PyObject * PyMmseg_BuildDict(PyObject * self, PyObject* args)
{
csfHelper_MMSegObject *self2 = (csfHelper_MMSegObject *)self;
char *type;
char *source_file;
char *target_file;
if (!PyArg_Parse(args, "(sss)", &type, &source_file, &target_file))
return NULL;
else
{
int ret = 0;
if(strncmp("unigram", type, 7) == 0) {
UnigramCorpusReader ur;
ur.open(source_file, NULL);
int ret = 0;
{
UnigramDict ud;
ret = ud.import(ur);
ud.save(target_file);
}
return Py_BuildValue("i",ret);
}
if(strncmp("thesaurus", type, 9) == 0 ) {
ThesaurusDict tdict;
ret = tdict.import(source_file, target_file);
return Py_BuildValue("i",ret);
}
return Py_None;
}
}
int main(int argc, char **argv) {
int c;
const char* corpus_file = NULL;
const char* uni_corpus_file = NULL;
const char* thesaurus_file = NULL;
const char* out_file = NULL;
const char* dict_path = NULL;
const char* target_file = NULL;
char out_buf[512];
if(argc < 2){
usage(argv[0]);
exit(0);
}
u1 bPlainText = 0;
u1 bUcs2 = 0;
while ((c = getopt(argc, argv, "t:b:u:d:o:rU")) != -1) {
switch (c) {
case 'o':
target_file = optarg;
break;
case 'u':
uni_corpus_file = optarg;
break;
case 'b':
corpus_file = optarg;
break;
case 'd':
dict_path = optarg;
break;
case 't':
thesaurus_file = optarg;
break;
case 'r':
bPlainText = 1;
break;
case 'U':
bUcs2 = 1;
break;
case 'h':
usage(argv[0]);
exit(0);
default:
fprintf(stderr, "Illegal argument \"%c\"\n", c);
return 1;
}
}
if(optind < argc) {
out_file = argv[optind];
}
if(thesaurus_file) {
ThesaurusDict tdict;
tdict.import(thesaurus_file, target_file);
//ThesaurusDict ldict;
//ldict.load("thesaurus.lib");
return 0;
}
if(corpus_file){
//build Synonyms dictionary
SynonymsDict dict;
dict.import(corpus_file);
if(target_file)
dict.save(target_file);
else
dict.save("synonyms.dat");
//debug use
//dict.load("synonyms.dat");
//printf("%s\n", dict.exactMatch("c#"));
return 0;
}
if(!corpus_file && !dict_path) {
//build unigram
if(!out_file) {
//build the output filename
size_t len = strlen(uni_corpus_file);
memcpy(out_buf,uni_corpus_file,len);
memcpy(&out_buf[len],".uni\0",5);
out_file = out_buf;
}
if(target_file) {
out_file = target_file;
}
UnigramCorpusReader ur;
ur.open(uni_corpus_file,bPlainText?"plain":NULL);
if(!bUcs2){
UnigramDict ud;
int ret = ud.import(ur);
ud.save(out_file);
//check
int i = 0;
for(i=0;i<ur.count();i++)
{
UnigramRecord* rec = ur.getAt(i);
if(ud.exactMatch(rec->key.c_str()) == rec->count){
continue;
}else{
printf("error!!!");
}
}//end for
}else{
printf("UCS2 used as inner encoding, is unsupported\n");
}
return 0;
}else
if(!dict_path){ //not segment mode.
//build bigram
if(!out_file) {
//build the output filename
size_t len = strlen(corpus_file);
memcpy(out_buf,corpus_file,len);
memcpy(&out_buf[len],".bi\0",4);
out_file = out_buf;
}
printf("Bigram build unsupported.\n");
}//end if(!corpus_file)
//Segment mode
{
SegmenterManager* mgr = new SegmenterManager();
int nRet = 0;
if(dict_path)
nRet = mgr->init(dict_path);
else{
usage(argv[0]);
exit(0);
}
if(nRet == 0){
//init ok, do segment.
Segmenter* seg = mgr->getSegmenter();
segment(out_file,seg);
}
delete mgr;
}
return 0;
}
