Rev 3718 | Go to most recent revision | Details | 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 | /* Check type of a file */ |
||
| 19 | |||
| 20 | /*************** Customization: change values hereafter ****************/ |
||
| 21 | |||
| 22 | /* limit of data buffers */ |
||
| 23 | #define buflim 1024*1024*16 |
||
| 24 | |||
| 25 | /***************** Nothing should need change hereafter *****************/ |
||
| 26 | |||
| 27 | #include "../wims.h" |
||
| 28 | #include "../hmname.c" |
||
| 29 | |||
| 30 | char fn1[1024]="", fn2[1024]=""; |
||
| 31 | char mathbuf[MAX_LINELEN+1]; |
||
| 32 | char *filebuf; |
||
| 33 | int filelen=0; |
||
| 34 | int latex2html=0; |
||
| 35 | FILE *outf; |
||
| 36 | |||
| 37 | struct { |
||
| 38 | char *name, *trans; |
||
| 39 | } backtrans[]={ |
||
| 40 | {"\\ge\\", " >= "}, |
||
| 41 | {"\\geq\\", " >= "}, |
||
| 42 | {"\\le\\", " <= "}, |
||
| 43 | {"\\leq\\", " <= "}, |
||
| 44 | {"\\to\\", " -> "}, |
||
| 45 | {"\\rightarrow\\", " -> "}, |
||
| 46 | {"\\longrightarrow\\", " --> "}, |
||
| 47 | {"\\Rightarrow\\", " => "}, |
||
| 48 | {"\\Longrightarrow\\", " ==> "}, |
||
| 49 | {"\\Leftrightarrow\\", " <=> "}, |
||
| 50 | {"\\Longleftrightarrow\\", " <==> "}, |
||
| 51 | {"\\Longleftarrow\\", " <== "}, |
||
| 52 | }; |
||
| 53 | |||
| 54 | #define backtransno (sizeof(backtrans)/sizeof(backtrans[0])) |
||
| 55 | |||
| 56 | void *xmalloc(size_t n) |
||
| 57 | { |
||
| 58 | void *p; |
||
| 59 | p=malloc(n); |
||
| 60 | if(p==NULL) exit(1); |
||
| 61 | return p; |
||
| 62 | } |
||
| 63 | |||
| 64 | /* Points to the end of the word */ |
||
| 65 | char *find_word_end(char *p) |
||
| 66 | { |
||
| 67 | int i; |
||
| 68 | for(i=0;!isspace(*p) && *p!=0 && i<MAX_LINELEN; p++,i++); |
||
| 69 | return p; |
||
| 70 | } |
||
| 71 | |||
| 72 | /* Strips leading spaces */ |
||
| 73 | char *find_word_start(char *p) |
||
| 74 | { |
||
| 75 | int i; |
||
| 76 | for(i=0; isspace(*p) && i<MAX_LINELEN; p++,i++); |
||
| 77 | return p; |
||
| 78 | } |
||
| 79 | |||
| 80 | char *find_tag_end(char *p) |
||
| 81 | { |
||
| 82 | char *pp, *old; |
||
| 83 | pp=p; if(*pp=='<') pp++; |
||
| 84 | for(; *pp && *pp!='>'; pp++) { |
||
| 85 | if(*pp=='"') { |
||
| 86 | pp=strchr(pp+1,'"'); |
||
| 87 | if(pp==NULL) {pp=p+strlen(p); break;} else continue; |
||
| 88 | } |
||
| 89 | } |
||
| 90 | /* this is probably an syntax error of the page */ |
||
| 91 | if(*pp==0 && pp>p+2048) { |
||
| 92 | old=p; if(*old=='<') old++; |
||
| 93 | pp=strchr(old,'>'); |
||
| 94 | if(pp==NULL) pp=strchr(old,'<'); |
||
| 95 | if(pp==NULL) pp=find_word_end(find_word_start(old)); |
||
| 96 | } |
||
| 97 | if(*pp=='>') pp++; return pp; |
||
| 98 | } |
||
| 99 | |||
| 100 | char *find_tag(char *p, char *tag) |
||
| 101 | { |
||
| 102 | char *pp; |
||
| 103 | int len; |
||
| 104 | len=strlen(tag); |
||
| 105 | for(pp=strchr(p,'<'); pp!=NULL && *pp; pp=strchr(pp+1,'<')) { |
||
| 106 | if(strncasecmp(pp+1,tag,len)==0 && !isalnum(*(pp+1+len))) return pp; |
||
| 107 | } |
||
| 108 | return p+strlen(p); |
||
| 109 | } |
||
| 110 | |||
| 111 | /* modify a string. Bufferlen must be ast least MAX_LINELEN */ |
||
| 112 | void string_modify(char *start, char *bad_beg, char *bad_end, char *good,...) |
||
| 113 | { |
||
| 114 | char buf[MAX_LINELEN+1]; |
||
| 115 | va_list vp; |
||
| 116 | |||
| 117 | va_start(vp,good); |
||
| 118 | vsnprintf(buf,sizeof(buf),good,vp); va_end(vp); |
||
| 119 | if(strlen(start)-(bad_end-bad_beg)+strlen(buf)>=MAX_LINELEN) { |
||
| 120 | return; |
||
| 121 | } |
||
| 122 | strcat(buf,bad_end); |
||
| 123 | strcpy(bad_beg,buf); |
||
| 124 | } |
||
| 125 | |||
| 126 | void cutamp(char *p) |
||
| 127 | { |
||
| 128 | char *pp; |
||
| 129 | for(pp=strchr(p,'&'); pp; pp=strchr(pp+1,'&')) { |
||
| 130 | if(strncmp(pp,"&",5)==0) { |
||
| 131 | strcpy(pp+1,pp+5); continue; |
||
| 132 | } |
||
| 133 | if(strncmp(pp,"<",4)==0) { |
||
| 134 | *pp='<'; strcpy(pp+1,pp+4); continue; |
||
| 135 | } |
||
| 136 | if(strncmp(pp,">",4)==0) { |
||
| 137 | *pp='>'; strcpy(pp+1,pp+4); continue; |
||
| 138 | } |
||
| 139 | |||
| 140 | } |
||
| 141 | } |
||
| 142 | |||
| 143 | /* get the file */ |
||
| 144 | void prepare_file(void) |
||
| 145 | { |
||
| 146 | FILE *f; |
||
| 147 | long int flen; |
||
| 148 | |||
| 149 | filelen=0; |
||
| 150 | f=fopen(fn1,"r"); if(f==NULL) return; |
||
| 151 | fseek(f,0,SEEK_END);flen=ftell(f); fseek(f,0,SEEK_SET); |
||
| 152 | if(flen>buflim) return; |
||
| 153 | filebuf=xmalloc(2*flen+1024);flen=fread(filebuf,1,flen,f); |
||
| 154 | fclose(f); |
||
| 155 | if(flen<0 || flen>=buflim) flen=0; filebuf[flen]=0; |
||
| 156 | filelen=flen; |
||
| 157 | outf=fopen(fn2,"w"); if(outf==NULL) return; |
||
| 158 | } |
||
| 159 | |||
| 160 | void getmath(char *p) |
||
| 161 | { |
||
| 162 | char *pt, *pv; |
||
| 163 | char *p1, *p2, buf[256]; |
||
| 164 | |||
| 165 | mathbuf[0]=0; |
||
| 166 | pt=find_word_start(p); |
||
| 167 | if(strncmp(pt,"\\begin{displaymath}", |
||
| 168 | strlen("\\begin{displaymath}"))==0) { |
||
| 169 | pt=strchr(pt,'}')+1; |
||
| 170 | pv=strstr(pt,"\\end{displaymath}"); |
||
| 171 | if(pv==NULL) return; |
||
| 172 | goto insmath; |
||
| 173 | } |
||
| 174 | if(*pt=='%') pt=strchr(pt,'$'); if(pt==NULL) return; |
||
| 175 | if(*pt!='$') return; do pt++; while(*pt=='$'); |
||
| 176 | pv=strchr(pt,'$'); if(pv==NULL) return; |
||
| 177 | insmath: if(pv-pt>=MAX_LINELEN-256) return; |
||
| 178 | memmove(mathbuf,pt,pv-pt); mathbuf[pv-pt]=0; |
||
| 179 | if(strstr(mathbuf,"...\n...")!=NULL) { |
||
| 180 | strcpy(mathbuf,"......"); return; |
||
| 181 | } |
||
| 182 | cutamp(mathbuf); latex2html=1; |
||
| 183 | for(p1=strstr(mathbuf,"\\mathbb");p1;p1=strstr(p1+1,"\\mathbb")) { |
||
| 184 | char c,*d; |
||
| 185 | p2=find_word_start(p1+strlen("\\mathbb")); c=0; |
||
| 186 | if(strchr("NZQRC",*p2)!=NULL) c=*p2; |
||
| 187 | else if(*p2=='{' && *(p2+2)=='}' && strchr("NZQRC",*(p2+1))!=NULL) { |
||
| 188 | c=*(p2+1); p2+=2; |
||
| 189 | } |
||
| 190 | if(c) { |
||
| 191 | p2=find_word_start(++p2); |
||
| 192 | if(isalnum(*p2)) d=" "; else d=""; |
||
| 193 | string_modify(mathbuf,p1,p2,"\\%c%c%s",c,c,d); |
||
| 194 | } |
||
| 195 | } |
||
| 196 | for(p1=strstr(mathbuf,"{\\"); p1; p1=strstr(p1+1,"{\\")) { |
||
| 197 | if(p1>mathbuf && isalpha(*(p1-1))) continue; |
||
| 198 | for(p2=p1+2; p2<p1+24 && isalpha(*p2); p2++); |
||
| 199 | if(*p2!='}' || isalnum(*(p2+1))) continue; |
||
| 200 | memmove(buf,p1+1,p2-p1-1); buf[p2-p1-1]='\\'; buf[p2-p1]=0; |
||
| 201 | if(strstr(hmsame,buf)==NULL) continue; |
||
| 202 | strcpy(p2,p2+1); strcpy(p1,p1+1); |
||
| 203 | } |
||
| 204 | if(strstr(mathbuf,"\\begin{")!=NULL) return; |
||
| 205 | for(p1=strchr(mathbuf,'{'); p1; p1=strchr(p1+1,'{')) { |
||
| 206 | if((p1>mathbuf && isalpha(*(p1-1))) || |
||
| 207 | !isalnum(*(p1+1)) || *(p1+2)!='}') continue; |
||
| 208 | *p1=*(p1+1); strcpy(p1+1,p1+3); |
||
| 209 | } |
||
| 210 | if(strchr(mathbuf,'[')!=NULL) { |
||
| 211 | char mbuf[MAX_LINELEN+1]; |
||
| 212 | snprintf(mbuf,sizeof(mbuf),"{%s}",mathbuf); |
||
| 213 | strcpy(mathbuf,mbuf); |
||
| 214 | } |
||
| 215 | /* try to simplify */ |
||
| 216 | if(strchr(mathbuf,'{')==NULL && strchr(mathbuf,'\\')!=NULL) { |
||
| 217 | int i, tt; |
||
| 218 | tt=0; |
||
| 219 | for(p1=strchr(mathbuf,'\\'); p1; p1=strchr(p1+1,'\\')) { |
||
| 220 | for(p2=p1+1;isalpha(*p2);p2++); |
||
| 221 | if(p2==p1+1 || p2>p1+24) {tt=1; break;} |
||
| 222 | memmove(buf,p1,p2-p1);buf[p2-p1]='\\';buf[p2-p1+1]=0; |
||
| 223 | for(i=0;i<backtransno && strcmp(buf,backtrans[i].name)!=0;i++); |
||
| 224 | if(i>=backtransno && strstr(hmsame,buf)==NULL) { |
||
| 225 | tt=1; break; |
||
| 226 | } |
||
| 227 | } |
||
| 228 | if(tt==0) { |
||
| 229 | for(p1=strchr(mathbuf,'\\'); p1; p1=strchr(p1+1,'\\')) { |
||
| 230 | for(p2=p1+1;isalpha(*p2);p2++); |
||
| 231 | if(p2==p1+1 || p2>p1+24) break; |
||
| 232 | memmove(buf,p1,p2-p1);buf[p2-p1]='\\';buf[p2-p1+1]=0; |
||
| 233 | for(i=0;i<backtransno && strcmp(buf,backtrans[i].name)!=0;i++); |
||
| 234 | if(i<backtransno) |
||
| 235 | string_modify(buf,p1,p2,backtrans[i].trans); |
||
| 236 | else *p1=' '; |
||
| 237 | } |
||
| 238 | } |
||
| 239 | } |
||
| 240 | } |
||
| 241 | |||
| 242 | void output(void) |
||
| 243 | { |
||
| 244 | char *p, *pp, *p2, *pt; |
||
| 245 | char buf[MAX_LINELEN+1]; |
||
| 246 | p=filebuf; |
||
| 247 | restart: |
||
| 248 | pp=find_tag(p,"body"); if(*pp!=0) { |
||
| 249 | p=find_tag_end(pp); goto restart; |
||
| 250 | } |
||
| 251 | pp=find_tag(p,"html"); if(*pp!=0) { |
||
| 252 | p=find_tag_end(pp); goto restart; |
||
| 253 | } |
||
| 254 | *find_tag(p,"/body")=0; *find_tag(p,"/html")=0; |
||
| 255 | for(pp=strstr(p,"\n\n"); pp; pp=strstr(pp+1,"\n\n")) *pp=' '; |
||
| 256 | for(pp=strchr(p,'<');pp!=NULL;pp=strchr(find_tag_end(pp),'<')) { |
||
| 257 | if(pp>p) {fwrite(p,1,pp-p,outf); p=pp;} |
||
| 258 | if(latex2html && strncasecmp(pp,"<br><hr>",8)==0 && |
||
| 259 | *find_word_start(pp+8)==0) break; |
||
| 260 | if(strncasecmp(pp+1,"!-- MATH",8)==0) { |
||
| 261 | p2=strstr(pp+8,"-->"); if(p2==NULL) continue; |
||
| 262 | *p2=0; getmath(pp+9); *p2='-'; |
||
| 263 | p=p2+3; pt=find_word_start(p); |
||
| 264 | if(mathbuf[0] && strncasecmp(pt,"<IMG",4)==0 && isspace(*(pt+4))) { |
||
| 265 | p=find_tag_end(pt); pp=pt; |
||
| 266 | fprintf(outf,"\\(%s\\)",mathbuf); |
||
| 267 | } |
||
| 268 | continue; |
||
| 269 | } |
||
| 270 | if(strncasecmp(pp+1,"a",1)==0 && isspace(*(pp+2))) { |
||
| 271 | |||
| 272 | |||
| 273 | |||
| 274 | continue; |
||
| 275 | } |
||
| 276 | if(strncasecmp(pp+1,"img",3)==0 && isspace(*(pp+4))) { |
||
| 277 | p2=find_tag_end(pp); |
||
| 278 | if(p2-pp>=MAX_LINELEN-256) continue; |
||
| 279 | memmove(buf,pp+1,p2-pp-2); buf[p2-pp-2]=0; |
||
| 280 | pt=strstr(buf,"ALT=\""); if(pt==NULL) pt=strstr(buf,"alt=\""); |
||
| 281 | if(pt!=NULL) { |
||
| 282 | pt+=strlen("ALT=\""); |
||
| 283 | getmath(pt); if(mathbuf[0]) { |
||
| 284 | fprintf(outf,"\\(%s\\)",mathbuf); p=p2; |
||
| 285 | } |
||
| 286 | } |
||
| 287 | } |
||
| 288 | } |
||
| 289 | if(pp==NULL) fprintf(outf,"%s",p); |
||
| 290 | } |
||
| 291 | |||
| 292 | int main(int argc, char *argv[]) |
||
| 293 | { |
||
| 294 | char *p, *pp; |
||
| 295 | char *mod; |
||
| 296 | |||
| 297 | mod=getenv("w_module"); |
||
| 298 | if(mod!=NULL && strncmp(mod,"adm/",4)!=0 && strcmp(mod,"home")!=0) return 1; |
||
| 299 | if(mod==NULL) p=argv[1]; else p=getenv("wims_exec_parm"); |
||
| 300 | if(p==NULL || *p==0) return 1; |
||
| 301 | p=find_word_start(p); pp=find_word_end(p); |
||
| 302 | if(pp<=p || pp-p>sizeof(fn1)-1) return 1; |
||
| 303 | memmove(fn1,p,pp-p); fn1[pp-p]=0; |
||
| 304 | p=find_word_start(pp); pp=find_word_end(p); |
||
| 305 | if(pp<=p || pp-p>sizeof(fn2)-1) strcpy(fn2,fn1); |
||
| 306 | else {memmove(fn2,p,pp-p); fn2[pp-p]=0;} |
||
| 307 | prepare_file(); |
||
| 308 | output(); |
||
| 309 | fclose(outf); |
||
| 310 | return 0; |
||
| 311 | } |
||
| 312 |