Package org.apache.tika.sax

Examples of org.apache.tika.sax.XHTMLContentHandler.startElement()

127
128
129
130
131
132
133
134
135
136
137
            XHTMLContentHandler xhtml =
                new XHTMLContentHandler(handler, metadata);
            xhtml.startDocument();

            xhtml.startElement("p");
            char[] buffer = new char[4096];
            int n = reader.read(buffer);
            while (n != -1) {
                xhtml.characters(buffer, 0, n);
                n = reader.read(buffer);
View Full Code Here
256
257
258
259
260
261
262
263
264
265
266
                // call next handler if no filtering
                if (completelyFiltered == 0) {
                    // special handling of text:h, that are directly passed
                    // to xhtml handler
                    if (TEXT_NS.equals(namespaceURI) && "h".equals(localName)) {
                        xhtml.startElement(headingStack.push(
                                getXHTMLHeaderTagName(atts)));
                    } else {
                        super.startElement(
                                namespaceURI, localName, qName, atts);
                    }
View Full Code Here
55
56
57
58
59
60
61
62
63
64
65
                metadata);

        XHTMLContentHandler xhtml =
            new XHTMLContentHandler(handler, metadata);
        xhtml.startDocument();
        xhtml.startElement("p");
        extractText(filesystem, new AppendableAdaptor(xhtml));
        xhtml.endElement("p");
        xhtml.endDocument();
    }
View Full Code Here
40
41
42
43
44
45
46
47
48
49
50
        Reader reader = Utils.getUTF8Reader(stream, metadata);
        metadata.set(Metadata.CONTENT_TYPE, "text/plain");

        XHTMLContentHandler xhtml = new XHTMLContentHandler(handler, metadata);
        xhtml.startDocument();
        xhtml.startElement("p");
        char[] buffer = new char[4096];
        for (int n = reader.read(buffer); n != -1; n = reader.read(buffer)) {
            xhtml.characters(buffer, 0, n);
        }
        xhtml.endElement("p");
View Full Code Here
100
101
102
103
104
105
106
107
108
109
110
        xp.extractContent(xmlDoc, "nbWord", "//meta:document-statistic/@meta:word-count", metadata);
        xp.extractContent(xmlDoc, "nbcharacter", "//meta:document-statistic/@meta:character-count", metadata);

        XHTMLContentHandler xhtml = new XHTMLContentHandler(handler, metadata);
        xhtml.startDocument();
        xhtml.startElement("p");
        xp.concatOccurrence(xmlDoc, "//*", " ", new AppendableAdaptor(xhtml));
        xhtml.endElement("p");
        xhtml.endDocument();
    }

View Full Code Here
135
136
137
138
139
140
141
142
143
144
145
            XHTMLContentHandler xhtml =
                new XHTMLContentHandler(handler, metadata);
            xhtml.startDocument();

            xhtml.startElement("p");
            char[] buffer = new char[4096];
            int n = reader.read(buffer);
            while (n != -1) {
                xhtml.characters(buffer, 0, n);
                n = reader.read(buffer);
View Full Code Here
79
80
81
82
83
84
85
86
87
88
89
        }

        final XHTMLContentHandler xhtml =
            new XHTMLContentHandler(handler, metadata);
        xhtml.startDocument();
        xhtml.startElement("p");

        getSAXParser(context).parse(
                new CloseShieldInputStream(stream),
                new OfflineContentHandler(
                        getContentHandler(handler, metadata)));
View Full Code Here
115
116
117
118
119
120
121
122
123
124
125
                    // Blank line is signal that we're transitioning to the content.
                    saveHeaderInMetadata(numEmails, metadata, multiLine);
                    parseState = ParseStates.IN_CONTENT;

                    // Mimic what PackageParser does between entries.
                    xhtml.startElement("div", "class", "email-entry");
                    xhtml.startElement("p");
                    inQuote = false;
                } else if (curLine.startsWith(" ") || curLine.startsWith("\t")) {
                    multiLine += " " + curLine.trim();
                } else {
View Full Code Here
116
117
118
119
120
121
122
123
124
125
126
                    saveHeaderInMetadata(numEmails, metadata, multiLine);
                    parseState = ParseStates.IN_CONTENT;

                    // Mimic what PackageParser does between entries.
                    xhtml.startElement("div", "class", "email-entry");
                    xhtml.startElement("p");
                    inQuote = false;
                } else if (curLine.startsWith(" ") || curLine.startsWith("\t")) {
                    multiLine += " " + curLine.trim();
                } else {
                    saveHeaderInMetadata(numEmails, metadata, multiLine);
View Full Code Here
144
145
146
147
148
149
150
151
152
153
154
                        if (!quoted) {
                            xhtml.endElement("q");
                            inQuote = false;
                        }
                    } else if (quoted) {
                        xhtml.startElement("q");
                        inQuote = true;
                    }

                    xhtml.characters(curLine);
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.