static PyObject *constituents(PyObject *self, PyObject *args) {
Dictionary dict;
Parse_Options opts;
Sentence sent;
Linkage linkage;
CNode * cn;
/// Link counts
int num_linkages;
const char *text;
PyObject *output_list;
if (!PyArg_ParseTuple(args, "s", &text))
return NULL;
opts = parse_options_create();
parse_options_set_verbosity(opts, -1);
setlocale(LC_ALL, "");
dict = dictionary_create_default_lang();
if (!dict) {
PyErr_SetString(PyExc_RuntimeError, "Fatal error: Unable to open the dictionary");
Py_INCREF(Py_None);
return Py_None;
}
sent = sentence_create(text, dict);
sentence_split(sent, opts);
num_linkages = sentence_parse(sent, opts);
if (num_linkages > 0) {
linkage = linkage_create(0, sent, opts);
cn = linkage_constituent_tree(linkage);
output_list = build_tree(cn, linkage);
if(output_list == Py_None) {
Py_INCREF(output_list);
return output_list;
}
linkage_free_constituent_tree(cn);
linkage_delete(linkage);
} else {
sentence_delete(sent);
dictionary_delete(dict);
parse_options_delete(opts);
Py_INCREF(Py_None);
return Py_None;
}
sentence_delete(sent);
dictionary_delete(dict);
parse_options_delete(opts);
return Py_BuildValue("S", output_list);
}
static void finish(per_thread_data *ptd)
{
if (ptd->sent)
sentence_delete(ptd->sent);
ptd->sent = NULL;
#if DO_PHRASE_TREE
if (tree)
linkage_free_constituent_tree(tree);
tree = NULL;
#endif
if (ptd->linkage)
linkage_delete(ptd->linkage);
ptd->linkage = NULL;
dictionary_delete(ptd->dict);
ptd->dict = NULL;
parse_options_delete(ptd->opts);
ptd->opts = NULL;
parse_options_delete(ptd->panic_parse_opts);
ptd->panic_parse_opts = NULL;
#ifdef USE_PTHREADS
pthread_setspecific(java_key, NULL);
#else
global_ptd = NULL;
#endif
free(ptd);
}
int main() {
Dictionary dict;
Parse_Options opts;
Sentence sent;
Linkage linkage;
char * diagram;
int i, num_linkages;
char * input_string[] = {
"Grammar is useless because there is nothing to say -- Gertrude Stein.",
"Computers are useless; they can only give you answers -- Pablo Picasso.",
};
opts = parse_options_create();
dict = dictionary_create("4.0.dict", "4.0.knowledge",
"4.0.constituent-knowledge", "4.0.affix");
for (i=0; i<2; ++i) {
sent = sentence_create(input_string[i], dict);
num_linkages = sentence_parse(sent, opts);
if (num_linkages > 0) {
linkage = linkage_create(0, sent, opts);
printf("%s\n", diagram = linkage_print_diagram(linkage));
string_delete(diagram);
linkage_delete(linkage);
}
sentence_delete(sent);
}
dictionary_delete(dict);
parse_options_delete(opts);
return 0;
}
int main(int argc, char * argv[])
{
FILE *input_fh = stdin;
Dictionary dict;
Sentence sent;
const char *language="en"; /* default to english, and not locale */
int pp_on=TRUE;
int af_on=TRUE;
int cons_on=TRUE;
int num_linkages, i;
char *input_string;
Label label = NO_LABEL;
const char *codeset;
/*############################################################################################*/
int num=0;
FILE *fp2,*fp_word_info,*fp_num,*fp_lname,*fp_rel,*fp_lcount,*fp_word_cat,*fp_word;
/*##############################################################################################*/
i = 1;
/*############################################################################################*/
if (argc > 4){
num =atoi(argv[4]);
num=num-1;
/// language = argv[1];
/*####################################################################################*/
//
// if ((argc > 1) && (argv[1][0] != '-')) {
/* the dictionary is the first argument if it doesn't begin with "-" */
// language = argv[1];
i++;
}
/*################################*/
language = argv[1];
/*################################*/
/* Get the locale from the environment...
* perhaps we should someday get it from the dictionary ??
*/
setlocale(LC_ALL, "");
/* Check to make sure the current locale is UTF8; if its not,
* then force-set this to the english utf8 locale
*/
codeset = nl_langinfo(CODESET);
if (!strstr(codeset, "UTF") && !strstr(codeset, "utf"))
{
setlocale(LC_CTYPE, "en_US.UTF-8");
}
/*########################################*/
for (; i<argc-3; i++) {
/*########################################*/
if (argv[i][0] == '-') {
if (strcmp("--version", argv[i])==0) {
printf("Version: %s\n", linkgrammar_get_version());
exit(0);
} else if (strcmp("-ppoff", argv[i])==0) {
pp_on = FALSE;
} else if (strcmp("-coff", argv[i])==0) {
cons_on = FALSE;
} else if (strcmp("-aoff", argv[i])==0) {
af_on = FALSE;
} else if (strcmp("-batch", argv[i])==0) {
} else if (strncmp("-!", argv[i],2)==0) {
} else {
print_usage(argv[0]);
}
} else {
print_usage(argv[0]);
}
}
opts = parse_options_create();
if (opts == NULL) {
fprintf(stderr, "%s: Fatal error: unable to create parse options\n", argv[0]);
exit(-1);
}
panic_parse_opts = parse_options_create();
if (panic_parse_opts == NULL) {
fprintf(stderr, "%s: Fatal error: unable to create panic parse options\n", argv[0]);
exit(-1);
}
setup_panic_parse_options(panic_parse_opts);
parse_options_set_max_sentence_length(opts, 170);
parse_options_set_panic_mode(opts, TRUE);
parse_options_set_max_parse_time(opts, 30);
parse_options_set_linkage_limit(opts, 1000);
parse_options_set_short_length(opts, 10);
/*##########################################*/
// parse_options_set_display_on(opts, TRUE);
/*##########################################*/
if(language && *language)
dict = dictionary_create_lang(language);
else
dict = dictionary_create_default_lang();
if (dict == NULL) {
/*###########################################################################*/
fprintf(stderr, "%s: Fatal error: Unable to open dictionary.\n", argv[1]);
/*##########################################################################*/
exit(-1);
}
/* process the command line like commands */
for (i=1; i<argc; i++) {
if ((strcmp("-pp", argv[i])==0) ||
(strcmp("-c", argv[i])==0) ||
(strcmp("-a", argv[i])==0))
{
i++;
}
else if ((argv[i][0] == '-') && (strcmp("-ppoff", argv[i])!=0) &&
(argv[i][0] == '-') && (strcmp("-coff", argv[i])!=0) &&
(argv[i][0] == '-') && (strcmp("-aoff", argv[i])!=0))
{
if (argv[i][1] == '!')
issue_special_command(argv[i]+2, opts, dict);
else
issue_special_command(argv[i]+1, opts, dict);
}
}
verbosity = parse_options_get_verbosity(opts);
/* Main input loop */
while (1)
{
input_string = fget_input_string(input_fh, stdout, opts);
if (NULL == input_string)
{
if (input_fh == stdin) break;
fclose (input_fh);
input_fh = stdin;
continue;
}
if ((strcmp(input_string, "quit\n")==0) ||
(strcmp(input_string, "exit\n")==0)) break;
/* We have to handle the !file command inline; its too hairy
* otherwise ... */
if (strncmp(input_string, "!file", 5) == 0)
{
char * filename = &input_string[6];
input_fh = fopen(filename, "r");
if (NULL == input_fh)
{
int perr = errno;
fprintf(stderr, "Error: %s (%d) %s\n",
filename, perr, strerror(perr));
input_fh = stdin;
continue;
}
continue;
}
if (special_command(input_string, dict)) continue;
if (parse_options_get_echo_on(opts)) {
printf("%s", input_string);
}
if (parse_options_get_batch_mode(opts)) {
label = strip_off_label(input_string);
}
sent = sentence_create(input_string, dict);
if (sent == NULL) continue;
if (sentence_length(sent) > parse_options_get_max_sentence_length(opts)) {
if (verbosity > 0) {
fprintf(stdout,
"Sentence length (%d words) exceeds maximum allowable (%d words)\n",
sentence_length(sent), parse_options_get_max_sentence_length(opts));
}
sentence_delete(sent);
continue;
}
/* First parse with cost 0 or 1 and no null links */
parse_options_set_disjunct_cost(opts, 2);
parse_options_set_min_null_count(opts, 0);
parse_options_set_max_null_count(opts, 0);
parse_options_reset_resources(opts);
num_linkages = sentence_parse(sent, opts);
//############################################################################################
if(num+1>num_linkages && num_linkages != 0)
{
sprintf(link_info_filename,"%s/%s_tmp/linkid_cat.txt",argv[2],argv[3]);
fp2 = fopen(link_info_filename,"a");
if(fp2==NULL) {printf("Could not open %s for writing\n",link_info_filename);exit(1);}
fprintf(fp2,";~~~~~~~~~~\n");
fclose(fp2);
sprintf(link_info_filename,"%s/%s_tmp/linkid_word.txt",argv[2],argv[3]);
fp_word= fopen(link_info_filename, "a");
if(fp_word==NULL) {printf("Could not open %s for writing\n",link_info_filename);exit(1);}
fprintf(fp_word,";~~~~~~~~~~\n");
fclose(fp_word);
sprintf(link_info_filename,"%s/%s_tmp/link_numeric_word.txt",argv[2],argv[3]);
fp_word_info= fopen(link_info_filename, "a");
if(fp_word_info==NULL) {printf("Could not open %s for writing\n",link_info_filename);exit(1);}
fprintf(fp_word_info,";~~~~~~~~~~\n");
fclose(fp_word_info);
sprintf(link_info_filename,"%s/%s_tmp/link_name_expand.txt",argv[2],argv[3]);
fp_lname =fopen(link_info_filename,"a");
if(fp_lname==NULL) {printf("Could not open %s for writing\n",link_info_filename);exit(1);}
fprintf(fp_lname,";~~~~~~~~~~\n");
fclose(fp_lname);
sprintf(link_info_filename,"%s/%s_tmp/link_relation_info.txt",argv[2],argv[3]);
fp_rel = fopen(link_info_filename,"a");
if(fp_rel==NULL) {printf("Could not open %s for writing\n",link_info_filename);exit(1);}
fprintf(fp_rel,";~~~~~~~~~~\n");
fclose(fp_rel);
/* sprintf(link_info_filename,"%s/%s_tmp/constituents.txt",argv[2],argv[3]);
fp = fopen(link_info_filename,"a");
if(fp==NULL) {printf("Could not open %s for writing\n",link_info_filename);exit(1);}
fprintf(fp,"\n");
fprintf(fp,";~~~~~~~~~~\n");
fclose(fp);*/
}
sprintf(link_info_filename,"%s/%s_tmp/linkage_count.txt",argv[2],argv[3]);
fp_lcount = fopen(link_info_filename,"a");
if(fp_lcount==NULL) {printf("Could not open %s for writing\n",link_info_filename);exit(1);}
fprintf(fp_lcount, "(Found maximum of %d linkages )\n",num_linkages);
fprintf(fp_lcount,";~~~~~~~~~~\n");
fclose(fp_lcount);
//############################################################################################
if (num_linkages < 0) continue;
/* Now parse with null links */
if ((num_linkages == 0) && (!parse_options_get_batch_mode(opts))) {
if (verbosity > 0) //fprintf(stdout, "No complete linkages found.\n");
//############################################################################################
/* This part of the code written by Maha Laxmi and Shirisha Manju
* if no complete linkage is found then redirect the output to standard output in clips format*/
sprintf(link_info_filename,"%s/%s_tmp/number.txt",argv[2],argv[3]);
fp_num =fopen(link_info_filename,"a");
if(fp_num==NULL) {printf("Could not open %s for writing\n",link_info_filename);exit(1);}
/* sprintf(link_info_filename,"%s/%s_tmp/constituents.txt",argv[2],argv[3]);
fp_cons= fopen(link_info_filename,"a");
if(fp_cons==NULL) {printf("Could not open %s for writing\n",link_info_filename);exit(1);}*/
sprintf(link_info_filename,"%s/%s_tmp/linkid_word.txt",argv[2],argv[3]);
fp_word= fopen(link_info_filename, "a");
if(fp_word==NULL) {printf("Could not open %s for writing\n",link_info_filename);exit(1);}
sprintf(link_info_filename,"%s/%s_tmp/link_numeric_word.txt",argv[2],argv[3]);
fp_word_info= fopen(link_info_filename, "a");
if(fp_word_info==NULL) {printf("Could not open %s for writing\n",link_info_filename);exit(1);}
sprintf(link_info_filename,"%s/%s_tmp/linkid_cat.txt",argv[2],argv[3]);
fp_word_cat= fopen(link_info_filename, "a");
if(fp_word_cat==NULL) {printf("Could not open %s for writing\n",link_info_filename);exit(1);}
sprintf(link_info_filename,"%s/%s_tmp/link_name_expand.txt",argv[2],argv[3]);
fp_lname =fopen(link_info_filename,"a");
if(fp_lname==NULL) {printf("Could not open %s for writing\n",link_info_filename);exit(1);}
sprintf(link_info_filename,"%s/%s_tmp/link_relation_info.txt",argv[2],argv[3]);
fp_rel = fopen(link_info_filename,"a");
if(fp_rel==NULL) {printf("Could not open %s for writing\n",link_info_filename);exit(1);}
fprintf(fp_word, "\n(No complete linkages found)\n");
fprintf(fp_word_cat, "\n(No complete linkages found)\n");
fprintf(fp_word_info, "\n(No complete linkages found)\n");
fprintf(fp_lname, "\n(No complete linkages found)\n");
fprintf(fp_rel,"\n(No complete linkages found)\n");
fprintf(fp_word_cat,";~~~~~~~~~~\n"); // fprintf(fp_cons,";~~~~~~~~~~\n");
fprintf(fp_lname,";~~~~~~~~~~\n"); fprintf(fp_rel,";~~~~~~~~~~\n");
fprintf(fp_word_info,";~~~~~~~~~~\n"); fprintf(fp_word,";~~~~~~~~~~\n");
/* fclose(fp_cons); */ fclose(fp_word_cat); fclose(fp_word_info);
fclose(fp_lname); fclose(fp_rel); fclose(fp_word);
continue;
/*###############################################################################################################*/
if (parse_options_get_allow_null(opts)) {
parse_options_set_min_null_count(opts, 1);
parse_options_set_max_null_count(opts, sentence_length(sent));
num_linkages = sentence_parse(sent, opts);
}
}
if (parse_options_timer_expired(opts)) {
if (verbosity > 0) fprintf(stdout, "Timer is expired!\n");
}
if (parse_options_memory_exhausted(opts)) {
if (verbosity > 0) fprintf(stdout, "Memory is exhausted!\n");
}
if ((num_linkages == 0) &&
parse_options_resources_exhausted(opts) &&
parse_options_get_panic_mode(opts)) {
/* print_total_time(opts); */
if (verbosity > 0) fprintf(stdout, "Entering \"panic\" mode...\n");
parse_options_reset_resources(panic_parse_opts);
parse_options_set_verbosity(panic_parse_opts, verbosity);
num_linkages = sentence_parse(sent, panic_parse_opts);
if (parse_options_timer_expired(panic_parse_opts)) {
if (verbosity > 0) fprintf(stdout, "Timer is expired!\n");
}
}
/* print_total_time(opts); */
if (parse_options_get_batch_mode(opts)) {
//#######################################################################
batch_process_some_linkages(label, sent, opts,argv[2],argv[3]);
//#########################################################################
// batch_process_some_linkages(label, sent, opts);
}
/*###############################################################################*/
// send num as an argument to process_some_linkages
else {
int c = process_some_linkages(sent, opts,num,argv[2],argv[3]);
if (c == EOF) break;
}
/*################################################################################*/
/* else {
int c = process_some_linkages(sent, opts);
if (c == EOF) break;
}*/
sentence_delete(sent);
}
if (parse_options_get_batch_mode(opts)) {
/* print_time(opts, "Total"); */
/* fprintf(stderr,
"%d error%s.\n", batch_errors, (batch_errors==1) ? "" : "s");*/
}
parse_options_delete(panic_parse_opts);
parse_options_delete(opts);
dictionary_delete(dict);
// printf ("Bye.\n");
return 0;
}
/*******************************************************************************************
* This functions translats a list of sentences in the input_string array,
* terminates when it finds "end" in the array
* Inputs: input_string: Array of sentences
* all_linkages: if true, translate all possible linkages of the sentence
* out_to_file : if true, outputs the results to seperate files in the out directory,
* the out files names would be something like :
src-x-y: Contains a linkage for the source sentence, x is the index
of sentence in the array and y is the index of linkage
trg-x-y-z: Contains a linklage for the target sentence, x is the index
of sentence in the array, y is the index of linkage and z
is the zth translation of that linkage x-y(a linkage in the
source language may have more than one correspondent target
* linkage)
*******************************************************************************************/
void translate(char input_string[][200], int all_linkages, int out_to_file){
Dictionary dict;
Parse_Options opts;
Sentence sent;
Linkage src_linkage, trg_linkage;
Transfer trans;
char * diagram;
FILE * fp;
int i, j, num_src_linkages, num_trg_linkages;
//char filename[30], txfilename[30], filenum[4];
char txfilename[30];
char output_string[200];
int n;
opts = parse_options_create();
parse_options_set_verbosity (opts, FALSE);
parse_options_set_display_walls(opts, TRUE);
parse_options_set_display_postscript(opts, TRUE);
dict = dictionary_create("4.0.dict", "4.0.knowledge", NULL, NULL, "morphology/morphemes.dict");
if (!dict){
fprintf(stderr, "%s\n", lperrmsg);
parse_options_delete(opts);
printf("size : %lld", space_in_use);
exit(0);
}
trans=transfer_create("translation/mapfile.txt", "translation/links_list.txt","translation/lexicon.txt", "target/trg.dict" );
if (!trans){
fprintf(stderr, "%s\n", maperrmsg);
fprintf(stderr, "%s\n", lperrmsg);
printf("size : %lld", space_in_use);
exit(0);
}
for (n=0; strcmp (input_string[n],"end")!=0; n++ ){
sent = sentence_create(input_string[n], dict);
if (!sent){
fprintf(stderr, "%s\n", lperrmsg);
parse_options_delete(opts);
printf("size : %lld", space_in_use);
exit(0);
}
printf ("\n************************************\n%d-%s\n************************************\n", n+1, input_string[n]);
num_src_linkages = sentence_parse(sent, opts);
num_src_linkages = (all_linkages) ? num_src_linkages: (num_src_linkages!=0)*1;
for (i=0; i<num_src_linkages;i++) {
src_linkage = linkage_create(i, sent, opts);
diagram = linkage_print_diagram(src_linkage);
printf ("\nLinkage No. %d-%d\n\n", n+1, i+1);
printf("%s\n", diagram);
if (out_to_file){
sprintf(txfilename,"out/src-%d-%d.txt", n+1, i+1);
fp=fopen(txfilename,"w+");
if (fp==NULL){
fprintf(stderr, "%s%s\n", "Unable to open ", txfilename);
printf("size : %lld", space_in_use);
exit(0);
}
fprintf(fp,"%s",diagram);
fclose(fp);
}
string_delete(diagram);
num_trg_linkages = transfer_linkage_driver(trans, src_linkage);
if(num_trg_linkages==0){
fprintf(stderr, "%s\n", maperrmsg);
}
else{
parse_options_set_display_walls (trans->opts, TRUE);
for (j=0; j<num_trg_linkages; j++){
//second parameter should always be 0 !?
trg_linkage=trans_linkage_create(trans, 0, trans->sent[j], trans->opts);
diagram = linkage_print_diagram(trg_linkage);
printf ("\nTranslation No. %d_%d_%d\n\n", n+1,i+1,j+1);
printf("Translation: %s\n", output_string);
printf("%s\n", diagram);
if (out_to_file){
sprintf(txfilename,"out/trg-%d-%d-%d.txt", n+1, i+1, j+1);
fp=fopen(txfilename,"w+");
if (fp==NULL){
fprintf(stderr, "%s%s\n", "Unable to open ", txfilename);
printf("size : %lld", space_in_use);
exit(0);
}
extract_sent (trg_linkage, output_string);
fprintf(fp, "Translation:%s\n", output_string);
fprintf(fp,"%s",diagram);
fclose(fp);
}
string_delete(diagram);
linkage_delete(trg_linkage);
}
}
linkage_delete(src_linkage);
}
fprintf(stderr, "%s\n", lperrmsg);
sentence_delete(sent);
}
transfer_delete(trans);
dictionary_delete(dict);
parse_options_delete(opts);
}
/*******************************************************************************************
* This functions parses a list of sentences and outputs the results to seperate files in
* the out directory,
* The files are indexed as out_x_y, where x is the index of the sentence in the array and y
* is the index of the linkage
*******************************************************************************************/
void normal_parse(char input_string[][200], int unify_features, int all_linkages, int out_to_file){
Dictionary dict;
Parse_Options opts;
int n;
Sentence sent;
Linkage linkage;
char * diagram;
int i, num_linkages;
char txfilename[30];
FILE * fp;
opts = parse_options_create();
parse_options_set_display_walls(opts, TRUE);
parse_options_set_display_postscript(opts, TRUE);
parse_options_set_unify_features(opts, unify_features);
dict = dictionary_create("4.0.dict", "4.0.knowledge", NULL, NULL, "morphology/morphemes.dict");
if (!dict){
fprintf(stderr, "%s\n", lperrmsg);
parse_options_delete(opts);
printf("size : %lld", space_in_use);
exit(0);
}
//setting opts->unify_features to TRUE enables unification
for (n=0; strcmp (input_string[n],"end")!=0; n++ ){
sent = sentence_create(input_string[n], dict);
if (!sent){
fprintf(stderr, "%s\n", lperrmsg);
dictionary_delete(dict);
parse_options_delete(opts);
exit(0);
}
printf ("\n************************************\n%d-%s\n************************************\n", n+1, input_string[n]);
num_linkages = sentence_parse(sent, opts);
num_linkages = (all_linkages) ? num_linkages : (num_linkages!=0)*1 ;
//Echos all linkages to screen, ps file and text file in ./out directory
for (i=0; i<num_linkages; ++i) {
linkage = linkage_create(i , sent, opts);
diagram = linkage_print_diagram(linkage);
printf ("\n Linkage No. %d-%d\n\n", n+1, i+1);
printf("%s\n", diagram);
if (out_to_file){
sprintf(txfilename,"out/src-%d-%d.txt", n+1, i+1);
fp=fopen(txfilename,"w+");
if (fp==NULL){
fprintf(stderr, "%s%s\n", "Unable to open ", txfilename);
printf("size : %lld", space_in_use);
exit(0);
}
fprintf(fp,"%s",diagram);
fclose(fp);
}
string_delete(diagram);
linkage_delete(linkage);
}
sentence_delete(sent);
}
dictionary_delete(dict);
parse_options_delete(opts);
}
int main(int argc, char * argv[]) {
Dictionary dict;
Sentence sent;
char *dictionary_file=NULL;
char *post_process_knowledge_file=NULL;
char *constituent_knowledge_file=NULL;
char *affix_file=NULL;
int pp_on=TRUE;
int af_on=TRUE;
int cons_on=TRUE;
int num_linkages, i;
char input_string[MAXINPUT];
Label label = NO_LABEL;
int parsing_space_leaked, reported_leak, dictionary_and_option_space;
i = 1;
if ((argc > 1) && (argv[1][0] != '-')) {
/* the dictionary is the first argument if it doesn't begin with "-" */
dictionary_file = argv[1];
i++;
}
for (; i<argc; i++) {
if (argv[i][0] == '-') {
if (strcmp("-pp", argv[i])==0) {
if ((post_process_knowledge_file != NULL) || (i+1 == argc))
print_usage(argv[0]);
post_process_knowledge_file = argv[i+1];
i++;
} else
if (strcmp("-c", argv[i])==0) {
if ((constituent_knowledge_file != NULL) || (i+1 == argc))
print_usage(argv[0]);
constituent_knowledge_file = argv[i+1];
i++;
} else
if (strcmp("-a", argv[i])==0) {
if ((affix_file != NULL) || (i+1 == argc)) print_usage(argv[0]);
affix_file = argv[i+1];
i++;
} else if (strcmp("-ppoff", argv[i])==0) {
pp_on = FALSE;
} else if (strcmp("-coff", argv[i])==0) {
cons_on = FALSE;
} else if (strcmp("-aoff", argv[i])==0) {
af_on = FALSE;
} else if (strcmp("-batch", argv[i])==0) {
} else if (strncmp("-!", argv[i],2)==0) {
} else {
print_usage(argv[0]);
}
} else {
print_usage(argv[0]);
}
}
if (!pp_on && post_process_knowledge_file != NULL) print_usage(argv[0]);
if (dictionary_file == NULL) {
dictionary_file = "4.0.dict";
fprintf(stderr, "No dictionary file specified. Using %s.\n",
dictionary_file);
}
if (af_on && affix_file == NULL) {
affix_file = "4.0.affix";
fprintf(stderr, "No affix file specified. Using %s.\n", affix_file);
}
if (pp_on && post_process_knowledge_file == NULL) {
post_process_knowledge_file = "4.0.knowledge";
fprintf(stderr, "No post process knowledge file specified. Using %s.\n",
post_process_knowledge_file);
}
if (cons_on && constituent_knowledge_file == NULL) {
constituent_knowledge_file = "4.0.constituent-knowledge";
fprintf(stderr, "No constituent knowledge file specified. Using %s.\n",
constituent_knowledge_file);
}
opts = parse_options_create();
if (opts == NULL) {
fprintf(stderr, "%s\n", lperrmsg);
exit(-1);
}
panic_parse_opts = parse_options_create();
if (panic_parse_opts == NULL) {
fprintf(stderr, "%s\n", lperrmsg);
exit(-1);
}
setup_panic_parse_options(panic_parse_opts);
parse_options_set_max_sentence_length(opts, 70);
parse_options_set_panic_mode(opts, TRUE);
parse_options_set_max_parse_time(opts, 30);
parse_options_set_linkage_limit(opts, 1000);
parse_options_set_short_length(opts, 10);
dict = dictionary_create(dictionary_file,
post_process_knowledge_file,
constituent_knowledge_file,
affix_file);
if (dict == NULL) {
fprintf(stderr, "%s\n", lperrmsg);
exit(-1);
}
/* process the command line like commands */
for (i=1; i<argc; i++) {
if ((strcmp("-pp", argv[i])==0) ||
(strcmp("-c", argv[i])==0) ||
(strcmp("-a", argv[i])==0)) {
i++;
} else if ((argv[i][0] == '-') && (strcmp("-ppoff", argv[i])!=0) &&
(argv[i][0] == '-') && (strcmp("-coff", argv[i])!=0) &&
(argv[i][0] == '-') && (strcmp("-aoff", argv[i])!=0)) {
issue_special_command(argv[i]+1, opts, dict);
}
}
dictionary_and_option_space = space_in_use;
reported_leak = external_space_in_use = 0;
verbosity = parse_options_get_verbosity(opts);
while (fget_input_string(input_string, stdin, stdout, opts)) {
if (space_in_use != dictionary_and_option_space + reported_leak) {
fprintf(stderr, "Warning: %d bytes of space leaked.\n",
space_in_use-dictionary_and_option_space-reported_leak);
reported_leak = space_in_use - dictionary_and_option_space;
}
if ((strcmp(input_string, "quit\n")==0) ||
(strcmp(input_string, "exit\n")==0)) break;
if (special_command(input_string, dict)) continue;
if (parse_options_get_echo_on(opts)) {
printf("%s", input_string);
}
if (parse_options_get_batch_mode(opts)) {
label = strip_off_label(input_string);
}
sent = sentence_create(input_string, dict);
if (sent == NULL) {
if (verbosity > 0) fprintf(stderr, "%s\n", lperrmsg);
if (lperrno != NOTINDICT) exit(-1);
else continue;
}
if (sentence_length(sent) > parse_options_get_max_sentence_length(opts)) {
sentence_delete(sent);
if (verbosity > 0) {
fprintf(stdout,
"Sentence length (%d words) exceeds maximum allowable (%d words)\n",
sentence_length(sent), parse_options_get_max_sentence_length(opts));
}
continue;
}
/* First parse with cost 0 or 1 and no null links */
parse_options_set_disjunct_cost(opts, 2);
parse_options_set_min_null_count(opts, 0);
parse_options_set_max_null_count(opts, 0);
parse_options_reset_resources(opts);
num_linkages = sentence_parse(sent, opts);
/* Now parse with null links */
if ((num_linkages == 0) && (!parse_options_get_batch_mode(opts))) {
if (verbosity > 0) fprintf(stdout, "No complete linkages found.\n");
if (parse_options_get_allow_null(opts)) {
parse_options_set_min_null_count(opts, 1);
parse_options_set_max_null_count(opts, sentence_length(sent));
num_linkages = sentence_parse(sent, opts);
}
}
if (parse_options_timer_expired(opts)) {
if (verbosity > 0) fprintf(stdout, "Timer is expired!\n");
}
if (parse_options_memory_exhausted(opts)) {
if (verbosity > 0) fprintf(stdout, "Memory is exhausted!\n");
}
if ((num_linkages == 0) &&
parse_options_resources_exhausted(opts) &&
parse_options_get_panic_mode(opts)) {
print_total_time(opts);
if (verbosity > 0) fprintf(stdout, "Entering \"panic\" mode...\n");
parse_options_reset_resources(panic_parse_opts);
parse_options_set_verbosity(panic_parse_opts, verbosity);
num_linkages = sentence_parse(sent, panic_parse_opts);
if (parse_options_timer_expired(panic_parse_opts)) {
if (verbosity > 0) fprintf(stdout, "Timer is expired!\n");
}
}
print_total_time(opts);
if (parse_options_get_batch_mode(opts)) {
batch_process_some_linkages(label, sent, opts);
}
else {
process_some_linkages(sent, opts);
}
sentence_delete(sent);
if (external_space_in_use != 0) {
fprintf(stderr, "Warning: %d bytes of external space leaked.\n",
external_space_in_use);
}
}
if (parse_options_get_batch_mode(opts)) {
print_time(opts, "Total");
fprintf(stderr,
"%d error%s.\n", batch_errors, (batch_errors==1) ? "" : "s");
}
parsing_space_leaked = space_in_use - dictionary_and_option_space;
if (parsing_space_leaked != 0) {
fprintf(stderr, "Warning: %d bytes of space leaked during parsing.\n",
parsing_space_leaked);
}
parse_options_delete(panic_parse_opts);
parse_options_delete(opts);
dictionary_delete(dict);
if (space_in_use != parsing_space_leaked) {
fprintf(stderr,
"Warning: %d bytes of dictionary and option space leaked.\n",
space_in_use - parsing_space_leaked);
}
else if (parsing_space_leaked == 0) {
fprintf(stderr, "Good news: no space leaked.\n");
}
if (external_space_in_use != 0) {
fprintf(stderr, "Warning: %d bytes of external space leaked.\n",
external_space_in_use);
}
return 0;
}
void command_options_delete(Command_Options* co)
{
parse_options_delete(co->panic_opts);
parse_options_delete(co->popts);
free(co);
}
static PyObject *domains(PyObject *self, PyObject *args) {
Dictionary dict;
Parse_Options opts;
Sentence sent;
Linkage linkage;
//CNode * cn;
/// Link counts
int num_linkages;
int links;
int i;
int j = 0;
int num_domains;
const char *text;
PyObject *output_list;
PyObject *temp;
output_list = PyList_New(0);
if (!PyArg_ParseTuple(args, "s", &text))
return NULL;
opts = parse_options_create();
parse_options_set_verbosity(opts, -1);
setlocale(LC_ALL, "");
dict = dictionary_create_default_lang();
if (!dict) {
PyErr_SetString(PyExc_RuntimeError, "Fatal error: Unable to open the dictionary");
Py_INCREF(Py_None);
return Py_None;
}
sent = sentence_create(text, dict);
sentence_split(sent, opts);
num_linkages = sentence_parse(sent, opts);
if (num_linkages > 0) {
linkage = linkage_create(0, sent, opts);
links = linkage_get_num_sublinkages(linkage);
for(i=0; i<=links; i++) {
num_domains = linkage_get_link_num_domains(linkage, i);
const char **temp1 = linkage_get_link_domain_names(linkage, i);
//for(j=0; j<=num_domains; j++){
while(num_domains < j) {
temp = PyString_FromString(temp1[j]);
PyList_Append(output_list, temp);
j++;
}
j = 0;
}
linkage_delete(linkage);
} else {
sentence_delete(sent);
dictionary_delete(dict);
parse_options_delete(opts);
Py_INCREF(Py_None);
return Py_None;
}
sentence_delete(sent);
dictionary_delete(dict);
parse_options_delete(opts);
return Py_BuildValue("Si", output_list, num_domains);
}
/// This is the basic sentence dissection
static PyObject *sentence(PyObject *self, PyObject *args) {
Dictionary dict;
Parse_Options opts;
Sentence sent;
Linkage linkage;
Linkage sub_linkage;
char * diagram;
/// Link counts
int num_linkages;
int links;
/// Index's for the iterators
int link_idx;
int word_idx;
int num_words;
long span;
long sub_linkages;
const char *text;
const char *d_output;
PyObject *output_list;
PyObject *word_list;
PyObject *word2_list;
PyObject *span_list;
PyObject *temp;
PyObject *sublinkage_list;
PyObject *_diagram;
output_list = PyList_New(0);
word_list = PyList_New(0);
word2_list = PyList_New(0);
sublinkage_list = PyList_New(0);
span_list = PyList_New(0);
if (!PyArg_ParseTuple(args, "s", &text))
return NULL;
opts = parse_options_create();
parse_options_set_verbosity(opts, -1);
parse_options_set_screen_width(opts, 50);
setlocale(LC_ALL, "");
dict = dictionary_create_default_lang();
if (!dict) {
PyErr_SetString(PyExc_RuntimeError, "Fatal error: Unable to open the dictionary");
Py_INCREF(Py_None);
return Py_None;
}
sent = sentence_create(text, dict);
sentence_split(sent, opts);
num_linkages = sentence_parse(sent, opts);
if (num_linkages > 0) {
linkage = linkage_create(0, sent, opts);
/// Get the lengths of everything
num_words = linkage_get_num_words(linkage);
links = linkage_get_num_links(linkage);
for(link_idx=0; link_idx < links; link_idx++) {
PyObject *temp_subLen;
diagram = linkage_print_diagram(linkage);
_diagram = PyString_FromString(diagram);
sub_linkage = linkage_create(link_idx, sent, opts);
sub_linkages = linkage_get_num_sublinkages(linkage);
temp_subLen = PyLong_FromLong(sub_linkages);
linkage_delete(sub_linkage);
PyList_Append(sublinkage_list, temp_subLen);
span = linkage_get_link_length(linkage, link_idx);
PyList_Append(span_list, PyInt_FromLong(span));
PyObject *temp_list;
temp_list = PyList_New(0);
/// Sub Group these (left and right labels)
const char *t1 = linkage_get_link_llabel(linkage, link_idx);
temp = PyString_FromString(t1);
PyList_Append(temp_list, temp);
const char *t2 = linkage_get_link_rlabel(linkage, link_idx);
temp = PyString_FromString(t2);
PyList_Append(temp_list, temp);
/// Then add to the main list
PyList_Append(output_list, temp_list);
/// Just the label
const char *t3 = linkage_get_link_label(linkage, link_idx);
temp = PyString_FromString(t3);
PyList_Append(word2_list, temp);
}
for(word_idx=0; word_idx < num_words; word_idx++) {
d_output = linkage_get_word(linkage, word_idx);
PyObject *word;
word = PyString_FromString(d_output);
PyList_Append(word_list, word);
}
linkage_free_diagram(diagram);
linkage_delete(linkage);
} else {
sentence_delete(sent);
dictionary_delete(dict);
parse_options_delete(opts);
Py_INCREF(Py_None);
return Py_None;
}
sentence_delete(sent);
dictionary_delete(dict);
parse_options_delete(opts);
return Py_BuildValue("SSSSSS", word_list, span_list, output_list, word2_list, sublinkage_list, _diagram);
}
void free_parse_options_ptr(ParseOptionsPtr *ptr) {
parse_options_delete(ptr->opts);
free(ptr);
}
void delete_morpho_structures(Morpho_structures morpho_structures){
dictionary_delete(morpho_structures->dict);
// free_morpho_list(morpho_structures->ecomposed_word);
parse_options_delete(morpho_structures->opts);
xfree(morpho_structures, sizeof(struct Morpho_structures_s));
}
