int main(int argc, char * argv[]) {
if (argc < 2) {
std::cerr << "cws [model path] [lexicon_file]" << std::endl;
return 1;
}
void * engine = 0;
if (argc == 2) {
engine = segmentor_create_segmentor(argv[1]);
} else if (argc == 3) {
engine = segmentor_create_segmentor(argv[1], argv[2]);
}
if (!engine) {
return -1;
}
std::vector<std::string> words;
const char * suite[2] = {
"What's wrong with you? 别灰心! http://t.cn/zQz0Rn",
"台北真的是天子骄子吗？",};
for (int i = 0; i < 2; ++ i) {
words.clear();
int len = segmentor_segment(engine, suite[i], words);
for (int i = 0; i < len; ++ i) {
std::cout << words[i];
if (i+1 == len) std::cout <<std::endl;
else std::cout<< "|";
}
}
segmentor_release_segmentor(engine);
return 0;
}
std::vector<std::string> segment(const std::string& sentence) {
std::vector<std::string> ret;
if (model == NULL) {
std::cerr << "Segmentor: Model not loaded!" << std::endl;
} else {
segmentor_segment(model, sentence.c_str(), ret);
}
return ret;
}
// integrate word segmentor into LTP
int LTP::wordseg(XML4NLP & xml) {
if (xml.QueryNote(NOTE_WORD)) {
return 0;
}
//
int ret = splitSentence_dummy(xml);
if (0 != ret) {
ERROR_LOG("in LTP::wordseg, failed to perform split sentence preprocess.");
return ret;
}
// get the segmentor pointer
void * segmentor = _resource.GetSegmentor();
if (0 == segmentor) {
ERROR_LOG("in LTP::wordseg, failed to init a segmentor");
return kWordsegError;
}
int stnsNum = xml.CountSentenceInDocument();
if (0 == stnsNum) {
ERROR_LOG("in LTP::wordseg, number of sentence equals 0");
return kEmptyStringError;
}
for (int i = 0; i < stnsNum; ++ i) {
std::string strStn = xml.GetSentence(i);
std::vector<std::string> vctWords;
if (ltp::strutils::codecs::length(strStn) > MAX_SENTENCE_LEN) {
ERROR_LOG("in LTP::wordseg, input sentence is too long");
return kSentenceTooLongError;
}
if (0 == segmentor_segment(segmentor, strStn, vctWords)) {
ERROR_LOG("in LTP::wordseg, failed to perform word segment on \"%s\"",
strStn.c_str());
return kWordsegError;
}
if (0 != xml.SetWordsToSentence(vctWords, i)) {
ERROR_LOG("in LTP::wordseg, failed to write segment result to xml");
return kWriteXmlError;
}
}
xml.SetNote(NOTE_WORD);
return 0;
}
JNIEXPORT jint JNICALL Java_edu_hit_ir_ltp4j_Segmentor_segment
(JNIEnv* env, jobject obj, jstring sent, jobject array_words) {
jclass array_list = env->GetObjectClass(array_words);
jmethodID list_add = env->GetMethodID(array_list, "add", "(Ljava/lang/Object;)Z");
const char* str_sent = env->GetStringUTFChars( sent , 0);
std::string sentence(str_sent);
std::vector<std::string> words;
int len = segmentor_segment(segmentor, sentence, words);
for(int i = 0; i < len; i++) {
jobject tmp = stringToJstring(env,words[i].c_str());
env->CallBooleanMethod(array_words,list_add,tmp);
}
env->ReleaseStringUTFChars(sent, str_sent);
return len;
}
int main(int argc, char * argv[]) {
if (argc < 2) {
std::cerr << "cws [model path] [lexicon_file]" << std::endl;
return 1;
}
void * engine = 0;
if (argc == 2) {
engine = segmentor_create_segmentor(argv[1]);
} else if (argc == 3) {
engine = segmentor_create_segmentor(argv[1], argv[2]);
}
if (!engine) {
return -1;
}
std::vector<std::string> words;
std::string sentence;
std::cerr << "TRACE: Model is loaded" << std::endl;
double tm = get_time();
while (std::getline(std::cin, sentence, '\n')) {
words.clear();
if (sentence.size() == 0) { continue; }
int len = segmentor_segment(engine, sentence, words);
for (int i = 0; i < len; ++ i) {
std::cout << words[i];
if (i+1 == len) std::cout <<std::endl;
else std::cout<< "\t";
}
}
segmentor_release_segmentor(engine);
tm = get_time() - tm;
std::cerr << "TRACE: cws-tm-consume "
<< tm
<< " seconds."
<< std::endl;
return 0;
}