Package org.apache.tika.sax.xpath

Examples of org.apache.tika.sax.xpath.Matcher

64
65
66
67
68
69
70
71
72
73
74
        return new TeeContentHandler(ch, branch);
    }

    private static ContentHandler getStatistic(
            ContentHandler ch, Metadata md, String name, String attribute) {
        Matcher matcher =
            META_XPATH.parse("//meta:document-statistic/@meta:"+attribute);
        ContentHandler branch = new MatchingContentHandler(
              new AttributeMetadataHandler(META_NS, attribute, md, name), matcher);
        return new TeeContentHandler(ch, branch);
    }
View Full Code Here
42
43
44
45
46
47
48
49
50
51
52
    private static final String META_NS = "urn:oasis:names:tc:opendocument:xmlns:meta:1.0";
    private static final XPathParser META_XPATH = new XPathParser("meta", META_NS);

    private static ContentHandler getMeta(
            ContentHandler ch, Metadata md, String name, String element) {
        Matcher matcher = new CompositeMatcher(
                META_XPATH.parse("//meta:" + element),
                META_XPATH.parse("//meta:" + element + "//text()"));
        ContentHandler branch =
            new MatchingContentHandler(new MetadataHandler(md, name), matcher);
        return new TeeContentHandler(ch, branch);
View Full Code Here
52
53
54
55
56
57
58
59
60
61
62
        return new TeeContentHandler(ch, branch);
    }

    private static ContentHandler getUserDefined(
            ContentHandler ch, Metadata md) {
        Matcher matcher = new CompositeMatcher(
                META_XPATH.parse("//meta:user-defined/@meta:name"),
                META_XPATH.parse("//meta:user-defined//text()"));
        // eg <meta:user-defined meta:name="Info1">Text1</meta:user-defined> becomes custom:Info1=Text1
        ContentHandler branch = new MatchingContentHandler(
              new AttributeDependantMetadataHandler(md, "meta:name", Metadata.USER_DEFINED_METADATA_NAME_PREFIX),
View Full Code Here
64
65
66
67
68
69
70
71
72
73
74
        return new TeeContentHandler(ch, branch);
    }

    private static ContentHandler getStatistic(
            ContentHandler ch, Metadata md, String name, String attribute) {
        Matcher matcher =
            META_XPATH.parse("//meta:document-statistic/@meta:"+attribute);
        ContentHandler branch = new MatchingContentHandler(
              new AttributeMetadataHandler(META_NS, attribute, md, name), matcher);
        return new TeeContentHandler(ch, branch);
    }
View Full Code Here
33
34
35
36
37
38
39
40
41
42
43
    private static final XPathParser DC_XPATH = new XPathParser(
            "dc", "http://purl.org/dc/elements/1.1/");

    private static ContentHandler getDublinCore(
            ContentHandler ch, Metadata md, String name, String element) {
        Matcher matcher = new CompositeMatcher(
                DC_XPATH.parse("//dc:" + element),
                DC_XPATH.parse("//dc:" + element + "//text()"));
        ContentHandler branch =
            new MatchingContentHandler(new MetadataHandler(md, name), matcher);
        return new TeeContentHandler(ch, branch);
View Full Code Here
98
99
100
101
102
103
104
105
106
107
108
     *  as HTML as a string, excluding the rest
     */
    public String parseOnePartToHTML() throws IOException, SAXException, TikaException {
        // Only get things under html -> body -> div (class=header)
        XPathParser xhtmlParser = new XPathParser("xhtml", XHTMLContentHandler.XHTML);
        Matcher divContentMatcher = xhtmlParser.parse(
                "/xhtml:html/xhtml:body/xhtml:div/descendant::node()");       
        ContentHandler handler = new MatchingContentHandler(
                new ToXMLContentHandler(), divContentMatcher);
       
        InputStream stream = ContentHandlerExample.class.getResourceAsStream("test2.doc");
View Full Code Here
TOP

Related Classes of org.apache.tika.sax.xpath.Matcher

  • com.cloudera.cdk.morphline.solrcell.SolrCellBuilder$SolrCell
  • org.apache.solr.handler.extraction.ExtractingDocumentLoader
  • org.apache.tika.example.ContentHandlerExample
  • org.apache.tika.parser.html.HtmlParser
  • org.apache.tika.parser.odf.OpenDocumentMetaParser
  • org.apache.tika.parser.opendocument.OpenOfficeMetaParser
  • org.apache.tika.parser.xml.DcXMLParser
  • org.kitesdk.morphline.solrcell.SolrCellBuilder$SolrCell
    • 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.