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 per[MAXLEN];
  43    char sites[MAXLEN];
  44    char report[MAXLEN];
  45    char period[100];
  46    const char *sortf;
  47    const char *sortt;
  48    long long int nacc;
  49    long long int nbytes;
  50    long long int ntime;
  51    long long int tnacc=0;
  52    long long int tnbytes=0;
  53    long long int tntime=0;
  54    long long int twork1=0, twork2=0, twork3=0;
  55    int regs=0;
  56    int cstatus;
  57    int url_len;
  58    int ourl_size=0;
  59    struct getwordstruct gwarea;
  60    struct longlinestruct line;
  61    struct generalitemstruct item;
  62
  63    if(Privacy)
  64       return;
  65
  66    sprintf(general,"%s/sarg-general",dirname);
  67    sprintf(sites,"%s/sarg-sites",dirname);
  68    sprintf(general2,"%s/sarg-general2",dirname);
  69    sprintf(general3,"%s/sarg-general3",dirname);
  70    sprintf(per,"%s/sarg-period",dirname);
  71
  72    if ((ReportType & REPORT_TYPE_TOPUSERS) == 0)
  73       sprintf(report,"%s/index.html",dirname);
  74    else
  75       sprintf(report,"%s/topsites.html",dirname);
  76
  77    if ((fp_in = fopen(per, "r")) == 0) {
  78       fprintf(stderr, "SARG: (topuser) %s: %s\n",_("Cannot open file"),per);
  79       exit(1);
  80    }
  81
  82    if (!fgets(period,sizeof(period),fp_in)) {
  83       fprintf(stderr,"SARG: (topuser) read error in %s\n",per);
  84       exit(1);
  85    }
  86    fclose(fp_in);
  87
  88    sprintf(csort,"sort -k 4,4 -o \"%s\" \"%s\"",general2,general);
  89    cstatus=system(csort);
  90    if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
  91       fprintf(stderr, "SARG: sort command return status %d\n",WEXITSTATUS(cstatus));
  92       fprintf(stderr, "SARG: sort command: %s\n",csort);
  93       exit(1);
  94    }
  95
  96    if((fp_in=fopen(general2,"r"))==NULL) {
  97      fprintf(stderr, "SARG: (topsite) %s: %s\n",_("Cannot open log file"),general2);
  98      fprintf(stderr, "SARG: sort command: %s\n",csort);
  99      exit(1);
 100    }
 101
 102    if((fp_ou=fopen(general3,"w"))==NULL) {
 103      fprintf(stderr, "SARG: (topsite) %s: %s\n",_("Cannot open log file"),general3);
 104      exit(1);
 105    }
 106
 107    if (longline_prepare(&line)<0) {
 108       debuga(_("Not enough memory to read file %s"),general2);
 109       exit(1);
 110    }
 111
 112    while((buf=longline_read(fp_in,&line))!=NULL) {
 113       ger_read(buf,&item,general2);
 114       if(item.total) continue;
 115
 116       if(!regs) {
 117          url_len=strlen(item.url);
 118          if (!ourl || url_len>=ourl_size) {
 119             ourl_size=url_len+1;
 120             ourl=realloc(ourl,ourl_size);
 121             if (!ourl) {
 122                debuga(_("Not enough memory to store the url"));
 123                exit(1);
 124             }
 125          }
 126          strcpy(ourl,item.url);
 127          regs++;
 128       }
 129
 130       if(strcmp(item.url,ourl) != 0) {
 131          fprintf(fp_ou,"%lld\t%lld\t%lld\t%s\n",tnacc,tnbytes,tntime,ourl);
 132          url_len=strlen(item.url);
 133          if (url_len>=ourl_size) {
 134             ourl_size=url_len+1;
 135             ourl=realloc(ourl,ourl_size);
 136             if (!ourl) {
 137                debuga(_("Not enough memory to store the url"));
 138                exit(1);
 139             }
 140          }
 141          strcpy(ourl,item.url);
 142          tnacc=0;
 143          tnbytes=0;
 144          tntime=0;
 145       }
 146
 147       tnacc+=item.nacc;
 148       tnbytes+=item.nbytes;
 149       tntime+=item.nelap;
 150    }
 151    fclose(fp_in);
 152    unlink(general2);
 153    longline_free(&line);
 154
 155    if (ourl) {
 156       fprintf(fp_ou,"%lld\t%lld\t%lld\t%s\n",tnacc,tnbytes,tntime,ourl);
 157       free(ourl);
 158    }
 159
 160    fclose(fp_ou);
 161
 162    strlow(TopsitesSortField);
 163    strlow(TopsitesSortType);
 164
 165    if(strcmp(TopsitesSortField,"connect") == 0)
 166       sortf="-k 1,1 -k 2,2";
 167    else if(strcmp(TopsitesSortField,"bytes") == 0)
 168       sortf="-k 2,2 -k 1,1";
 169    else
 170       sortf="";
 171    if(strcmp(TopsitesSortType,"a") == 0)
 172       sortt="";
 173    else if(strcmp(TopsitesSortType,"d") == 0)
 174       sortt="-r";
 175    else
 176       sortt="";
 177
 178    sprintf(csort,"sort %s -n %s -o \"%s\" \"%s\"",sortt,sortf,sites,general3);
 179    cstatus=system(csort);
 180    if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 181       fprintf(stderr, "SARG: sort command return status %d\n",WEXITSTATUS(cstatus));
 182       fprintf(stderr, "SARG: sort command: %s\n",csort);
 183       exit(1);
 184    }
 185    if((fp_in=fopen(sites,"r"))==NULL) {
 186       fprintf(stderr, "SARG: (topsite) %s: %s\n",_("Cannot open log file"),sites);
 187       fprintf(stderr, "SARG: sort command: %s\n",csort);
 188       exit(1);
 189    }
 190
 191    unlink(general2);
 192    unlink(general3);
 193
 194    if((fp_ou=fopen(report,"w"))==NULL) {
 195      fprintf(stderr, "SARG: (topsite) %s: %s\n",_("Cannot open log file"),report);
 196      exit(1);
 197    }
 198
 199    write_html_header(fp_ou,(IndexTree == INDEX_TREE_DATE) ? 3 : 1,_("Top sites"));
 200    fprintf(fp_ou,"<tr><td class=\"header_c\">%s: %s</td></tr>\n",_("Period"),period);
 201    fprintf(fp_ou,"<tr><th class=\"header_c\">%s %d %s</th></tr>\n",_("Top"),TopSitesNum,_("sites"));
 202    close_html_header(fp_ou);
 203
 204    fputs("<div class=\"report\"><table cellpadding=\"1\" cellspacing=\"2\">\n",fp_ou);
 205    fputs("<tr><td></td></tr>\n",fp_ou);
 206    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"));
 207
 208    regs=0;
 209    ntopsites = 0;
 210
 211    if (longline_prepare(&line)<0) {
 212       debuga(_("Not enough memory to read file %s"),sites);
 213       exit(1);
 214    }
 215
 216    while(regs<TopSitesNum && (buf=longline_read(fp_in,&line))!=NULL) {
 217       getword_start(&gwarea,buf);
 218       if (getword_atoll(&nacc,&gwarea,'\t')<0) {
 219          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",sites);
 220          exit(1);
 221       }
 222       if (nacc == 0) continue;
 223       if (getword_atoll(&nbytes,&gwarea,'\t')<0 || getword_atoll(&ntime,&gwarea,'\t')<0) {
 224          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",sites);
 225          exit(1);
 226       }
 227       if (getword_ptr(buf,&url,&gwarea,'\t')<0) {
 228          debuga(_("Maybe you have a broken url in your %s file"),sites);
 229          exit(1);
 230       }
 231
 232       twork1=nacc;
 233       twork2=nbytes;
 234       twork3=ntime;
 235
 236       strcpy(wwork1,fixnum(twork1,1));
 237       strcpy(wwork2,fixnum(twork2,1));
 238       strcpy(wwork3,fixtime(twork3));
 239
 240       fprintf(fp_ou,"<tr><td class=\"data\">%d</td><td class=\"data2 link\">",++regs);
 241
 242       if(BlockIt[0] != '\0') {
 243          fprintf(fp_ou,"<a href=\"%s%s?url=\"",wwwDocumentRoot,BlockIt);
 244          output_html_url(fp_ou,url);
 245          fputs("\"><img src=\"../images/sarg-squidguard-block.png\"></a>&nbsp;",fp_ou);
 246       }
 247
 248       fputs("<a href=\"http://",fp_ou);
 249       output_html_url(fp_ou,url);
 250       fputs("\">",fp_ou);
 251       output_html_string(fp_ou,url,100);
 252       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);
 253    }
 254    fclose(fp_in);
 255    longline_free(&line);
 256
 257    fputs("</table></div>\n",fp_ou);
 258    write_html_trailer(fp_ou);
 259    fclose(fp_ou);
 260
 261    return;
 262
 263 }