of contained internal nodes", 1, j);
        assertTrue(insideNodes[0] instanceof ImageTag);
View Full Code Here
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
            "<a class=rlbA href=/news/866201.asp?0sl=-"
                + "32>Shoe bomber handed life sentence</a>");
        parser.registerScanners();
        parseAndAssertNodeCount(1);
        assertType("node type", LinkTag.class, node[0]);
        LinkTag linkTag = (LinkTag) node[0];
        assertStringEquals(
            "link text",
            "Shoe bomber handed life sentence",
            linkTag.getLinkText());
        assertStringEquals(
            "link url",
            "/news/866201.asp?0sl=-32",
            linkTag.getLink());
    }
View Full Code Here
746
747
748
749
750
751
752
753
754
755
756
            "<a href=\"<%=Application(\"sURL\")% "
                + ">/literature/index.htm\">Literature</a>");
        parser.registerScanners();
        parseAndAssertNodeCount(1);
        assertType("should be link tag", LinkTag.class, node[0]);
        LinkTag linkTag = (LinkTag) node[0];
        assertStringEquals(
            "expected link",
            "<%=Application(\"sURL\")%>/literature/index.htm",
            linkTag.getLink());
    }
View Full Code Here
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
    protected URL[] extractLinks(String url) throws ParserException
    {
        Parser parser;
        Vector vector;
        Node node;
        LinkTag link;
        URL[] ret;

        parser = new Parser(url);
        parser.registerScanners();
        ObjectFindingVisitor visitor = new ObjectFindingVisitor(LinkTag.class);
        parser.visitAllNodesWith(visitor);
        Node[] nodes = visitor.getTags();
        vector = new Vector();
        for (int i = 0; i < nodes.length; i++)
            try
            {
                link = (LinkTag) nodes[i];
                vector.add(new URL(link.getLink()));
            }
            catch (MalformedURLException murle)
            {
                //vector.remove (i);
                //i--;
View Full Code Here
TOP

Related Classes of org.htmlparser.tags.LinkTag

  • com.almilli.movierentals.blockbuster.BlockbusterService
  • com.almilli.movierentals.netflix.NetFlixService
  • com.knowgate.hipermail.HtmlMimeBodyPart
  • de.jungblut.crawl.extraction.OutlinkExtractor
  • edu.umd.cloud9.webgraph.ClueExtractLinks$Map
  • edu.umd.cloud9.webgraph.TrecExtractLinks$Map
  • hudson.plugins.rubyMetrics.rcov.RcovParser
  • hudson.plugins.rubyMetrics.saikuro.SaikuroParser
  • modelcrawler.Page
  • net.sf.regain.crawler.preparator.HtmlPreparator
    • 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.