@Override
public JapaneseTokenizer init(final byte[] txt) {
String source = string(txt);
if (wc) { // convert wide-space to space
source = source.replace('\u3000', '\u0020');
}
final ArrayList<?> morpheme = (ArrayList<?>) Reflect.invoke(parse, tagger, source);
final ArrayList<Morpheme> list = new ArrayList<>();
try {
int prev = 0;
final int ms = morpheme.size();
for (int i = 0; i < ms; i++) {
final Object m = morpheme.get(i);
final String srfc = surface.get(m).toString();
final String ftr = feature.get(m).toString();
final int strt = start.getInt(m);
if (i != 0) {
final int l = strt - prev;
if (l != 0) {
list.add(new Morpheme(source.substring(strt - 1, strt + l - 1), KIGOU_FEATURE));
}
}
prev = srfc.length() + strt;
// separates continuous mark (ASCII)
boolean cont = true;
final ArrayList<Morpheme> marks = new ArrayList<>();
final int sl = srfc.length();
for (int s = 0; s < sl; s++) {
final String c = String.valueOf(srfc.charAt(s));
final byte[] t = token(c);
if (t.length == 1) {
if (letter(t[0]) || digit(t[0])) cont = false;
else marks.add(new Morpheme(c, KIGOU_FEATURE));
} else {
cont = false;
}
}
if (cont) list.addAll(marks);
else list.add(new Morpheme(srfc, ftr));
}
} catch (final Exception ex) {
Util.errln(Util.className(this) + ": " + ex);
}
tokenList = list;
tokens = list.iterator();
return this;
}
/**
* Converts an HTML document to XML.
*
* @param io io reference
* @param opts html options
* @return parser
* @throws IOException I/O exception
*/
private static IO toXML(final IO io, final HtmlOptions opts) throws IOException {
// reader could not be initialized; fall back to XML
if (READER == null) return io;
try {
// tries to extract the encoding from the input
final TextInput ti = new TextInput(io);
String enc = ti.encoding();
final byte[] content = ti.content();
// looks for a charset definition
final byte[] encoding = token("charset=");
int cs = indexOf(content, encoding);
if (cs > 0) {
// extracts the encoding string
cs += encoding.length;
int ce = cs;
final int cl = content.length;
while (++ce < cl && content[ce] > 0x28) ;
enc = string(substring(content, cs, ce));
}
// define input
final InputSource is = new InputSource(new ArrayInput(content));
is.setEncoding(supported(enc) ? normEncoding(enc) : UTF8);
// define output
final StringWriter sw = new StringWriter();
final XMLReader reader = (XMLReader) Reflect.get(READER);
final Object writer = Reflect.get(WRITER, sw);
// set TagSoup options
if (opts.get(HtmlOptions.HTML)) {
reader.setFeature("http://xml.org/sax/features/namespaces", false);
opt("method", "html");
opt("omit-xml-declaration", "yes");
}
if (opts.get(HtmlOptions.NONS))
reader.setFeature("http://xml.org/sax/features/namespaces", false);
if (opts.get(HtmlOptions.OMITXML)) opt("omit-xml-declaration", "yes");
if (opts.get(HtmlOptions.NOBOGONS)) reader.setFeature(FEATURES + "ignore-bogons", true);
if (opts.get(HtmlOptions.NODEFAULTS))
reader.setFeature(FEATURES + "default-attributes", false);
if (opts.get(HtmlOptions.NOCOLONS)) reader.setFeature(FEATURES + "translate-colons", true);
if (opts.get(HtmlOptions.NORESTART)) reader.setFeature(FEATURES + "restart-elements", false);
if (opts.get(HtmlOptions.IGNORABLE))
reader.setFeature(FEATURES + "ignorable-whitespace", true);
if (opts.get(HtmlOptions.EMPTYBOGONS)) reader.setFeature(FEATURES + "bogons-empty", true);
if (opts.get(HtmlOptions.ANY)) reader.setFeature(FEATURES + "bogons-empty", false);
if (opts.get(HtmlOptions.NOROOTBOGONS)) reader.setFeature(FEATURES + "root-bogons", false);
if (opts.get(HtmlOptions.NOCDATA)) reader.setFeature(FEATURES + "cdata-elements", false);
if (opts.get(HtmlOptions.LEXICAL))
reader.setProperty("http://xml.org/sax/properties/lexical-handler", writer);
if (opts.contains(HtmlOptions.METHOD)) opt("method", opts.get(HtmlOptions.METHOD));
if (opts.contains(HtmlOptions.DOCTYPESYS))
opt("doctype-system", opts.get(HtmlOptions.DOCTYPESYS));
if (opts.contains(HtmlOptions.DOCTYPEPUB))
opt("doctype-public", opts.get(HtmlOptions.DOCTYPEPUB));
if (opts.contains(HtmlOptions.ENCODING)) is.setEncoding(opts.get(HtmlOptions.ENCODING));
// end TagSoup options
reader.setContentHandler((ContentHandler) writer);
reader.parse(is);
return new IOContent(token(sw.toString()), io.name());
} catch (final SAXException ex) {
Util.errln(ex);
return io;
}
}
/** Writes the properties to disk. */
public final synchronized void write() {
final StringBuilder user = new StringBuilder();
BufferedReader br = null;
try {
// caches options specified by the user
if (file.exists()) {
br = new BufferedReader(new FileReader(file.file()));
for (String line; (line = br.readLine()) != null; ) {
if (line.equals(PROPUSER)) break;
}
for (String line; (line = br.readLine()) != null; ) {
user.append(line).append(NL);
}
}
} catch (final Exception ex) {
Util.debug(ex);
} finally {
if (br != null)
try {
br.close();
} catch (final IOException e) {
}
}
BufferedWriter bw = null;
try {
bw = new BufferedWriter(new FileWriter(file.file()));
bw.write(PROPHEADER + NL);
for (final Field f : getClass().getFields()) {
final Object obj = f.get(null);
if (!(obj instanceof Object[])) continue;
final String key = ((Object[]) obj)[0].toString();
final Object val = props.get(key);
if (val instanceof String[]) {
final String[] str = (String[]) val;
bw.write(key + " = " + str.length + NL);
final int is = str.length;
for (int i = 0; i < is; ++i) {
if (str[i] != null) bw.write(key + (i + 1) + " = " + str[i] + NL);
}
} else if (val instanceof int[]) {
final int[] num = (int[]) val;
final int ns = num.length;
for (int i = 0; i < ns; ++i) {
bw.write(key + i + " = " + num[i] + NL);
}
} else {
bw.write(key + " = " + val + NL);
}
}
bw.write(NL + PROPUSER + NL);
bw.write(user.toString());
} catch (final Exception ex) {
Util.errln("% could not be written.", file);
Util.debug(ex);
} finally {
if (bw != null)
try {
bw.close();
} catch (final IOException e) {
}
}
}
