public DataBag exec(Tuple input) throws IOException {
try {
if (!input.isNull()) {
// Create the output like a databag {(res1,res2),(res3,res4)..}
DataBag output_databag = mBagFactory.newDefaultBag();
// Unpack tuple in order to get the bag {(1,2),(3,4),...}
String input_time = (String) input.get(0);
try {
DateFormat formatter = new SimpleDateFormat("MM/dd/yyyy kk:mm:ss");
Date date =
formatter.parse(
String.format(
"%s/%s/%s %s:%s:%s",
input_time.substring(5, 7),
input_time.substring(8, 10),
input_time.substring(0, 4),
input_time.substring(11, 13),
input_time.substring(14, 16),
input_time.substring(17, 18)));
Calendar calendar = Calendar.getInstance();
calendar.setTime(date);
int dayOfWeek = calendar.get(Calendar.DAY_OF_WEEK);
int dayOfMonth = calendar.get(Calendar.DAY_OF_MONTH);
int hour = calendar.get(Calendar.HOUR_OF_DAY);
// Add items to output
Tuple items = TupleFactory.getInstance().newTuple(1);
items.set(0, String.format("%d:%d:%d", dayOfWeek, dayOfMonth, hour));
output_databag.add(items);
} catch (Exception e) {
Tuple items = TupleFactory.getInstance().newTuple(1);
items.set(0, "petting #1" + e.getMessage());
output_databag.add(items);
return output_databag;
}
return output_databag;
} else {
DataBag output_databag = mBagFactory.newDefaultBag();
Tuple items = TupleFactory.getInstance().newTuple(1);
items.set(0, "petting #2");
output_databag.add(items);
return output_databag;
}
} catch (Exception e) {
System.err.println("Error with ?? ..");
DataBag output_databag = mBagFactory.newDefaultBag();
Tuple items = TupleFactory.getInstance().newTuple(1);
items.set(0, "petting #3" + e.getMessage());
output_databag.add(items);
return output_databag;
}
}
@Override
public String exec(Tuple input) throws IOException {
String output = "";
if (input != null && !input.isNull()) {
String _html = (String) input.get(0);
if (_html != null && !_html.isEmpty()) {
output = _html;
int start_index = _html.indexOf("<body");
start_index += 5;
int stop_index = _html.indexOf("</body");
if (start_index >= 0
&& start_index < _html.length()
&& stop_index >= start_index
&& stop_index < _html.length()) output = _html.substring(start_index, stop_index);
Pattern REMOVE_TAGS = Pattern.compile("<.+?>");
Matcher m = REMOVE_TAGS.matcher(output);
output = m.replaceAll("");
REMOVE_TAGS = Pattern.compile("<!--(.+?)-->");
m = REMOVE_TAGS.matcher(output);
output = m.replaceAll("");
/* boolean intag = false;
String inp = output;
String outp = "";
for (int i=0; i < inp.length(); ++i)
{
if (!intag && inp.charAt(i) == '<')
{
intag = true;
continue;
}
if (intag && inp.charAt(i) == '>')
{
intag = false;
continue;
}
if (!intag)
{
outp = outp + inp.charAt(i);
}
}
output = outp;
// output = output.replaceAll("\\s+","");
*/
}
}
return output;
}
public DataBag exec(Tuple input) throws IOException {
if (input == null || input.size() < 1 || input.isNull(0)) return null;
// Output bag
DataBag bagOfTokens = bagFactory.newDefaultBag();
StringReader textInput = new StringReader(input.get(0).toString());
PTBTokenizer ptbt = new PTBTokenizer(textInput, new CoreLabelTokenFactory(), "");
for (CoreLabel label; ptbt.hasNext(); ) {
label = (CoreLabel) ptbt.next();
if (label.value().length() > 2) {
System.err.println(label.toString());
Tuple termText = tupleFactory.newTuple(label.word());
bagOfTokens.add(termText);
}
}
return bagOfTokens;
}
@Override
public boolean isNull(int fieldNum) throws ExecException {
return t.isNull(fieldNum);
}
