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.

 */

/*  This is an internal program,

 * used to index modules for search engine.

 */

#include "../wims.h"

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

#include "translator_.h"

#include "suffix.h"

#define MAX_LANGS    MAX_LANGUAGES

#define MAX_MODULES    65536

char *moduledir=    "public_html/modules";

char *sheetdir=     "public_html/bases/sheet";

char *dicdir=       "public_html/bases";

char *outdir=       "public_html/bases/site2";

char *maindic=      "sys/words";

char *groupdic=     "sys/wgrp/wgrp";

char *suffixdic=    "sys/suffix";

char *domaindic=    "sys/domaindic";

char *ignoredic=    "sys/indignore";

char *conffile=     "log/wims.conf";

char *mlistbase=    "list";

char lang[MAX_LANGS][4]={

    "en","fr","cn","es","it","nl","si","ca","pt"

};

#define DEFAULT_LANGCNT    6

char allang[MAX_LANGS][4]={

    "en","fr","cn","es","it","nl","de","si","ca","pt"

};

#define allangcnt 8

char ignore[MAX_LANGS][MAX_LINELEN+1];

char mlistfile[MAX_LANGS][256];

int langcnt;

FILE *langf, *titf, *descf, *weightf, *robotf, *indf, *listf, *addrf, *serialf, *authorf, *versionf, *remf;

struct cat {

    char *name;

    char typ;

} cat[]={

    {"all_types", 'A'},

    {"exercise",  'X'},

    {"oef",       'O'},

    {"tool",      'T'},

    {"recreation",'R'},

    {"reference", 'Y'},

    {"document",  'D'},

    {"popup",     'P'},

    {"datamodule",'M'}

};

#define catno (sizeof(cat)/sizeof(cat[0]))

struct mod {

    char *name;

    unsigned char langs[MAX_LANGS];

    int counts[MAX_LANGS];

    int  langcnt;

} mod[MAX_MODULES];

int modcnt;

char *mlist;

/*  fold known accented letters to unaccented, other strange characters to space

 *  apostrophe is among the exceptions to be kept (important for multi-word expressions)

 */

void deaccent2(char *p)

{

    char *sp;

    char *v;

    for(sp=p;*sp;sp++) {

    if(*sp<0 && (v=strchr(acctab,*sp))!=NULL)

      *sp=*(deatab+(v-acctab));

    if(!isalnum(*sp) && strchr(",.&$+*",*sp)==0) *sp=' ';

    else *sp=tolower(*sp);

    }

}

/*  translate everything non-alphanumeric into space */

void towords(char *p)

{

    char *pp;

    for(pp=p;*pp;pp++) if(!isalnum(*pp) && strchr("&$+*",*pp)==0) *pp=' ';

}

/*  Find first occurrence of word */

char *wordchr2(char *p, char *w)

{

    char *r;

    for(r=strstr(p,w);r!=NULL &&

    ( (r>p && !isspace(*(r-1))) || (!isspace(*(r+strlen(w))) && *(r+strlen(w))!=0) );

    r=strstr(r+1,w));

    return r;

}

/*  strip trailing spaces; return string end. */

char *strip_trailing_spaces2(char *p)

{

    char *pp;

    if(*p==0) return p;

    for(pp=p+strlen(p)-1; pp>=p && isspace(*pp); *(pp--)=0);

    return pp;

}

char *find_tag_end(char *p)

{

    char *pp;

    pp=p; if(*pp=='<') pp++;

    for(; *pp && *pp!='>'; pp++) {

    if(*pp=='<') {

        pp=find_tag_end(pp)-1; continue;

    }

    if(*pp=='"') {

        pp=strchr(pp+1,'"');

        if(pp==NULL) return p+strlen(p); else continue;

    }

    if(*pp=='\'') {

        pp=strchr(pp+1,'\'');

        if(pp==NULL) return p+strlen(p); else continue;

    }

    }

    if(*pp=='>') pp++; return pp;

}

char *find_tag(char *p, char *tag)

{

    char *pp;

    int len;

    len=strlen(tag);

    for(pp=strchr(p,'<'); pp!=NULL && *pp; pp=strchr(pp+1,'<')) {

    if(strncasecmp(pp+1,tag,len)==0 && !isalnum(*(pp+1+len))) return pp;

    }

    return p+strlen(p);

}

/*  remove all html tags */

void detag(char *p)

{

    char *pp, *p2;

    for(pp=strchr(p,'<'); pp!=NULL; pp=strchr(pp,'<')) {

    p2=find_tag_end(pp);

    if(*p2==0) {*pp=0; return; }

    ovlstrcpy(pp,p2);

    }

}

/*  modify a string. Bufferlen must be at least MAX_LINELEN */

void string_modify3(char *start, char *bad_beg, char *bad_end, char *good,...)

{

    char buf[MAX_LINELEN+1];

    va_list vp;

    va_start(vp,good);

    vsnprintf(buf,sizeof(buf),good,vp); va_end(vp);

    if(strlen(start)-(bad_end-bad_beg)+strlen(buf)>=MAX_LINELEN)

      return; /* this is an error situation. */

    strcat(buf,bad_end);

    ovlstrcpy(bad_beg,buf);

}

/* add a space after comma to see end of words */

void comma(char *p)

{

    char *pp;

    for(pp=strchr(p,','); pp; pp=strchr(pp+1,','))

      string_modify3(p,pp,pp+1,", ");

}

void _getdef(char buf[], char *name, char value[])

{

    char *p1, *p2, *p3;

    value[0]=0;

    for(p1=strstr(buf,name); p1!=NULL; p1=strstr(p1+1,name)) {

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

    if((p1>buf && !isspace(*(p1-1))) || *p2!='=') continue;

    p3=p1; while(p3>buf && isspace(*(p3-1)) && *(p3-1)!='\n') p3--;

    if(p3>buf && *(p3-1)!='\n') continue;

    p3=strchr(p2,'\n');

    p2=find_word_start(p2+1);

    if(p3 <= p2) continue;

    snprintf(value,MAX_LINELEN,"%s",p2);

    if(p3!=NULL && p3-p2<MAX_LINELEN) value[p3-p2]=0;

    strip_trailing_spaces2(value);

    break;

    }

}

/*  Get variable definition from a file.

 * Result stored in buffer value of length MAX_LINELEN.

 */

void getdef(char *fname, char *name, char value[])

{

    FILE *f;

    char *buf;

    int l;

    value[0]=0;

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

    fseek(f,0,SEEK_END); l=ftell(f); fseek(f,0,SEEK_SET);

    buf=xmalloc(l+256); l=fread(buf,1,l,f);

    fclose(f);

    if(l<=0) return; else buf[l]=0;

    _getdef(buf,name,value);

    free(buf);

}

char *mdicbuf, *gdicbuf, *ddicbuf, *gentry, *mentry, *dentry;

int gentrycount, mentrycount, dentrycount;

/*  Preparation of data */

void prep(void)

{

    char buf[MAX_LINELEN+1];

    char *p1,*p2,*s,*old;

    int i,l,thislang,t;

    FILE *f;

    s=getenv("modind_outdir"); if(s!=NULL && *s!=0) outdir=s;

    s=getenv("modind_sheetdir"); if(s!=NULL && *s!=0) sheetdir=s;

    snprintf(buf,sizeof(buf),"%s/addr",outdir);

    addrf=fopen(buf,"w");

    snprintf(buf,sizeof(buf),"%s/serial",outdir);

    serialf=fopen(buf,"w");

    modcnt=langcnt=0;

/* take the langs declared in conffile */

    getdef(conffile,"site_languages",buf);

    for(p1=buf;*p1;p1++) if(!isalnum(*p1)) *p1=' ';

    for(p1=find_word_start(buf); *p1 && langcnt<MAX_LANGS; p1=find_word_start(p2)) {

    p2=find_word_end(p1);

    if(p2!=p1+2 || !isalpha(*p1) || !isalpha(*(p1+1))) continue;

    memmove(lang[langcnt],p1,2); lang[langcnt++][2]=0;

    }

    if(langcnt==0) {/*  default languages */

    langcnt=DEFAULT_LANGCNT;

    }

    s=getenv("mlist"); if(s==NULL) exit(1);

    l=strlen(s); if(l<0 || l>100*MAX_LINELEN) exit(1);

    mlist=xmalloc(l+16); ovlstrcpy(mlist,s); old="";

    for(i=0;i<langcnt;i++) {

    snprintf(buf,sizeof(buf),"%s/%s.%s",dicdir,ignoredic,lang[i]);

    f=fopen(buf,"r"); if(f==NULL) continue;

    l=fread(ignore[i],1,MAX_LINELEN,f);fclose(f);

    if(l<0 || l>=MAX_LINELEN) l=0;

    ignore[i][l]=0;

    }

    for(t=0, p1=find_word_start(mlist);

    *p1 && modcnt<MAX_MODULES;

    p1=find_word_start(p2), t++) {

    p2=find_word_end(p1);

    l=p2-p1; if(*p2) *p2++=0;

    fprintf(addrf,"%d:%s\n",t,p1);

    fprintf(serialf,"%s:%d\n",p1,t);

    thislang=-1;

/* language is taken from the address */

    if(l>3 && p1[l-3]=='.') {

        for(i=0;i<langcnt;i++) if(strcasecmp(lang[i],p1+l-2)==0) break;

        if(i<langcnt) {p1[l-3]=0; thislang=i;}

        else {/*  unknown language, not referenced */

        continue;

        }

    }

    if(modcnt>0 && strcmp(old,p1)==0 && thislang>=0) {

        if(mod[modcnt-1].langcnt<langcnt) {

        mod[modcnt-1].langs[mod[modcnt-1].langcnt]=thislang;

        mod[modcnt-1].counts[mod[modcnt-1].langcnt]=t;

        (mod[modcnt-1].langcnt)++;

        }

    }

    else {

        mod[modcnt].name=old=p1;

        if(thislang>=0) {

        mod[modcnt].langs[0]=thislang;

        mod[modcnt].langcnt=1;

        }

        else mod[modcnt].langcnt=0;

        mod[modcnt].counts[0]=t;

        modcnt++;

    }

    }

    snprintf(buf,sizeof(buf),"%s/language",outdir);

    langf=fopen(buf,"w");

    snprintf(buf,sizeof(buf),"%s/title",outdir);

    titf=fopen(buf,"w");

    snprintf(buf,sizeof(buf),"%s/description",outdir);

    descf=fopen(buf,"w");

    snprintf(buf,sizeof(buf),"%s/author",outdir);

    authorf=fopen(buf,"w");

    snprintf(buf,sizeof(buf),"%s/version",outdir);

    versionf=fopen(buf,"w");

    snprintf(buf,sizeof(buf),"%s/lists/robot.phtml",outdir);

    robotf=fopen(buf,"w");

    fclose(addrf); fclose(serialf);

    if(!robotf || !versionf || !authorf || !descf || !titf || !descf) {

    fprintf(stderr,"modind: error creating output files.\n");

    exit(1);

    }

}

void sprep(void)

{

    char *p1,*p2,*s;

    int i,l,thislang;

    modcnt=0;

    s=getenv("slist"); if(s==NULL) return;

    l=strlen(s); if(l<0 || l>100*MAX_LINELEN) return;

    mlist=xmalloc(l+16); ovlstrcpy(mlist,s);

    for(p1=find_word_start(mlist); *p1 && modcnt<MAX_MODULES; p1=find_word_start(p2)) {

    p2=find_word_end(p1);

    l=p2-p1; if(*p2) *p2++=0;

    for(i=0;i<langcnt;i++) if(strncasecmp(lang[i],p1,2)==0) break;

    if(i<langcnt) thislang=i; else continue;

    mod[modcnt].name=p1;

    mod[modcnt].langs[0]=thislang;

    mod[modcnt].langcnt=1;

    modcnt++;

    }

}

void clean(void)

{

    fclose(langf); fclose(titf); fclose(descf); fclose(robotf);

    fclose(authorf); fclose(versionf);

}

char *sheetindex[]={

      "title", "description",

      "duration", "severity",

      "level", "domain",

      "keywords", "reserved1", "reserved2", "information"

};

#define SHEETINDEX_NO (sizeof(sheetindex)/sizeof(sheetindex[0]))

char sindbuf[SHEETINDEX_NO][MAX_LINELEN+1];

enum{s_title, s_description,

      s_duration, s_severity,

      s_level, s_domain,

      s_keywords, s_reserved1, s_reserved2,

      s_information

};

char *modindex[]={

      "title", "description",

      "author", "address", "copyright",

      "version", "wims_version", "language",

      "category", "level", "domain", "keywords",

      "keywords_ca", "keywords_en", "keywords_fr", "keywords_it", "keywords_nl",

      "title_ca", "title_en", "title_fr", "title_it", "title_nl",

      "require"

};

#define MODINDEX_NO (sizeof(modindex)/sizeof(modindex[0]))

char indbuf[MODINDEX_NO][MAX_LINELEN+1];

enum{i_title, i_description,

      i_author,i_address,i_copyright,

      i_version,i_wims_version,i_language,

      i_category,i_level,i_domain,i_keywords,

      i_keywords_ca,i_keywords_en,i_keywords_fr,i_keywords_it,i_keywords_nl,

      i_title_ca,i_title_en,i_title_fr,i_title_it,i_title_nl,

      i_require

};

char *module_special_file[]={

    "intro","help","about"

};

#define MODSPEC_NO (sizeof(module_special_file)/sizeof(module_special_file[0]))

char module_language[4];

/*  read and treat module's INDEX file */

int module_index(const char *name)

{

    char *p, fbuf[MAX_LINELEN+1], ibuf[MAX_LINELEN+1];

    FILE *indf;

    int i,l;

    snprintf(fbuf,sizeof(fbuf),"%s/%s/INDEX",moduledir,name);

    indf=fopen(fbuf,"r"); if(indf==NULL) return -1;

    l=fread(ibuf,1,MAX_LINELEN,indf); fclose(indf);

    if(l>0 && l<MAX_LINELEN) ibuf[l]=0; else return -1;

/* treate all fields in *modindex */

    for(i=0;i<MODINDEX_NO;i++) {

    _getdef(ibuf,modindex[i],indbuf[i]);

/*  compatibility precaution */

    if(indbuf[i][0]==':') indbuf[i][0]='.';

    }

    p=find_word_start(indbuf[i_language]);

    if(isalpha(*p) && isalpha(*(p+1))) {

    memmove(module_language,p,2); module_language[2]=0;

    }

    else ovlstrcpy(module_language,"en");

    return 0;

}

int sheet_index(int serial)

{

    char *p1, *p2, fbuf[MAX_LINELEN+1], ibuf[MAX_LINELEN+1];

    FILE *indf;

    int i,l;

    snprintf(fbuf,sizeof(fbuf),"%s/%s.def",sheetdir,mod[serial].name);

    indf=fopen(fbuf,"r"); if(indf==NULL) return -1;

    l=fread(ibuf,1,MAX_LINELEN,indf); fclose(indf);

    if(l>0 && l<MAX_LINELEN) ibuf[l]=0; else return -1;

    for(i=0;i<SHEETINDEX_NO;i++) sindbuf[i][0]=0;

    for(i=0,p1=find_word_start(ibuf);

    i<SHEETINDEX_NO-1 && *p1!=':' && *p1!=0;

    i++,p1=p2) {

    p2=strchr(p1,'\n');

    if(p2!=NULL) *p2++=0; else p2=p1+strlen(p1);

    p1=find_word_start(p1); strip_trailing_spaces2(p1);

    snprintf(sindbuf[i],MAX_LINELEN,"%s",p1);

    }

    p2=strstr(p1,"\n:"); if(p2==NULL) p2=p1+strlen(p1);

    else *p2=0;

    p1=find_word_start(p1); strip_trailing_spaces2(p1);

    for(p2=p1;*p2;p2++) if(*p2=='\n') *p2=' ';

    ovlstrcpy(sindbuf[s_information],p1);

    return 0;

}

unsigned char categories[16];

char taken[MAX_LINELEN+1];

int catcnt, takenlen, tweight;

void appenditem(char *word, int lind, int serial, int weight, char *l)

{

    char nbuf[MAX_LINELEN+1], buf[MAX_LINELEN+1];

    int i, ll;

    char *p;

    FILE *f;

    if(!isalnum(*word) || (ll=strlen(word))<2 ||

       wordchr2(taken,word)!=NULL ||

       wordchr2(ignore[lind],word)!=NULL ||

       takenlen>=MAX_LINELEN-ll-16)

      return;

    if(ll==2 && (!isdigit(word[0]) || !isalpha(word[1]))) return;

    for(p=word;*p;p++) if(!isalnum(*p) && *p!=' ') return;

    taken[takenlen++]=' '; taken[takenlen++]=' ';

    ovlstrcpy(taken+takenlen,word);

    takenlen+=ll; tweight+=weight;

    snprintf(buf,sizeof(buf),"%s:%d?%d\n",word,serial,weight);

    for(i=0;i<catcnt;i++) {

    snprintf(nbuf,sizeof(nbuf),"%s/%c.%s",

         outdir,categories[i],lang[lind]);

    f=fopen(nbuf,"a");

    if(f!=NULL) {fputs(buf,f); fclose(f);}

    }

}

void appenditem1 (char *buf, int lind, int serial, int weight, char *l )

{

  char *p1, *p2 ;

  for(p1=find_word_start(buf); *p1;

    p1=find_word_start(p2)) {

    p2=strchr(p1,',');

    if(p2!=NULL) *p2++=0; else p2=p1+strlen(p1);

    if(strlen(p1)<=0) continue;

    appenditem(p1,lind,serial,weight,module_language);

  }

}

void appenditem2 (char *buf, int lind, int serial, int weight, char *l )

{

  char *p1, *p2 ;

  for(p1=find_word_start(buf);*p1;

    p1=find_word_start(p2)) {

    p2=find_word_end(p1); if(*p2) *p2++=0;

    appenditem(p1,lind,serial,weight,module_language);

  }

}

void onemodule(const char *name, int serial, int lind)

{

    int i;

    unsigned char trlist[]={

    i_title,i_description,i_category,i_domain,i_keywords,

      i_require,i_author,

      i_keywords_ca,i_keywords_en,i_keywords_fr,i_keywords_it,i_keywords_nl,

      i_title_ca,i_title_en,i_title_fr,i_title_it,i_title_nl

    };

    #define trcnt (sizeof(trlist)/sizeof(trlist[0]))

    char *p1, *p2, *pp, *q, buf[MAX_LINELEN+1], lbuf[16];

    FILE *f;

    if(module_index(name)) return;

    towords(indbuf[i_category]);

/*   list the categories (among A=all,X=eXercise,O,D,...) corresponding

 *   to this module

 */

    for(i=catcnt=0;i<catno && catcnt<16;i++) {

    if(wordchr2(indbuf[i_category],cat[i].name)!=NULL)

      categories[catcnt++]=cat[i].typ;

    }

    if(catcnt==0) return;

    if(categories[0]!=cat[0].typ)

      categories[catcnt++]=cat[0].typ;

/*  write module's name in the category.language files, for instance lists/X.fr

 * for french exercises

 */

    for(i=0;i<catcnt;i++) {

    snprintf(buf,sizeof(buf),"%s/lists/%c.%s",

         outdir,categories[i],lang[lind]);

    f=fopen(buf,"a");

    if(f!=NULL) {fprintf(f,"%s\n",name); fclose(f);}

    }

/*   add serial number and language (resp.title, ...) to corresponding file  */

    fprintf(langf,"%d:%s\n",serial,module_language);

    fprintf(titf,"%d:%s\n",serial,indbuf[i_title]);

    fprintf(descf,"%d:%s\n",serial,indbuf[i_description]);

    fprintf(authorf,"%d:%s\n",serial,indbuf[i_author]);

    fprintf(versionf,"%d:%s\n",serial,indbuf[i_version]);

/*   add module's information in html page for robots  */

    snprintf(buf,sizeof(buf),"%s",indbuf[i_description]);

    for(pp=strchr(buf,','); pp; pp=strchr(pp,','))

      string_modify3(buf,pp,pp+1,",");

    if(strcmp(module_language,lang[lind])==0)

      fprintf(robotf,"%s ,%s,%s,%s,%s\n",name,module_language,name,

          indbuf[i_title], buf);

/*   Normalize the information of trlist, using dictionary

 *  -- bases/sys/domain.xx without suffix translation (--> english version)

 */

    entrycount=dentrycount; dicbuf=ddicbuf;

    memmove(entry,dentry,dentrycount*sizeof(entry[0]));

    unknown_type=unk_leave;

    for(i=0;i<trcnt;i++) {

    detag(indbuf[trlist[i]]);

    deaccent2(indbuf[trlist[i]]);

    comma(indbuf[trlist[i]]);

    singlespace2(indbuf[trlist[i]]);

    translate(indbuf[trlist[i]]);

    }

/*   Normalize the information, using dictionary

 *   bases/sys/words.xx with suffix translation

 */

    entrycount=mentrycount; dicbuf=mdicbuf;

    memmove(entry,mentry,mentrycount*sizeof(entry[0]));

    unknown_type=unk_leave;/*  used in translator_.c */

    for(i=0;i<trcnt;i++) {

    suffix_translate(indbuf[trlist[i]]);

    translate(indbuf[trlist[i]]);

    }

/* taken contains all words already seen in the module index */

    taken[0]=0; takenlen=tweight=0;

/*  append words of title  */

    ovlstrcpy(buf,indbuf[i_title]); towords(buf);

    appenditem2(buf,lind,serial,4,module_language);

/*  extract words of every other information except level */

    snprintf(buf,sizeof(buf),"%s %s %s %s %s %s %s %s %s %s %s %s %s %s %s",

         indbuf[i_description],indbuf[i_keywords],

         indbuf[i_keywords_ca],indbuf[i_keywords_en],indbuf[i_keywords_fr],

         indbuf[i_keywords_it],indbuf[i_keywords_nl],

         indbuf[i_title_ca],indbuf[i_title_en],indbuf[i_title_fr],

         indbuf[i_title_it],indbuf[i_title_nl],

         indbuf[i_domain],indbuf[i_require],indbuf[i_author]);

    towords(buf);

    appenditem2(buf,lind,serial,2,module_language);

/*   this time the dictionary is the group dictionary  sys/wgrp/wgrp

 *   with a g (groupdic), not an m (maindic) . see below main, suffix, group.

 *   and delete unknown ?? and translate

 */

    entrycount=gentrycount; dicbuf=gdicbuf;

    memmove(entry,gentry,gentrycount*sizeof(entry[0]));

/*  append words of every title information  */