Subversion Repositories wimsdev

/*    Copyright (C) 1998-2003 XIAO, Gang of Universite de Nice - Sophia Antipolis

 *

 *  This program is free software; you can redistribute it and/or modify

 *  it under the terms of the GNU General Public License as published by

 *  the Free Software Foundation; either version 2 of the License, or

 *  (at your option) any later version.

 *

 *  This program is distributed in the hope that it will be useful,

 *  but WITHOUT ANY WARRANTY; without even the implied warranty of

 *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the

 *  GNU General Public License for more details.

 *

 *  You should have received a copy of the GNU General Public License

 *  along with this program; if not, write to the Free Software

 *  Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.

 */

/* Versatile translation according to a dictionary */

/*************** Customization: change values hereafter ****************/

#include "suffix.h"

#include "../Lib/libwims.h"

#include "translator_.h"

/* limit of dictionary entries */

#define entrylim 32768

/* limit of dictionary length */

#define diclim 1024*1024

/***************** Nothing should need change hereafter *****************/

char inpbuf[MAX_LINELEN+1], outbuf[2*MAX_LINELEN+2];

char *dicbuf;

struct entry entry[entrylim];

int entrycount;

int entry_size=sizeof(entry);

int has_digits=0;

int unknown_type=unk_delete;

int nocase=0,leaveline=0;

char *unknown, unkbuf[1024];

int compare(int i1, const char *s2)

{

    int k;

    if(nocase) k=strncasecmp((char*)entry[i1].original,s2,entry[i1].olen);

    else k=strncmp((char*)entry[i1].original,s2,entry[i1].olen);

    if(k==0 && (isalnum(*(s2+entry[i1].olen)) || (*(s2+entry[i1].olen)&128)!=0)) return -1;

    else return k;

}

/* searches a list. Returns index if found, -1 if nomatch.

 * Uses binary search, list must be sorted. */

int search_list2(struct entry *list, int items, size_t item_size, const char *str)

{

    int i1,i2,j,k,t,t1;

    unsigned char c;

    if(items<=0) return -1;

    j=0; c=str[0];

    k=list[0].original[0]-c; if(k==0) k=compare(0,str);

    if(k==0) goto more;

    if(k>0) return -1;

    j=items-1; k=list[j].original[0]-c; if(k==0) k=compare(j,str);

    if(k==0) return j;

    if(k>0) for(i1=0,i2=j;i2>i1+1;) {

      j=i1+(i2-i1)/2;

      k=list[j].original[0]-c; if(k==0) k=compare(j,str);

      if(k==0) goto more;

      if(k>0) {i2=j; continue;}

      if(k<0) {i1=j; continue;}

    }

    if(k>0) {j--;k=compare(j,str);}

    more:

    if((t=list[j].earlier)<0) {

      if(k==0) return j; else return -1;

    }

    if(compare(t,str)!=0) return -1;

    for(j=t1=t,k=0;j<items && list[j].earlier==t1 && (k=compare(j,str))<=0; j++) {

      if(k==0) t=j;

    }

    return t;

}

/* change all spaces into ' ', and collapse multiple occurences */

void singlespace2(char *p)

{

    char *pp, *p2;

    for(pp=p;*pp;pp++) {

      if(!isspace(*pp)) continue;

      if(leaveline) {

          if(*pp==13) ovlstrcpy(pp,pp+1);

          if(*pp=='\n') {

            pp++;

            gopt: for(p2=pp; isspace(*p2) && *p2!='\n'; p2++);

            if(p2>pp) ovlstrcpy(pp,p2);

            pp--;

          }

          else {

            pp++; if(!isspace(*pp) || *pp=='\n') continue;

            goto gopt;

          }

      }

      else {

          if(*pp!=' ') *pp=' ';

          if(!isspace(*(pp+1))) continue;

          for(pp++,p2=pp;isspace(*p2);p2++);

          ovlstrcpy(pp,p2); pp--;

      }

    }

}

/* Prepare dictionary */

void prepare_dic(char *fname)

{

    int i,l;

    FILE *dicf;

    char *p1, *p2, *pp;

    long int flen;

    entrycount=0;

    dicf=fopen(fname,"r"); if(dicf==NULL) return;

    fseek(dicf,0,SEEK_END);flen=ftell(dicf); fseek(dicf,0,SEEK_SET);

    if(flen>diclim) return;

    dicbuf=xmalloc(flen+16);flen=fread(dicbuf,1,flen,dicf);

    fclose(dicf);

    if(flen>0 && flen<diclim) dicbuf[flen]=0;

    else return;

    for(i=0,p1=dicbuf;p1!=NULL && *p1!=0 && i<entrylim;p1=p2) {

      p2=strchr(p1+1,'\n'); if(p2>p1) *p2++=0;

      pp=strchr(p1,':'); if(pp==NULL) continue;

      *pp++=0;

      strip_trailing_spaces2(p1); strip_trailing_spaces2(pp);

      singlespace2(p1);

      p1=find_word_start(p1); pp=find_word_start(pp);

      if(*p1==0) continue;

      if(has_digits==0) {

          char *p;

          for(p=p1;*p!=0 && p<pp && !isdigit(*p);p++);

          if(isdigit(*p)) has_digits=1;

      }

      entry[i].original=(unsigned char*)p1;

      entry[i].replace=(unsigned char*)pp;

      entry[i].olen=l=strlen(p1); entry[i].earlier=-1;

      if(i>0) {

          int l1,l2;

          l1=entry[i-1].earlier; if(l1>=0) l2=entry[l1].olen;

          else {l2=entry[i-1].olen;l1=i-1;}

          if(l>l2 && isspace(p1[l2])

             && strncmp((char*)entry[l1].original,p1,l2)==0)

            entry[i].earlier=entry[i-1].earlier=l1;

      }

      i++;

    }

    entrycount=i;

}

/* now make the translation. */

void translate(char *p)

{

  char *p1, *p2, *pp;

  int t;

  if(entrycount<=0 && suffixcnt<=0) return;

  snprintf(outbuf,sizeof(outbuf),"%s",p);

  for(p1=find_word_start(outbuf);

        p1!=NULL && p1-outbuf<MAX_LINELEN && *p1!=0;

        p1=p2) {

    p2=find_word_end(p1);

    for(pp=p1;pp<p2 &&

      ((!has_digits && isalpha(*pp)) ||

       (has_digits && isalnum(*pp)) || (*pp&128)!=0 ||

      strchr("_",*pp)!=NULL);pp++);

    p2=find_word_start(p2);

    if(pp==p1 ||

      (has_digits==0 && isdigit(*pp)) ||

      (*pp!=0 && !isspace(*pp) && strchr(",.?!/;",*pp)==NULL)) continue;

    t=search_list2(entry,entrycount,sizeof(entry[0]),p1);

    if(t<0) {

      switch(unknown_type) {

        case unk_leave: break;

        case unk_delete: {

          ovlstrcpy(p1,find_word_start(pp)); p2=p1;

          break;

        }

        case unk_replace: {

          string_modify3(outbuf,p1,pp,unkbuf);

          p2=find_word_start(p1+strlen(unkbuf));

        }

      }

      continue;

    }

    string_modify3(outbuf,p1,p1+strlen((char*)entry[t].original),

                  (char*)entry[t].replace);

    p2=find_word_start(p1+strlen((char*)entry[t].replace));

  }

  snprintf(p,MAX_LINELEN,"%s",outbuf);

}