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          /*
 118          This complicated printf is due to Microsoft's inability to comply with any standard. Msvcrt is unable
 119          to print a long long int unless it is exactly 64-bits long.
 120          */
 121          fprintf(fp_ou,"%"PRIu64"\t%"PRIu64"\t%"PRIu64"\t%s\n",(uint64_t)tnacc,(uint64_t)tnbytes,(uint64_t)tntime,ourl);
 122          url_len=strlen(item.url);
 123          if (url_len>=ourl_size) {
 124             ourl_size=url_len+1;
 125             ourl=realloc(ourl,ourl_size);
 126             if (!ourl) {
 127                debuga(_("Not enough memory to store the url\n"));
 128                exit(EXIT_FAILURE);
 129             }
 130          }
 131          strcpy(ourl,item.url);
 132          tnacc=0;
 133          tnbytes=0;
 134          tntime=0;
 135       }
 136
 137       tnacc+=item.nacc;
 138       tnbytes+=item.nbytes;
 139       tntime+=item.nelap;
 140    }
 141    fclose(fp_in);
 142    unlink(general2);
 143    longline_destroy(&line);
 144
 145    if (ourl) {
 146       /*
 147       This complicated printf is due to Microsoft's inability to comply with any standard. Msvcrt is unable
 148       to print a long long int unless it is exactly 64-bits long.
 149       */
 150       fprintf(fp_ou,"%"PRIu64"\t%"PRIu64"\t%"PRIu64"\t%s\n",(uint64_t)tnacc,(uint64_t)tnbytes,(uint64_t)tntime,ourl);
 151       free(ourl);
 152    }
 153
 154    fclose(fp_ou);
 155
 156    if((TopsitesSort & TOPSITE_SORT_CONNECT) != 0) {
 157       sortf="-k 1,1 -k 2,2";
 158    } else {
 159       sortf="-k 2,2 -k 1,1";
 160    }
 161    if((TopsitesSort & TOPSITE_SORT_REVERSE) != 0) {
 162       sortt="-r";
 163    } else {
 164       sortt="";
 165    }
 166
 167    sprintf(csort,"sort %s -n %s -o \"%s\" \"%s\"",sortt,sortf,sites,general3);
 168    cstatus=system(csort);
 169    if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 170       debuga(_("sort command return status %d\n"),WEXITSTATUS(cstatus));
 171       debuga(_("sort command: %s\n"),csort);
 172       exit(EXIT_FAILURE);
 173    }
 174    if((fp_in=fopen(sites,"r"))==NULL) {
 175       debuga(_("(topsites) Cannot open log file %s\n"),sites);
 176       debuga(_("sort command: %s\n"),csort);
 177       exit(EXIT_FAILURE);
 178    }
 179
 180    unlink(general2);
 181    unlink(general3);
 182
 183    if((fp_ou=fopen(report,"w"))==NULL) {
 184      debuga(_("(topsites) Cannot open log file %s\n"),report);
 185      exit(EXIT_FAILURE);
 186    }
 187
 188    write_html_header(fp_ou,(IndexTree == INDEX_TREE_DATE) ? 3 : 1,_("Top sites"),HTML_JS_SORTTABLE);
 189    fputs("<tr><td class=\"header_c\">",fp_ou);
 190    fprintf(fp_ou,_("Period: %s"),period.html);
 191    fputs("</td></tr>\n",fp_ou);
 192    fputs("<tr><th class=\"header_c\">",fp_ou);
 193    fprintf(fp_ou,_("Top %d sites"),TopSitesNum);
 194    fputs("</th></tr>\n",fp_ou);
 195    close_html_header(fp_ou);
 196
 197    fputs("<div class=\"report\"><table cellpadding=\"1\" cellspacing=\"2\"",fp_ou);
 198    if (SortTableJs[0]) fputs(" class=\"sortable\"",fp_ou);
 199    fputs(">\n",fp_ou);
 200    fprintf(fp_ou,"<thead><tr><th class=\"header_l\">%s</th><th class=\"header_l",_("NUM"));
 201    if (SortTableJs[0]) fputs(" sorttable_alpha",fp_ou);
 202    fprintf(fp_ou,"\">%s</th><th class=\"header_l\">%s</th><th class=\"header_l\">%s</th><th class=\"header_l\">%s</th></tr></thead>\n",_("ACCESSED SITE"),_("CONNECT"),_("BYTES"),_("TIME"));
 203
 204    regs=0;
 205    ntopsites = 0;
 206
 207    if ((line=longline_create())==NULL) {
 208       debuga(_("Not enough memory to read file %s\n"),sites);
 209       exit(EXIT_FAILURE);
 210    }
 211
 212    while(regs<TopSitesNum && (buf=longline_read(fp_in,line))!=NULL) {
 213       getword_start(&gwarea,buf);
 214       if (getword_atoll(&nacc,&gwarea,'\t')<0) {
 215          debuga(_("Maybe you have a broken record or garbage in your %s file\n"),sites);
 216          exit(EXIT_FAILURE);
 217       }
 218       if (nacc == 0) continue;
 219       if (getword_atoll(&nbytes,&gwarea,'\t')<0 || getword_atoll(&ntime,&gwarea,'\t')<0) {
 220          debuga(_("Maybe you have a broken record or garbage in your %s file\n"),sites);
 221          exit(EXIT_FAILURE);
 222       }
 223       if (getword_ptr(buf,&url,&gwarea,'\t')<0) {
 224          debuga(_("The url is invalid in file %s\n"),sites);
 225          exit(EXIT_FAILURE);
 226       }
 227
 228       twork1=nacc;
 229       twork2=nbytes;
 230       twork3=ntime;
 231
 232       fprintf(fp_ou,"<tr><td class=\"data\">%d</td><td class=\"data2 link\">",++regs);
 233
 234       if(BlockIt[0] != '\0') {
 235          fprintf(fp_ou,"<a href=\"%s%s?url=\"",wwwDocumentRoot,BlockIt);
 236          output_html_url(fp_ou,url);
 237          fputs("\"><img src=\"../images/sarg-squidguard-block.png\"></a>&nbsp;",fp_ou);
 238       }
 239
 240       fputs("<a href=\"http://",fp_ou);
 241       output_html_url(fp_ou,url);
 242       fputs("\">",fp_ou);
 243       output_html_string(fp_ou,url,100);
 244       fputs("</a></td><td class=\"data\"",fp_ou);
 245       if (SortTableJs[0]) fprintf(fp_ou," sorttable_customkey=\"%"PRId64"\"",(uint64_t)twork1);
 246       fprintf(fp_ou,">%s</td>",fixnum(twork1,1));
 247       fputs("<td class=\"data\"",fp_ou);
 248       if (SortTableJs[0]) fprintf(fp_ou," sorttable_customkey=\"%"PRId64"\"",(uint64_t)twork2);
 249       fprintf(fp_ou,">%s</td>",fixnum(twork2,1));
 250       fputs("<td class=\"data\"",fp_ou);
 251       if (SortTableJs[0]) fprintf(fp_ou," sorttable_customkey=\"%"PRId64"\"",(uint64_t)twork3);
 252       fprintf(fp_ou,">%s</td></tr>\n",fixtime(twork3));
 253    }
 254    fclose(fp_in);
 255    longline_destroy(&line);
 256
 257    fputs("</table></div>\n",fp_ou);
 258    if (write_html_trailer(fp_ou)<0)
 259       debuga(_("Write error in file %s\n"),report);
 260    if (fclose(fp_ou)==EOF)
 261       debuga(_("Failed to close file %s - %s\n"),report,strerror(errno));
 262
 263    return;
 264
 265 }