Subversion Repositories wimsdev

Rev

Rev 12248 | Blame | Compare with Previous | Last modification | View Log | RSS feed

  1. /*    Copyright (C) 1998-2003 XIAO, Gang of Universite de Nice - Sophia Antipolis
  2.  *
  3.  *  This program is free software; you can redistribute it and/or modify
  4.  *  it under the terms of the GNU General Public License as published by
  5.  *  the Free Software Foundation; either version 2 of the License, or
  6.  *  (at your option) any later version.
  7.  *
  8.  *  This program is distributed in the hope that it will be useful,
  9.  *  but WITHOUT ANY WARRANTY; without even the implied warranty of
  10.  *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  11.  *  GNU General Public License for more details.
  12.  *
  13.  *  You should have received a copy of the GNU General Public License
  14.  *  along with this program; if not, write to the Free Software
  15.  *  Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  16.  */
  17.  
  18. /* Versatile translation according to a dictionary */
  19.  
  20. /*************** Customization: change values hereafter ****************/
  21.  
  22. /* limit of dictionary entries */
  23. #define entrylim 32768
  24. /* limit of dictionary length */
  25. #define diclim 2*1024*1024
  26. /* limit of source length */
  27. #define sourcelim 16*1024*1024
  28.  
  29. /***************** Nothing should need change hereafter *****************/
  30. #include "../Lib/libwims.h"
  31. #include "suffix.h"
  32.  
  33. char *inpbuf, outbuf[2*(MAX_LINELEN+1)];
  34. char *dicbuf;
  35. struct entry {
  36.   unsigned char *original, *replace;
  37.   int olen,earlier;
  38. } entry[entrylim];
  39. int entrycount;
  40.  
  41. enum {
  42.   unk_delete, unk_leave, unk_replace
  43. };
  44.  
  45. int has_digits=0;
  46. int unknown_type=unk_delete;
  47. int nocase=0,leaveline=0,fromfile=0;
  48. char *unknown, unkbuf[1024];
  49.  
  50. /* Exit without translating anything */
  51. void escape(void)
  52. {
  53.   printf("%s",inpbuf); exit(0);
  54. }
  55.  
  56. int compare(int i1, const char *s2)
  57. {
  58.   int k;
  59.   if(nocase) k=strncasecmp((char*)entry[i1].original,s2,entry[i1].olen);
  60.   else k=strncmp((char*)entry[i1].original,s2,entry[i1].olen);
  61.   if(k==0 && (isalnum(*(s2+entry[i1].olen)) || (*(s2+entry[i1].olen)&128)!=0)) return -1;
  62.   else return k;
  63. }
  64.  
  65. /* searches a list. Returns index if found, -1 if nomatch.
  66.  * Uses binary search, list must be sorted.
  67.  */
  68. int search_list2(struct entry *list, int items, size_t item_size, const char *str)
  69. {
  70.   int i1,i2,j,k,t,t1;
  71.   unsigned char c;
  72.  
  73.   if(items<=0) return -1;
  74.   j=0; c=str[0];
  75.   k=list[0].original[0]-c; if(k==0) k=compare(0,str);
  76.   if(k==0) goto more;
  77.   if(k>0) return -1;
  78.   j=items-1; k=list[j].original[0]-c; if(k==0) k=compare(j,str);
  79.   if(k==0) return j;
  80.   if(k>0) for(i1=0,i2=j;i2>i1+1;) {
  81.     j=i1+(i2-i1)/2;
  82.     k=list[j].original[0]-c; if(k==0) k=compare(j,str);
  83.     if(k==0) goto more;
  84.     if(k>0) {i2=j; continue;}
  85.     if(k<0) {i1=j; continue;}
  86.   }
  87.   if(k>0) {j--;k=compare(j,str);}
  88.   more:
  89.   if((t=list[j].earlier)<0) {
  90.     if(k==0) return j; else return -1;
  91.   }
  92.   if(compare(t,str)!=0) return -1;
  93.   for(j=t1=t,k=0;j<items && list[j].earlier==t1 && (k=compare(j,str))<=0; j++) {
  94.     if(k==0) t=j;
  95.   }
  96.   return t;
  97. }
  98.  
  99. /* change all spaces into ' ', and collapse multiple occurences */
  100. void singlespace2(char *p)
  101. {
  102.   char *pp, *p2;
  103.   for(pp=p;*pp;pp++) {
  104.     if(!isspace(*pp)) continue;
  105.     if(leaveline) {
  106.       if(*pp==13) ovlstrcpy(pp,pp+1);
  107.       if(*pp=='\n') {
  108.         pp++;
  109.         gopt: for(p2=pp; isspace(*p2) && *p2!='\n'; p2++);
  110.         if(p2>pp) ovlstrcpy(pp,p2);
  111.         pp--;
  112.       }
  113.       else {
  114.         pp++; if(!isspace(*pp) || *pp=='\n') continue;
  115.         goto gopt;
  116.       }
  117.     }
  118.     else {
  119.       if(*pp!=' ') *pp=' ';
  120.       pp++; if(!isspace(*pp)) continue;
  121.       for(p2=pp;isspace(*p2);p2++);
  122.       ovlstrcpy(pp,p2); pp--;
  123.     }
  124.   }
  125. }
  126.  
  127. /* Prepare dictionary */
  128. static void prepare_dict(void)
  129. {
  130.   int i,l;
  131.   FILE *dicf;
  132.   char *fname, *p1, *p2, *pp, buf[1024];
  133.   long int flen;
  134.   fname=getenv("w_dictionary");
  135.   if(fname==NULL || *fname==0 || *fname=='/' || strstr(fname,"..")) {
  136.     p1=getenv("w_module");
  137.     if(p1 && strncmp(p1,"classes/",strlen("classes/"))==0) {
  138.       p1=getenv("w_wims_class"); p2=getenv("w_wims_home");
  139.       if(p1 && p2) {
  140.         snprintf(buf,sizeof(buf),"%s/log/classes/%s/",p2,p1);
  141.         if(strncmp(fname,buf,strlen(buf))!=0) escape();
  142.       }
  143.       else escape();
  144.     }
  145.     else {
  146.       p1=getenv("untrust"); if(p1 && strstr(p1,"yes")) escape();
  147.     }
  148.   }
  149. /* replace escape() by return if there is some suffix dictionary, */
  150.  
  151.   dicf=fopen(fname,"r"); if(dicf==NULL) return;
  152.   fseek(dicf,0,SEEK_END);flen=ftell(dicf); fseek(dicf,0,SEEK_SET);
  153.   if(flen>diclim) escape();
  154.   dicbuf=xmalloc(flen+16);flen=fread(dicbuf,1,flen,dicf);
  155.   fclose(dicf);
  156.   if(flen>0 && flen<diclim) dicbuf[flen]=0;
  157.   else escape();
  158.   for(i=0,p1=dicbuf;p1!=NULL && *p1!=0 && i<entrylim;p1=p2) {
  159.     p2=strchr(p1+1,'\n'); if(p2>p1) *p2++=0;
  160.     pp=strchr(p1,':'); if(pp==NULL) continue;
  161.     *pp++=0;
  162.     strip_trailing_spaces2(p1); strip_trailing_spaces2(pp);
  163.     singlespace2(p1);
  164.     p1=find_word_start(p1); pp=find_word_start(pp);
  165.     if(*p1==0) continue;
  166.     if(has_digits==0) {
  167.       char *p;
  168.       for(p=p1;*p!=0 && p<pp && !isdigit(*p);p++);
  169.       if(isdigit(*p)) has_digits=1;
  170.     }
  171.     entry[i].original=(unsigned char*)p1;
  172.     entry[i].replace=(unsigned char*)pp;
  173.     entry[i].olen=l=strlen(p1); entry[i].earlier=-1;
  174.     if(i>0) {
  175.       int l1,l2;
  176.       l1=entry[i-1].earlier; if(l1>=0) l2=entry[l1].olen;
  177.       else {l2=entry[i-1].olen;l1=i-1;}
  178.       if(l>l2 && isspace(p1[l2])
  179.            && strncmp((char*)entry[l1].original,p1,l2)==0)
  180.         entry[i].earlier=entry[i-1].earlier=l1;
  181.     }
  182.     i++;
  183.   }
  184.   entrycount=i; if(entrycount<=0) escape();
  185. }
  186.  
  187. /* now make the translation. */
  188. void translate(void)
  189. {
  190.   char *p1, *p2, *pp;
  191.   int t;
  192.  
  193.   for(p1=find_word_start(outbuf);
  194.       p1!=NULL && p1-outbuf<MAX_LINELEN && *p1!=0;
  195.       p1=p2) {
  196.     p2=find_word_end(p1);
  197.     for(pp=p1;pp<p2 &&
  198.         ((!has_digits && isalpha(*pp)) ||
  199.          (has_digits && isalnum(*pp)) || (*pp&128)!=0 ||
  200.       strchr("_",*pp)!=NULL);pp++);
  201.     p2=find_word_start(p2);
  202.     if(pp==p1 ||
  203.        (has_digits==0 && isdigit(*pp)) ||
  204.        (*pp!=0 && !isspace(*pp) && strchr(",.?!/;",*pp)==NULL)) continue;
  205.     t=search_list2(entry,entrycount,sizeof(entry[0]),p1);
  206.     if(t<0) {
  207.       switch(unknown_type) {
  208.         case unk_leave: break;
  209.         case unk_delete: {
  210.           ovlstrcpy(p1,find_word_start(pp)); p2=p1;
  211.           break;
  212.         }
  213.         case unk_replace: {
  214.           string_modify3(outbuf,p1,pp,unkbuf);
  215.           p2=find_word_start(p1+strlen(unkbuf));
  216.       }
  217.     }
  218.     continue;
  219.   }
  220.   string_modify3(outbuf,p1,p1+strlen((char*)entry[t].original),
  221.                 (char*)entry[t].replace);
  222.   p2=find_word_start(p1+strlen((char*)entry[t].replace));
  223.   }
  224.   outbuf[MAX_LINELEN]=0; printf("%s",outbuf);
  225. }
  226.  
  227. void switches(void)
  228. {
  229.   char *sw;
  230.   unknown=getenv("w_translator_unknown");
  231.   if(unknown==NULL) unknown="";
  232.   snprintf(unkbuf,sizeof(unkbuf),"%s",find_word_start(unknown));
  233.   *find_word_end(unkbuf)=0;
  234.   if(strcasecmp(unkbuf,"leave")==0) unknown_type=unk_leave;
  235.   else if(unkbuf[0]!=0) unknown_type=unk_replace;
  236.   sw=getenv("w_translator_switch");
  237.   if(sw==NULL) return;
  238.   if(strstr(sw,"leaveline")) leaveline=1;
  239.   if(strstr(sw,"nocase")) nocase=1;
  240.   if(strstr(sw,"file")) fromfile=1;
  241. }
  242.  
  243. int main()
  244. {
  245.   char c, *p1, *p2, *s;
  246.   unsigned int l;
  247.  
  248.   switches();
  249.   if(!fromfile) {
  250.     s=getenv("wims_exec_parm");
  251.     if(s==NULL || *s==0) return 0; /* Nothing to translate */
  252.     l=strlen(s); if(l<=0 || l>sourcelim) return 0; /* too long */
  253.     inpbuf=xmalloc(l+16); memmove(inpbuf,s,l+1);
  254.   }
  255.   else {
  256.     FILE *f;
  257.     s=getenv("translator_input"); if(s==NULL || *s==0) return 0;
  258.     f=fopen(s,"r"); if(f==NULL) return 0; /* no file */
  259.     fseek(f,0,SEEK_END); l=ftell(f); fseek(f,0,SEEK_SET);
  260.     if(l<=0 || l>sourcelim) return 0; /* too long */
  261.     inpbuf=xmalloc(l+16); (void)fread(inpbuf,1,l,f); fclose(f); inpbuf[l]=0;
  262.   }
  263.   p1=inpbuf; prepare_dict();
  264.   if(leaveline) c='\n'; else c=' ';
  265.   do {
  266.     l=strlen(p1);
  267.     if(l>MAX_LINELEN-1024) l=MAX_LINELEN-1024;
  268.     p2=p1+l;
  269.     if(*p2) {
  270.       while(p2>p1 && *p2!=c) p2--;
  271.     }
  272.     if(p2<=p1) return 0;
  273.     memmove(outbuf,p1,p2-p1); outbuf[p2-p1]=0;
  274.     singlespace2(outbuf);
  275.     s=getenv("w_suffix_dictionary");
  276.     if(s!=NULL && *s!=0) suffix(outbuf,s);
  277.     translate();
  278.     if(*p2==c) {printf("%c",c); p1=++p2;}
  279.   }
  280.   while(*p2);
  281.   return 0;
  282. }
  283.