source: src/gue/main.cc @ d6a59ca

#include <time.h>
#include <stdlib.h>
#include "../lib/iotools.h"
#define _CMDLINE_FILE "../gue/cmdline.h"
#define CONFIGFILE1 "/home/ynka/utt/utt-0.9/conf/gue.conf"
#define CONFIGFILE2 "/home/ynka/utt/utt-0.9/conf/gue.conf"
#include "../common/common.h"
#include "common_guess.h"
#include "guess.h"
#include "cmdline.h"

#define W_SUFF 0.6
#define W_PREF 0.4


int main(int argc, char** argv) {

  //  int non_standard_config=0;

  gengetopt_args_info args;

  if(cmdline_parser(argc, argv, &args) != 0)
    exit(1);

  process_config_files(&args,argv[0]);
  process_common_options(&args,argv[0]);
  process_guess_options(&args);

  char line[MAX_LINE];
  char outline[MAX_LINE];
  char parms[MAX_LINE], desc[MAX_LINE], lemma[MAX_LINE];
  long line_count = 0;
  
  Guess guess(dictionary);
  int words_count=0;
  time_t start_time = time(NULL);
  
  //  Segment seg;
  Words tab;
  while (fgets(line, MAX_LINE, inputf))
    {
      line_count++;
      int start, len;
      
      //      line[strlen(line)-1] = '\0';
      
      if (!process_seg(line, args))
        fputs(line,outputf);
      else
        {
          char form[MAX_FORM];
          words_count++;
          tab.clear();
          getfield(line,input_field_prefix,form);
          if (form==NULL) continue;//BZDURA
          
          guess.ana(form, tab);
          
          if ((tab.count()==0)) // no guesses - analysis was unsuccessful
            fputs(line, outputf);
          else
            {

//            if (copy_processed)
//              fputs(line, stdout);
//            continue;
//          }
          // we've got some guesses. Do we want to print it?
//            if (args.only_fail_flag)
//          continue;

              float last_weight=0;
              int i=0;
              int count=0;
              unsigned first=1;
              char* parms_end = parms;
              char last_lemma[MAX_LINE];

              count = 1;

                      tab.sort();
              
              while (count < tab.count() && count <= guess_count)
                if (first || tab[count].w_suf() >= cut_off && tab[count].w_suf() >= delta * last_weight)
                  {
                    first=0;
                    last_weight = tab[i].w_suf();
                    count++;
                  }
                else
                  break;
              
              // drukujemy count pierwszych z tab


              if(one_line)
                {
                  char* descp=desc;
                  for (int i=0; i< count; ++i)
                    {
                      descp += sprintf(descp," %s%s,%s", output_field_prefix, tab[i].lemma(), tab[i].descr());
                      if(weights) descp += sprintf(descp,":%d",(int)tab[i].w_suf());
                    }
                  strcpy(outline,line);
                  outline[strlen(outline)-1]='\0';
                  strcat(outline,desc);
                  strcat(outline,"\n");
                  fputs(outline, outputf);
                  if (copy_processed)
                    fputs(line,outputf);
                }
              else if(one_field)
                {
                  char* descp=desc;
                  for (int i=0; i< count; ++i)
                    if(i==0)
                      {
                        descp += sprintf(descp," %s%s,%s", output_field_prefix, tab[i].lemma(), tab[i].descr());
                        if(weights) descp += sprintf(descp,":%d",(int)tab[i].w_suf());
                      }
                    else
                      {
                        if(strcmp(tab[i].lemma(),tab[i-1].lemma())==0)
                          descp += sprintf(descp,",%s",tab[i].descr());
                        else
                          descp += sprintf(descp,";%s,%s",tab[i].lemma(),tab[i].descr());
                        if(weights) descp += sprintf(descp,":%d",(int)tab[i].w_suf());
                      }
                  
                  strcpy(outline,line);
                  outline[strlen(outline)-1]='\0';
                  strcat(outline,desc);
                  strcat(outline,"\n");
                  fputs(outline, outputf);
                  if (copy_processed)
                    fputs(line,outputf);
                }
              else
                {
                  for (int i=0; i< count; ++i)
                    {
                      // kolejne opisy - kolejne linie.
                      char* descp=desc;
                      descp += sprintf(desc, " %s%s,%s", output_field_prefix, tab[i].lemma(), tab[i].descr());
                      if(weights) descp += sprintf(descp,":%d",(int)tab[i].w_suf());
                      descp += sprintf(descp,"\n");
                      strcpy(outline,line);
                      outline[strlen(outline)-1]='\0';
                      strcat(outline,desc);
                      fputs(outline, outputf);
                    }
                  if (copy_processed)
                    fputs(line,outputf);
                }
            }
        }
      if(args.interactive_flag) fflush(outputf);
      
    }
  cmdline_parser_free(&args);
}








//            while ((i=tab.next()) != -1 && count++<guess_count) {
//              /* if we have "one-line" flag then everything goes in one segment as many fields,
//               * if we have "one-field" flag everything goes in one segment as ONE field:
//               * - diferent lemmas are separated with ';', sequent descriptions to one lemma
//               *   are separated with ','
//               */
//              if ((!first) && (tab[i].w_suf() < cut_off) || (tab[i].w_suf() < delta * last_weight)) {
//                break;
//              }
//              if (first) {
//                parms_end += sprintf(parms_end, "%s", output_field_prefix);
//              } else if (!args.one_field_flag)
//                parms_end += sprintf(parms_end, "%s", output_field_prefix);
                
//              if (!args.one_field_flag || strcmp(last_lemma, tab[i].lemma()) != 0) {
//                if (args.one_field_flag && !first)
//                  parms_end += sprintf(parms_end, ";");
//                parms_end += sprintf(parms_end, "%s", tab[i].lemma());
//                strcpy(last_lemma, tab[i].lemma());
//              }
                
//              first=0;
                
//              last_weight = tab[i].w_suf();
//              if (!weights)
//                parms_end += sprintf(parms_end, ",%s:%d",  tab[i].descr(), (int)tab[i].w_suf());
//              else
//                parms_end += sprintf(parms_end, ",%s", tab[i].descr());
                
//              if (!args.one_field_flag) {
//                seg.addfield(parms);
//                parms_end = parms;
//              }
                
//              if (!(args.one_field_flag || args.one_line_flag)) {
//                seg.print(outline);
//                fputs(outline, outputf);
//                --seg.auxn;
//              }
//              //if (copy_processed)
//              //  fputs(outline, stdout);
//            } //while
              
//            if (args.one_field_flag)
//              seg.addfield(parms);
              
//            if (args.one_field_flag || args.one_line_flag){
//              seg.print(outline);
//              fputs(outline, outputf);
//            }
//          } else { // if (process_segment)
//          // jak to nie jest wyraz - to przepisz token na wyjscie.
//          //      printtok(line, start, len, cat, form);
//          seg.print(outline);
//          fputs(outline, outputf);
//          if (copy_processed)
//            fputs(outline, stdout);
//        }
//      }
//       time_t end_time = time(NULL);
//       if (per_info) {
//      printf("Liczba s³ów: %d\n", words_count);
//      printf("Czas analizy: %d sekund\n", end_time-start_time);
//       }
//       cmdline_parser_free(&args);
//     }