Package org.apache.tika.parser.xml

Examples of org.apache.tika.parser.xml.XMLParser.extractContent()

83
84
85
86
87
88
89
90
91
92
93
            InputStream stream, ContentHandler handler, Metadata metadata)
            throws IOException, SAXException, TikaException {
        Document xmlDoc = parse(stream);
        XMLParser xp = new XMLParser();
        xp.getAllDocumentNs(xmlDoc);
        xp.extractContent(xmlDoc, Metadata.TITLE, "//dc:title", metadata);
        xp.extractContent(xmlDoc, Metadata.SUBJECT, "//dc:subject", metadata);
        xp.extractContent(xmlDoc, Metadata.CREATOR, "//dc:creator", metadata);
        xp.extractContent(xmlDoc, Metadata.DESCRIPTION, "//dc:description", metadata);
        xp.extractContent(xmlDoc, Metadata.LANGUAGE, "//dc:language", metadata);
        xp.extractContent(xmlDoc, Metadata.KEYWORDS, "//meta:keyword", metadata);
View Full Code Here
84
85
86
87
88
89
90
91
92
93
94
            throws IOException, SAXException, TikaException {
        Document xmlDoc = parse(stream);
        XMLParser xp = new XMLParser();
        xp.getAllDocumentNs(xmlDoc);
        xp.extractContent(xmlDoc, Metadata.TITLE, "//dc:title", metadata);
        xp.extractContent(xmlDoc, Metadata.SUBJECT, "//dc:subject", metadata);
        xp.extractContent(xmlDoc, Metadata.CREATOR, "//dc:creator", metadata);
        xp.extractContent(xmlDoc, Metadata.DESCRIPTION, "//dc:description", metadata);
        xp.extractContent(xmlDoc, Metadata.LANGUAGE, "//dc:language", metadata);
        xp.extractContent(xmlDoc, Metadata.KEYWORDS, "//meta:keyword", metadata);
        xp.extractContent(xmlDoc, Metadata.DATE, "//dc:date", metadata);
View Full Code Here
85
86
87
88
89
90
91
92
93
94
95
        Document xmlDoc = parse(stream);
        XMLParser xp = new XMLParser();
        xp.getAllDocumentNs(xmlDoc);
        xp.extractContent(xmlDoc, Metadata.TITLE, "//dc:title", metadata);
        xp.extractContent(xmlDoc, Metadata.SUBJECT, "//dc:subject", metadata);
        xp.extractContent(xmlDoc, Metadata.CREATOR, "//dc:creator", metadata);
        xp.extractContent(xmlDoc, Metadata.DESCRIPTION, "//dc:description", metadata);
        xp.extractContent(xmlDoc, Metadata.LANGUAGE, "//dc:language", metadata);
        xp.extractContent(xmlDoc, Metadata.KEYWORDS, "//meta:keyword", metadata);
        xp.extractContent(xmlDoc, Metadata.DATE, "//dc:date", metadata);
        xp.extractContent(xmlDoc, "nbTab", "//meta:document-statistic/@meta:table-count", metadata);
View Full Code Here
86
87
88
89
90
91
92
93
94
95
96
        XMLParser xp = new XMLParser();
        xp.getAllDocumentNs(xmlDoc);
        xp.extractContent(xmlDoc, Metadata.TITLE, "//dc:title", metadata);
        xp.extractContent(xmlDoc, Metadata.SUBJECT, "//dc:subject", metadata);
        xp.extractContent(xmlDoc, Metadata.CREATOR, "//dc:creator", metadata);
        xp.extractContent(xmlDoc, Metadata.DESCRIPTION, "//dc:description", metadata);
        xp.extractContent(xmlDoc, Metadata.LANGUAGE, "//dc:language", metadata);
        xp.extractContent(xmlDoc, Metadata.KEYWORDS, "//meta:keyword", metadata);
        xp.extractContent(xmlDoc, Metadata.DATE, "//dc:date", metadata);
        xp.extractContent(xmlDoc, "nbTab", "//meta:document-statistic/@meta:table-count", metadata);
        xp.extractContent(xmlDoc, "nbObject", "//meta:document-statistic/@meta:object-count", metadata);
View Full Code Here
87
88
89
90
91
92
93
94
95
96
97
        xp.getAllDocumentNs(xmlDoc);
        xp.extractContent(xmlDoc, Metadata.TITLE, "//dc:title", metadata);
        xp.extractContent(xmlDoc, Metadata.SUBJECT, "//dc:subject", metadata);
        xp.extractContent(xmlDoc, Metadata.CREATOR, "//dc:creator", metadata);
        xp.extractContent(xmlDoc, Metadata.DESCRIPTION, "//dc:description", metadata);
        xp.extractContent(xmlDoc, Metadata.LANGUAGE, "//dc:language", metadata);
        xp.extractContent(xmlDoc, Metadata.KEYWORDS, "//meta:keyword", metadata);
        xp.extractContent(xmlDoc, Metadata.DATE, "//dc:date", metadata);
        xp.extractContent(xmlDoc, "nbTab", "//meta:document-statistic/@meta:table-count", metadata);
        xp.extractContent(xmlDoc, "nbObject", "//meta:document-statistic/@meta:object-count", metadata);
        xp.extractContent(xmlDoc, "nbImg", "//meta:document-statistic/@meta:image-count", metadata);
View Full Code Here
88
89
90
91
92
93
94
95
96
97
98
        xp.extractContent(xmlDoc, Metadata.TITLE, "//dc:title", metadata);
        xp.extractContent(xmlDoc, Metadata.SUBJECT, "//dc:subject", metadata);
        xp.extractContent(xmlDoc, Metadata.CREATOR, "//dc:creator", metadata);
        xp.extractContent(xmlDoc, Metadata.DESCRIPTION, "//dc:description", metadata);
        xp.extractContent(xmlDoc, Metadata.LANGUAGE, "//dc:language", metadata);
        xp.extractContent(xmlDoc, Metadata.KEYWORDS, "//meta:keyword", metadata);
        xp.extractContent(xmlDoc, Metadata.DATE, "//dc:date", metadata);
        xp.extractContent(xmlDoc, "nbTab", "//meta:document-statistic/@meta:table-count", metadata);
        xp.extractContent(xmlDoc, "nbObject", "//meta:document-statistic/@meta:object-count", metadata);
        xp.extractContent(xmlDoc, "nbImg", "//meta:document-statistic/@meta:image-count", metadata);
        xp.extractContent(xmlDoc, "nbPage", "//meta:document-statistic/@meta:page-count", metadata);
View Full Code Here
89
90
91
92
93
94
95
96
97
98
99
        xp.extractContent(xmlDoc, Metadata.SUBJECT, "//dc:subject", metadata);
        xp.extractContent(xmlDoc, Metadata.CREATOR, "//dc:creator", metadata);
        xp.extractContent(xmlDoc, Metadata.DESCRIPTION, "//dc:description", metadata);
        xp.extractContent(xmlDoc, Metadata.LANGUAGE, "//dc:language", metadata);
        xp.extractContent(xmlDoc, Metadata.KEYWORDS, "//meta:keyword", metadata);
        xp.extractContent(xmlDoc, Metadata.DATE, "//dc:date", metadata);
        xp.extractContent(xmlDoc, "nbTab", "//meta:document-statistic/@meta:table-count", metadata);
        xp.extractContent(xmlDoc, "nbObject", "//meta:document-statistic/@meta:object-count", metadata);
        xp.extractContent(xmlDoc, "nbImg", "//meta:document-statistic/@meta:image-count", metadata);
        xp.extractContent(xmlDoc, "nbPage", "//meta:document-statistic/@meta:page-count", metadata);
        xp.extractContent(xmlDoc, "nbPara", "//meta:document-statistic/@meta:paragraph-count", metadata);
View Full Code Here
90
91
92
93
94
95
96
97
98
99
100
        xp.extractContent(xmlDoc, Metadata.CREATOR, "//dc:creator", metadata);
        xp.extractContent(xmlDoc, Metadata.DESCRIPTION, "//dc:description", metadata);
        xp.extractContent(xmlDoc, Metadata.LANGUAGE, "//dc:language", metadata);
        xp.extractContent(xmlDoc, Metadata.KEYWORDS, "//meta:keyword", metadata);
        xp.extractContent(xmlDoc, Metadata.DATE, "//dc:date", metadata);
        xp.extractContent(xmlDoc, "nbTab", "//meta:document-statistic/@meta:table-count", metadata);
        xp.extractContent(xmlDoc, "nbObject", "//meta:document-statistic/@meta:object-count", metadata);
        xp.extractContent(xmlDoc, "nbImg", "//meta:document-statistic/@meta:image-count", metadata);
        xp.extractContent(xmlDoc, "nbPage", "//meta:document-statistic/@meta:page-count", metadata);
        xp.extractContent(xmlDoc, "nbPara", "//meta:document-statistic/@meta:paragraph-count", metadata);
        xp.extractContent(xmlDoc, "nbWord", "//meta:document-statistic/@meta:word-count", metadata);
View Full Code Here
91
92
93
94
95
96
97
98
99
100
101
        xp.extractContent(xmlDoc, Metadata.DESCRIPTION, "//dc:description", metadata);
        xp.extractContent(xmlDoc, Metadata.LANGUAGE, "//dc:language", metadata);
        xp.extractContent(xmlDoc, Metadata.KEYWORDS, "//meta:keyword", metadata);
        xp.extractContent(xmlDoc, Metadata.DATE, "//dc:date", metadata);
        xp.extractContent(xmlDoc, "nbTab", "//meta:document-statistic/@meta:table-count", metadata);
        xp.extractContent(xmlDoc, "nbObject", "//meta:document-statistic/@meta:object-count", metadata);
        xp.extractContent(xmlDoc, "nbImg", "//meta:document-statistic/@meta:image-count", metadata);
        xp.extractContent(xmlDoc, "nbPage", "//meta:document-statistic/@meta:page-count", metadata);
        xp.extractContent(xmlDoc, "nbPara", "//meta:document-statistic/@meta:paragraph-count", metadata);
        xp.extractContent(xmlDoc, "nbWord", "//meta:document-statistic/@meta:word-count", metadata);
        xp.extractContent(xmlDoc, "nbcharacter", "//meta:document-statistic/@meta:character-count", metadata);
View Full Code Here
92
93
94
95
96
97
98
99
100
101
102
        xp.extractContent(xmlDoc, Metadata.LANGUAGE, "//dc:language", metadata);
        xp.extractContent(xmlDoc, Metadata.KEYWORDS, "//meta:keyword", metadata);
        xp.extractContent(xmlDoc, Metadata.DATE, "//dc:date", metadata);
        xp.extractContent(xmlDoc, "nbTab", "//meta:document-statistic/@meta:table-count", metadata);
        xp.extractContent(xmlDoc, "nbObject", "//meta:document-statistic/@meta:object-count", metadata);
        xp.extractContent(xmlDoc, "nbImg", "//meta:document-statistic/@meta:image-count", metadata);
        xp.extractContent(xmlDoc, "nbPage", "//meta:document-statistic/@meta:page-count", metadata);
        xp.extractContent(xmlDoc, "nbPara", "//meta:document-statistic/@meta:paragraph-count", metadata);
        xp.extractContent(xmlDoc, "nbWord", "//meta:document-statistic/@meta:word-count", metadata);
        xp.extractContent(xmlDoc, "nbcharacter", "//meta:document-statistic/@meta:character-count", metadata);
View Full Code Here
TOP
Copyright © 2018 www.massapi.com. 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.