Rev 10 | Rev 3808 | Go to most recent revision | Details | Compare with Previous | Last modification | View Log | RSS feed
| Rev | Author | Line No. | Line |
|---|---|---|---|
| 10 | reyssat | 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 | char inpbuf[MAX_LINELEN+1], troutbuf[2*MAX_LINELEN+2]; |
||
| 21 | struct entry { |
||
| 22 | unsigned char *original, *replace; |
||
| 23 | int olen,earlier; |
||
| 24 | } entry[MAX_DICENTRIES]; |
||
| 25 | int entrycount=0; |
||
| 26 | |||
| 27 | struct dic { |
||
| 28 | char name[MAX_FNAME+1]; |
||
| 29 | char unknown[256]; |
||
| 30 | char *buf; |
||
| 31 | int unknown_type; |
||
| 32 | int start; |
||
| 33 | int len; |
||
| 34 | } dic[MAX_DICS]; |
||
| 35 | int diccnt; |
||
| 36 | int transdic, macrodic; |
||
| 37 | |||
| 38 | enum { |
||
| 39 | unk_delete, unk_leave, unk_replace |
||
| 40 | }; |
||
| 41 | |||
| 42 | int compare(struct entry *e, const char *s2) |
||
| 43 | { |
||
| 44 | int k; |
||
| 45 | k=strncmp(e->original,s2,e->olen); |
||
| 46 | if(k==0 && isalnum(*(s2+e->olen))) return -1; |
||
| 47 | else return k; |
||
| 48 | } |
||
| 49 | |||
| 50 | /* searches a list. Returns index if found, -1 if nomatch. |
||
| 51 | * Uses binary search, list must be sorted. */ |
||
| 52 | int search_dic(struct entry *list, int items, size_t item_size, const char *str) |
||
| 53 | { |
||
| 54 | int i1,i2,j,k,t,t1; |
||
| 55 | unsigned char c; |
||
| 56 | |||
| 57 | if(items<=0) return -1; |
||
| 58 | j=0; c=str[0]; |
||
| 59 | k=list[0].original[0]-c; if(k==0) k=compare(list,str); |
||
| 60 | if(k==0) goto more; if(k>0) return -1; |
||
| 61 | j=items-1; k=list[j].original[0]-c; if(k==0) k=compare(list+j,str); |
||
| 62 | if(k==0) return j; |
||
| 63 | if(k>0) for(i1=0,i2=j;i2>i1+1;) { |
||
| 64 | j=i1+(i2-i1)/2; |
||
| 65 | k=list[j].original[0]-c; if(k==0) k=compare(list+j,str); |
||
| 66 | if(k==0) goto more; |
||
| 67 | if(k>0) {i2=j; continue;} |
||
| 68 | if(k<0) {i1=j; continue;} |
||
| 69 | } |
||
| 70 | if(k>0) {j--;k=compare(list+j,str);} |
||
| 71 | more: |
||
| 72 | if((t=list[j].earlier)<0) { |
||
| 73 | if(k==0) return j; else return -1; |
||
| 74 | } |
||
| 75 | if(compare(entry+t,str)!=0) return -1; |
||
| 76 | for(j=t1=t,k=0;j<items+(list-entry) && entry[j].earlier==t1 && (k=compare(entry+j,str))<=0; j++) |
||
| 77 | if(k==0) t=j; |
||
| 78 | return t-(list-entry); |
||
| 79 | } |
||
| 80 | |||
| 81 | #include "suffix.c" |
||
| 82 | |||
| 83 | /* Prepare dictionary */ |
||
| 84 | struct dic *prepare_dic(char *fname) |
||
| 85 | { |
||
| 86 | int i,l; |
||
| 87 | struct dic *thisdic; |
||
| 88 | FILE *dicf; |
||
| 89 | char *p1, *p2, *pp; |
||
| 90 | char tbuf[256], buf[MAX_LINELEN+1]; |
||
| 91 | long int flen; |
||
| 92 | |||
| 93 | if(diccnt>=MAX_DICS) error("too_many_dictionaries"); |
||
| 94 | thisdic=dic+diccnt; diccnt++; |
||
| 95 | thisdic->len=0; |
||
| 96 | thisdic->start=entrycount; |
||
| 97 | snprintf(thisdic->name,sizeof(thisdic->name),"%s",fname); |
||
| 98 | dicf=fopen(mkfname(NULL,"%s/%s",styledir,fname),"r"); if(dicf==NULL) return NULL; |
||
| 99 | fseek(dicf,0,SEEK_END);flen=ftell(dicf); fseek(dicf,0,SEEK_SET); |
||
| 100 | if(flen>=MAX_DICSIZE) return NULL; |
||
| 101 | thisdic->buf=xmalloc(flen+16);flen=fread(thisdic->buf,1,flen,dicf); |
||
| 102 | fclose(dicf); |
||
| 103 | if(flen>0 && flen<MAX_DICSIZE) thisdic->buf[flen]=0; |
||
| 104 | else return NULL; |
||
| 105 | for(i=entrycount,p1=thisdic->buf;p1!=NULL && *p1!=0 && i<MAX_DICENTRIES;p1=p2) { |
||
| 106 | p2=strchr(p1+1,'\n'); if(p2>p1) *p2++=0; |
||
| 107 | pp=strchr(p1,':'); if(pp==NULL) continue; |
||
| 108 | *pp++=0; |
||
| 109 | strip_trailing_spaces(p1); strip_trailing_spaces(pp); |
||
| 110 | singlespace(p1); |
||
| 111 | p1=find_word_start(p1); pp=find_word_start(pp); |
||
| 112 | if(*p1==0) continue; |
||
| 113 | if(i>entrycount && compare(entry+i-1,p1)>0) |
||
| 114 | error("unsorted_dictionary %s: %s > %s.\n", |
||
| 115 | fname,entry[i-1].original,p1); |
||
| 116 | if(i>entrycount && strcmp(entry[i-1].original,p1)==0) |
||
| 117 | error("duplication_in_dictionary %s: %s.\n", |
||
| 118 | fname,p1); |
||
| 119 | entry[i].original=p1; entry[i].replace=pp; |
||
| 120 | entry[i].olen=l=strlen(p1); entry[i].earlier=-1; |
||
| 121 | if(i>0) { |
||
| 122 | int l1,l2; |
||
| 123 | l1=entry[i-1].earlier; if(l1>=0) l2=entry[l1].olen; |
||
| 124 | else {l2=entry[i-1].olen;l1=i-1;} |
||
| 125 | if(l>l2 && isspace(p1[l2]) |
||
| 126 | && strncmp(entry[l1].original,p1,l2)==0) |
||
| 127 | entry[i].earlier=entry[i-1].earlier=l1; |
||
| 128 | } |
||
| 129 | i++; |
||
| 130 | } |
||
| 131 | thisdic->len=i-entrycount; |
||
| 132 | pp=strrchr("fname",'/'); if(pp==NULL) pp=fname; |
||
| 133 | snprintf(tbuf,sizeof(tbuf),"unknown_%s",pp); |
||
| 134 | _getdef(defbuf,tbuf,buf); |
||
| 135 | p1=find_word_start(buf); *find_word_end(p1)=0; |
||
| 136 | for(pp=p1; *pp; pp++) *pp=tolower(*pp); |
||
| 137 | thisdic->unknown_type=unk_delete; |
||
| 138 | if(strcmp(p1,"leave")==0) thisdic->unknown_type=unk_leave; |
||
| 139 | else if(strcmp(p1,"delete")!=0) { |
||
| 140 | thisdic->unknown_type=unk_replace; |
||
| 141 | snprintf(thisdic->unknown,sizeof(thisdic->unknown),"%s",p1); |
||
| 142 | } |
||
| 143 | entrycount=i; |
||
| 144 | if(debug) fprintf(stderr,"Dictionary %d: %s, %d entries.\n", |
||
| 145 | diccnt,fname,thisdic->len); |
||
| 146 | return thisdic; |
||
| 147 | } |
||
| 148 | |||
| 149 | /* make the translation. */ |
||
| 150 | void _translate(char *p, int i) |
||
| 151 | { |
||
| 152 | char *p1, *p2, *pp; |
||
| 153 | int t; |
||
| 154 | |||
| 155 | if(i<0 || i>=diccnt) return; |
||
| 156 | if(dic[i].len<=0) return; |
||
| 157 | snprintf(troutbuf,sizeof(troutbuf),"%s",p); |
||
| 158 | for(p1=find_word_start(troutbuf); |
||
| 159 | p1!=NULL && p1-troutbuf<MAX_LINELEN && *p1!=0; |
||
| 160 | p1=p2) { |
||
| 161 | p2=find_word_end(p1); |
||
| 162 | for(pp=p1;pp<p2 && (isalnum(*pp) || strchr("_",*pp)!=NULL);pp++); |
||
| 163 | p2=find_word_start(p2); |
||
| 164 | if(pp==p1 || (*pp!=0 && strchr(" ,.?!",*pp)==NULL)) continue; |
||
| 165 | t=search_dic(entry+dic[i].start,dic[i].len,sizeof(entry[0]),p1); |
||
| 166 | if(t<0) { |
||
| 167 | switch(dic[i].unknown_type) { |
||
| 168 | case unk_leave: break; |
||
| 169 | case unk_delete: { |
||
| 3718 | reyssat | 170 | ovlstrcpy(p1,find_word_start(pp)); p2=p1; |
| 10 | reyssat | 171 | break; |
| 172 | } |
||
| 173 | case unk_replace: { |
||
| 174 | string_modify(troutbuf,p1,pp,dic[i].unknown); |
||
| 175 | p2=find_word_start(p1+strlen(dic[i].unknown)); |
||
| 176 | } |
||
| 177 | } |
||
| 178 | continue; |
||
| 179 | } |
||
| 180 | t+=dic[i].start; |
||
| 181 | string_modify(troutbuf,p1,p1+strlen(entry[t].original), |
||
| 182 | entry[t].replace); |
||
| 183 | p2=find_word_start(p1+strlen(entry[t].replace)); |
||
| 184 | } |
||
| 185 | snprintf(p,MAX_LINELEN,"%s",troutbuf); |
||
| 186 | } |
||
| 187 | |||
| 188 | /* make translation using file name */ |
||
| 189 | void translate(char *p, char *dicname) |
||
| 190 | { |
||
| 191 | int i; |
||
| 192 | for(i=0;i<diccnt && strcmp(dicname,dic[i].name)!=0;i++); |
||
| 193 | if(i<diccnt) _translate(p,i); |
||
| 194 | } |
||
| 195 | |||
| 196 | /* Returns dictionary index, or -1 if not found */ |
||
| 197 | int getdic(char *dicname) |
||
| 198 | { |
||
| 199 | int i; |
||
| 200 | char *p1, *p2, buf[MAX_LINELEN+1]; |
||
| 201 | for(i=0;i<diccnt && strcmp(dicname,dic[i].name)!=0;i++); |
||
| 202 | if(i<diccnt) return i; |
||
| 203 | _getdef(defbuf,"dictionaries",buf); |
||
| 204 | p1=wordchr(buf,dicname); if(p1==NULL) return -1; |
||
| 205 | for(p2=p1; myisalnum(*p2) || *p2=='.'; p2++); |
||
| 206 | if(p2-p1 >= MAX_NAMELEN) return -1; |
||
| 207 | *p2=0; i=diccnt; |
||
| 208 | prepare_dic(dicname); return i; |
||
| 209 | } |
||
| 210 |