// ----------------------------------------------
// PLAIN DECODER
// ----------------------------------------------
SegmentDecoder::SegmentDecoder(
Model *model,
int agenda) {
this->m_Model = model;
this->m_Agenda = agenda;
Alphabet *labelAlpha = model->getAlphabet("LABELS");
this->m_NumLabels = labelAlpha->size();
m_Legal = new int *[m_NumLabels + 1];
for (int i = 0; i <= m_NumLabels; ++ i) {
char prev = 'X';
if (i < m_NumLabels)
prev = labelAlpha->rlookup(i)[0];
m_Legal[i] = new int[m_NumLabels];
for (int j = 0; j < m_NumLabels; ++ j) {
char curr = labelAlpha->rlookup(j)[0];
m_Legal[i][j] = 0;
if ((prev == 'X' || prev == 'S' || prev == 'E')
&& (curr == 'S' || curr == 'B'))
m_Legal[i][j] = 1;
if ((prev == 'M' || prev == 'B') &&
(curr == 'M' || curr == 'E'))
m_Legal[i][j] = 1;
}
}
}
int OTWS_Wordseg_x(otws_t handle,
const string& sent,
vector<string>& words) {
OTWS_Engine *engine = reinterpret_cast<OTWS_Engine *>(handle);
words.clear();
RawSentence *tag_sent = new TagSent();
vector<string> chars;
int numChars = UTF::getCharactersFromUTF8String(sent, &chars);
// something for debug.
cerr << "TOKEN: ";
for (int i = 0; i < numChars; ++ i) {
cerr << chars[i] << " | ";
tag_sent->append(new TagItem(chars[i], "X"));
}
cerr << endl;
Instance *inst = (engine)->extractor->extract(tag_sent, false);
Items *items = inst->items();
Labels* labels = engine->decoder->decode(inst,
engine->model->getParameter("PARAMETER"))->best();
Alphabet *labelsDict = engine->model->getAlphabet("LABELS");
// something for debug.
for (int i = 0; i < labels->size(); ++ i) {
cerr << labelsDict->rlookup(labels->at(i)) <<
"(" << labels->at(i) << ") ";
}
cerr << "| label size: " << labels->size() << endl;
string tag;
string word;
for (int i = 0; i < items->size(); ) {
tag = labelsDict->rlookup(labels->at(i));
if ("S" == tag) {
word = chars[i];
words.push_back(word);
++ i;
} else if ("B" == tag) {
word = "";
while ("E" != tag && i < items->size()){
word = word + chars[i];
tag = labelsDict->rlookup(labels->at(i));
++ i;
}
words.push_back(word);
} else {
cerr << "Exception asserted." << endl;
words.clear();
return -1;
}
}
return words.size();
}