topsites.c

   1 /*
   2  * SARG Squid Analysis Report Generator      http://sarg.sourceforge.net
   3  *                                                            1998, 2010
   4  *
   5  * SARG donations:
   6  *      please look at http://sarg.sourceforge.net/donations.php
   7  * Support:
   8  *     http://sourceforge.net/projects/sarg/forums/forum/363374
   9  * ---------------------------------------------------------------------
  10  *
  11  *  This program is free software; you can redistribute it and/or modify
  12  *  it under the terms of the GNU General Public License as published by
  13  *  the Free Software Foundation; either version 2 of the License, or
  14  *  (at your option) any later version.
  15  *
  16  *  This program is distributed in the hope that it will be useful,
  17  *  but WITHOUT ANY WARRANTY; without even the implied warranty of
  18  *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  19  *  GNU General Public License for more details.
  20  *
  21  *  You should have received a copy of the GNU General Public License
  22  *  along with this program; if not, write to the Free Software
  23  *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111, USA.
  24  *
  25  */
  26
  27 #include "include/conf.h"
  28 #include "include/defs.h"
  29
  30 void topsites(void)
  31 {
  32
  33    FILE *fp_in, *fp_ou;
  34
  35    char *buf;
  36    char *url;
  37    char *ourl=NULL;
  38    char csort[255];
  39    char general[MAXLEN];
  40    char general2[MAXLEN];
  41    char general3[MAXLEN];
  42    char sites[MAXLEN];
  43    char report[MAXLEN];
  44    const char *sortf;
  45    const char *sortt;
  46    long long int nacc;
  47    long long int nbytes;
  48    long long int ntime;
  49    long long int tnacc=0;
  50    long long int tnbytes=0;
  51    long long int tntime=0;
  52    long long int twork1=0, twork2=0, twork3=0;
  53    int regs=0;
  54    int cstatus;
  55    int url_len;
  56    int ourl_size=0;
  57    struct getwordstruct gwarea;
  58    longline line;
  59    struct generalitemstruct item;
  60
  61    if(Privacy)
  62       return;
  63
  64    sprintf(general,"%s/sarg-general",outdirname);
  65    sprintf(sites,"%s/sarg-sites",outdirname);
  66    sprintf(general2,"%s/sarg-general2",outdirname);
  67    sprintf(general3,"%s/sarg-general3",outdirname);
  68
  69    if ((ReportType & REPORT_TYPE_TOPUSERS) == 0)
  70       sprintf(report,"%s/index.html",outdirname);
  71    else
  72       sprintf(report,"%s/topsites.html",outdirname);
  73
  74    sprintf(csort,"sort -k 4,4 -o \"%s\" \"%s\"",general2,general);
  75    cstatus=system(csort);
  76    if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
  77       debuga(_("sort command return status %d\n"),WEXITSTATUS(cstatus));
  78       debuga(_("sort command: %s\n"),csort);
  79       exit(EXIT_FAILURE);
  80    }
  81
  82    if((fp_in=fopen(general2,"r"))==NULL) {
  83      debuga(_("(topsites) Cannot open log file %s\n"),general2);
  84      debuga(_("sort command: %s\n"),csort);
  85      exit(EXIT_FAILURE);
  86    }
  87
  88    if((fp_ou=fopen(general3,"w"))==NULL) {
  89      debuga(_("(topsites) Cannot open log file %s\n"),general3);
  90      exit(EXIT_FAILURE);
  91    }
  92
  93    if ((line=longline_create())==NULL) {
  94       debuga(_("Not enough memory to read file %s\n"),general2);
  95       exit(EXIT_FAILURE);
  96    }
  97
  98    while((buf=longline_read(fp_in,line))!=NULL) {
  99       ger_read(buf,&item,general2);
 100       if(item.total) continue;
 101
 102       if(!regs) {
 103          url_len=strlen(item.url);
 104          if (!ourl || url_len>=ourl_size) {
 105             ourl_size=url_len+1;
 106             ourl=realloc(ourl,ourl_size);
 107             if (!ourl) {
 108                debuga(_("Not enough memory to store the url\n"));
 109                exit(EXIT_FAILURE);
 110             }
 111          }
 112          strcpy(ourl,item.url);
 113          regs++;
 114       }
 115
 116       if(strcmp(item.url,ourl) != 0) {
 117          fprintf(fp_ou,"%lld\t%lld\t%lld\t%s\n",tnacc,tnbytes,tntime,ourl);
 118          url_len=strlen(item.url);
 119          if (url_len>=ourl_size) {
 120             ourl_size=url_len+1;
 121             ourl=realloc(ourl,ourl_size);
 122             if (!ourl) {
 123                debuga(_("Not enough memory to store the url\n"));
 124                exit(EXIT_FAILURE);
 125             }
 126          }
 127          strcpy(ourl,item.url);
 128          tnacc=0;
 129          tnbytes=0;
 130          tntime=0;
 131       }
 132
 133       tnacc+=item.nacc;
 134       tnbytes+=item.nbytes;
 135       tntime+=item.nelap;
 136    }
 137    fclose(fp_in);
 138    unlink(general2);
 139    longline_destroy(&line);
 140
 141    if (ourl) {
 142       fprintf(fp_ou,"%lld\t%lld\t%lld\t%s\n",tnacc,tnbytes,tntime,ourl);
 143       free(ourl);
 144    }
 145
 146    fclose(fp_ou);
 147
 148    strlow(TopsitesSortField);
 149    strlow(TopsitesSortType);
 150
 151    if(strcmp(TopsitesSortField,"connect") == 0)
 152       sortf="-k 1,1 -k 2,2";
 153    else if(strcmp(TopsitesSortField,"bytes") == 0)
 154       sortf="-k 2,2 -k 1,1";
 155    else
 156       sortf="";
 157    if(strcmp(TopsitesSortType,"a") == 0)
 158       sortt="";
 159    else if(strcmp(TopsitesSortType,"d") == 0)
 160       sortt="-r";
 161    else
 162       sortt="";
 163
 164    sprintf(csort,"sort %s -n %s -o \"%s\" \"%s\"",sortt,sortf,sites,general3);
 165    cstatus=system(csort);
 166    if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 167       debuga(_("sort command return status %d\n"),WEXITSTATUS(cstatus));
 168       debuga(_("sort command: %s\n"),csort);
 169       exit(EXIT_FAILURE);
 170    }
 171    if((fp_in=fopen(sites,"r"))==NULL) {
 172       debuga(_("(topsites) Cannot open log file %s\n"),sites);
 173       debuga(_("sort command: %s\n"),csort);
 174       exit(EXIT_FAILURE);
 175    }
 176
 177    unlink(general2);
 178    unlink(general3);
 179
 180    if((fp_ou=fopen(report,"w"))==NULL) {
 181      debuga(_("(topsites) Cannot open log file %s\n"),report);
 182      exit(EXIT_FAILURE);
 183    }
 184
 185    write_html_header(fp_ou,(IndexTree == INDEX_TREE_DATE) ? 3 : 1,_("Top sites"));
 186    fputs("<tr><td class=\"header_c\">",fp_ou);
 187    fprintf(fp_ou,_("Period: %s"),period.html);
 188    fputs("</td></tr>\n",fp_ou);
 189    fputs("<tr><th class=\"header_c\">",fp_ou);
 190    fprintf(fp_ou,_("Top %d sites"),TopSitesNum);
 191    fputs("</th></tr>\n",fp_ou);
 192    close_html_header(fp_ou);
 193
 194    fputs("<div class=\"report\"><table cellpadding=\"1\" cellspacing=\"2\">\n",fp_ou);
 195    fprintf(fp_ou,"<tr><th class=\"header_l\">%s</th><th class=\"header_l\">%s</th><th class=\"header_l\">%s</th><th class=\"header_l\">%s</th><th class=\"header_l\">%s</th></tr>\n",_("NUM"),_("ACCESSED SITE"),_("CONNECT"),_("BYTES"),_("TIME"));
 196
 197    regs=0;
 198    ntopsites = 0;
 199
 200    if ((line=longline_create())==NULL) {
 201       debuga(_("Not enough memory to read file %s\n"),sites);
 202       exit(EXIT_FAILURE);
 203    }
 204
 205    while(regs<TopSitesNum && (buf=longline_read(fp_in,line))!=NULL) {
 206       getword_start(&gwarea,buf);
 207       if (getword_atoll(&nacc,&gwarea,'\t')<0) {
 208          debuga(_("Maybe you have a broken record or garbage in your %s file\n"),sites);
 209          exit(EXIT_FAILURE);
 210       }
 211       if (nacc == 0) continue;
 212       if (getword_atoll(&nbytes,&gwarea,'\t')<0 || getword_atoll(&ntime,&gwarea,'\t')<0) {
 213          debuga(_("Maybe you have a broken record or garbage in your %s file\n"),sites);
 214          exit(EXIT_FAILURE);
 215       }
 216       if (getword_ptr(buf,&url,&gwarea,'\t')<0) {
 217          debuga(_("The url is invalid in file %s\n"),sites);
 218          exit(EXIT_FAILURE);
 219       }
 220
 221       twork1=nacc;
 222       twork2=nbytes;
 223       twork3=ntime;
 224
 225       strcpy(wwork1,fixnum(twork1,1));
 226       strcpy(wwork2,fixnum(twork2,1));
 227       strcpy(wwork3,fixtime(twork3));
 228
 229       fprintf(fp_ou,"<tr><td class=\"data\">%d</td><td class=\"data2 link\">",++regs);
 230
 231       if(BlockIt[0] != '\0') {
 232          fprintf(fp_ou,"<a href=\"%s%s?url=\"",wwwDocumentRoot,BlockIt);
 233          output_html_url(fp_ou,url);
 234          fputs("\"><img src=\"../images/sarg-squidguard-block.png\"></a>&nbsp;",fp_ou);
 235       }
 236
 237       fputs("<a href=\"http://",fp_ou);
 238       output_html_url(fp_ou,url);
 239       fputs("\">",fp_ou);
 240       output_html_string(fp_ou,url,100);
 241       fprintf(fp_ou,"</a></td><td class=\"data\">%s</td><td class=\"data\">%s</td><td class=\"data\">%s</td></tr>\n",wwork1,wwork2,wwork3);
 242    }
 243    fclose(fp_in);
 244    longline_destroy(&line);
 245
 246    fputs("</table></div>\n",fp_ou);
 247    if (write_html_trailer(fp_ou)<0)
 248       debuga(_("Write error in file %s\n"),report);
 249    if (fclose(fp_ou)==EOF)
 250       debuga(_("Failed to close file %s - %s\n"),report,strerror(errno));
 251
 252    return;
 253
 254 }