001 /*
002 * (c) Copyright 2009 University of Bristol
003 * All rights reserved.
004 * [See end of file]
005 */
006
007 package net.rootdev.javardfa;
008
009 import java.io.OutputStream;
010 import java.io.OutputStreamWriter;
011 import java.io.PrintWriter;
012 import java.io.UnsupportedEncodingException;
013 import java.io.Writer;
014 import java.util.regex.Pattern;
015
016
017 /**
018 * A pretty ropey NTriple serialiser.
019 * Advantages: streams, no dependencies.
020 *
021 * @author pldms
022 */
023 public class NTripleSink implements StatementSink {
024 private final PrintWriter out;
025 private final String[] comments;
026
027 public NTripleSink(OutputStream os, String... comments) throws UnsupportedEncodingException {
028 this(new OutputStreamWriter(os, "US-ASCII"), comments); // N-Triples is 7-bit ascii
029 }
030
031 public NTripleSink(Writer writer, String... comments) {
032 this.out = new PrintWriter(writer);
033 this.comments = comments;
034 }
035
036 public void start() {
037 for (String line: comments) {
038 out.print("# ");
039 out.println(line);
040 }
041 }
042
043 public void end() {
044 out.flush();
045 }
046
047 public void addObject(String subject, String predicate, String object) {
048 out.print(toNode(subject));
049 out.print(toNode(predicate));
050 out.print(toNode(object));
051 out.println(".");
052 }
053
054 public void addLiteral(String subject, String predicate, String lex, String lang, String datatype) {
055 out.print(toNode(subject));
056 out.print(toNode(predicate));
057 out.print(toLiteral(lex, lang, datatype));
058 out.println(".");
059 }
060
061 public void addPrefix(String prefix, String uri) {}
062
063 protected final String toNode(String node) {
064 if (node.startsWith("_:") || node.startsWith("?"))
065 return node + " ";
066 return "<" + node + "> ";
067 }
068
069 protected final String toLiteral(String lex, String lang, String datatype) {
070 if (lang != null)
071 return quote(lex) + "@" + lang + " ";
072 if (datatype != null)
073 return quote(lex) + "^^<" + datatype + "> ";
074 return quote(lex) + " ";
075 }
076
077 private Pattern quotePattern = Pattern.compile("\"");
078 protected final String quote(String lex) {
079 return "\"" + encode(lex) + "\"";
080 }
081
082 protected final String encode(String s) {
083 StringBuilder b = new StringBuilder();
084 for (int i = 0; i < s.length(); i++) {
085 int c = s.codePointAt(i);
086 if (c <= 8) b.append(enc(c));
087 else if (c == '\t') b.append("\\t");
088 else if (c == '\n') b.append("\\n");
089 else if (c == '\r') b.append("\\r");
090 else if (c == '"') b.append("\\\"");
091 else if (c == '\\') b.append("\\\\");
092 else if (c <= 127) b.appendCodePoint(c);
093 else if (c <= 0xFFFF) b.append(enc(c));
094 else b.append(longenc(c));
095 }
096 return b.toString();
097 }
098
099 protected final String enc(int codepoint) {
100 return String.format("\\u%04x", codepoint);
101 }
102
103 protected final String longenc(int codepoint) {
104 return String.format("\\u%08x", codepoint);
105 }
106 }
107
108 /*
109 * (c) Copyright 2009 University of Bristol
110 * All rights reserved.
111 *
112 * Redistribution and use in source and binary forms, with or without
113 * modification, are permitted provided that the following conditions
114 * are met:
115 * 1. Redistributions of source code must retain the above copyright
116 * notice, this list of conditions and the following disclaimer.
117 * 2. Redistributions in binary form must reproduce the above copyright
118 * notice, this list of conditions and the following disclaimer in the
119 * documentation and/or other materials provided with the distribution.
120 * 3. The name of the author may not be used to endorse or promote products
121 * derived from this software without specific prior written permission.
122 *
123 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
124 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
125 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
126 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
127 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
128 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
129 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
130 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
131 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
132 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
133 */