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         FILE *fp_in, *fp_ou;
  33
  34         char *buf;
  35         char *url;
  36         char *ourl=NULL;
  37         char csort[255];
  38         char general[MAXLEN];
  39         char general2[MAXLEN];
  40         char general3[MAXLEN];
  41         char sites[MAXLEN];
  42         char report[MAXLEN];
  43         const char *sortf;
  44         const char *sortt;
  45         long long int nacc;
  46         long long int nbytes;
  47         long long int ntime;
  48         long long int tnacc=0;
  49         long long int tnbytes=0;
  50         long long int tntime=0;
  51         long long int twork1=0, twork2=0, twork3=0;
  52         int regs=0;
  53         int cstatus;
  54         int url_len;
  55         int ourl_size=0;
  56         struct getwordstruct gwarea;
  57         longline line;
  58         struct generalitemstruct item;
  59
  60         if(Privacy)
  61                 return;
  62
  63         sprintf(general,"%s/sarg-general",outdirname);
  64         sprintf(sites,"%s/sarg-sites",outdirname);
  65         sprintf(general2,"%s/sarg-general2",outdirname);
  66         sprintf(general3,"%s/sarg-general3",outdirname);
  67
  68         if ((ReportType & REPORT_TYPE_TOPUSERS) == 0)
  69                 sprintf(report,"%s/index.html",outdirname);
  70         else
  71                 sprintf(report,"%s/topsites.html",outdirname);
  72
  73         sprintf(csort,"sort -k 4,4 -o \"%s\" \"%s\"",general2,general);
  74         cstatus=system(csort);
  75         if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
  76                 debuga(_("sort command return status %d\n"),WEXITSTATUS(cstatus));
  77                 debuga(_("sort command: %s\n"),csort);
  78                 exit(EXIT_FAILURE);
  79         }
  80
  81         if((fp_in=fopen(general2,"r"))==NULL) {
  82                 debuga(_("(topsites) Cannot open log file %s\n"),general2);
  83                 debuga(_("sort command: %s\n"),csort);
  84                 exit(EXIT_FAILURE);
  85         }
  86
  87         if((fp_ou=fopen(general3,"w"))==NULL) {
  88                 debuga(_("(topsites) Cannot open log file %s\n"),general3);
  89                 exit(EXIT_FAILURE);
  90         }
  91
  92         if ((line=longline_create())==NULL) {
  93                 debuga(_("Not enough memory to read file %s\n"),general2);
  94                 exit(EXIT_FAILURE);
  95         }
  96
  97         while((buf=longline_read(fp_in,line))!=NULL) {
  98                 ger_read(buf,&item,general2);
  99                 if(item.total) continue;
 100
 101                 if(!regs) {
 102                         url_len=strlen(item.url);
 103                         if (!ourl || url_len>=ourl_size) {
 104                                 ourl_size=url_len+1;
 105                                 ourl=realloc(ourl,ourl_size);
 106                                 if (!ourl) {
 107                                         debuga(_("Not enough memory to store the url\n"));
 108                                         exit(EXIT_FAILURE);
 109                                 }
 110                         }
 111                         strcpy(ourl,item.url);
 112                         regs++;
 113                 }
 114
 115                 if(strcmp(item.url,ourl) != 0) {
 116                         /*
 117                         This complicated printf is due to Microsoft's inability to comply with any standard. Msvcrt is unable
 118                         to print a long long int unless it is exactly 64-bits long.
 119                         */
 120                         fprintf(fp_ou,"%"PRIu64"\t%"PRIu64"\t%"PRIu64"\t%s\n",(uint64_t)tnacc,(uint64_t)tnbytes,(uint64_t)tntime,ourl);
 121                         url_len=strlen(item.url);
 122                         if (url_len>=ourl_size) {
 123                                 ourl_size=url_len+1;
 124                                 ourl=realloc(ourl,ourl_size);
 125                                 if (!ourl) {
 126                                         debuga(_("Not enough memory to store the url\n"));
 127                                         exit(EXIT_FAILURE);
 128                                 }
 129                         }
 130                         strcpy(ourl,item.url);
 131                         tnacc=0;
 132                         tnbytes=0;
 133                         tntime=0;
 134                 }
 135
 136                 tnacc+=item.nacc;
 137                 tnbytes+=item.nbytes;
 138                 tntime+=item.nelap;
 139         }
 140         fclose(fp_in);
 141         if (unlink(general2)) {
 142                 debuga(_("Cannot delete %s - %s\n"),general2,strerror(errno));
 143                 exit(EXIT_FAILURE);
 144         }
 145         longline_destroy(&line);
 146
 147         if (ourl) {
 148                 /*
 149                 This complicated printf is due to Microsoft's inability to comply with any standard. Msvcrt is unable
 150                 to print a long long int unless it is exactly 64-bits long.
 151                 */
 152                 fprintf(fp_ou,"%"PRIu64"\t%"PRIu64"\t%"PRIu64"\t%s\n",(uint64_t)tnacc,(uint64_t)tnbytes,(uint64_t)tntime,ourl);
 153                 free(ourl);
 154         }
 155
 156         fclose(fp_ou);
 157
 158         if((TopsitesSort & TOPSITE_SORT_CONNECT) != 0) {
 159                 sortf="-k 1,1 -k 2,2";
 160         } else if((TopsitesSort & TOPSITE_SORT_BYTES) != 0) {
 161                 sortf="-k 2,2 -k 1,1";
 162         } else if((TopsitesSort & TOPSITE_SORT_TIME) != 0) {
 163                 sortf="-k 3,3";
 164         } else {
 165                 sortf="-k 2,2 -k 1,1"; //default is BYTES
 166         }
 167         if((TopsitesSort & TOPSITE_SORT_REVERSE) != 0) {
 168                 sortt="-r";
 169         } else {
 170                 sortt="";
 171         }
 172
 173         sprintf(csort,"sort %s -n %s -o \"%s\" \"%s\"",sortt,sortf,sites,general3);
 174         cstatus=system(csort);
 175         if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 176                 debuga(_("sort command return status %d\n"),WEXITSTATUS(cstatus));
 177                 debuga(_("sort command: %s\n"),csort);
 178                 exit(EXIT_FAILURE);
 179         }
 180         if((fp_in=fopen(sites,"r"))==NULL) {
 181                 debuga(_("(topsites) Cannot open log file %s\n"),sites);
 182                 debuga(_("sort command: %s\n"),csort);
 183                 exit(EXIT_FAILURE);
 184         }
 185
 186         if (unlink(general3)) {
 187                 debuga(_("Cannot delete %s - %s\n"),general3,strerror(errno));
 188                 exit(EXIT_FAILURE);
 189         }
 190
 191         if((fp_ou=fopen(report,"w"))==NULL) {
 192                 debuga(_("(topsites) Cannot open log file %s\n"),report);
 193                 exit(EXIT_FAILURE);
 194         }
 195
 196         write_html_header(fp_ou,(IndexTree == INDEX_TREE_DATE) ? 3 : 1,_("Top sites"),HTML_JS_SORTTABLE);
 197         fputs("<tr><td class=\"header_c\">",fp_ou);
 198         fprintf(fp_ou,_("Period: %s"),period.html);
 199         fputs("</td></tr>\n",fp_ou);
 200         fputs("<tr><th class=\"header_c\">",fp_ou);
 201         fprintf(fp_ou,_("Top %d sites"),TopSitesNum);
 202         fputs("</th></tr>\n",fp_ou);
 203         close_html_header(fp_ou);
 204
 205         fputs("<div class=\"report\"><table cellpadding=\"1\" cellspacing=\"2\"",fp_ou);
 206         if (SortTableJs[0]) fputs(" class=\"sortable\"",fp_ou);
 207         fputs(">\n",fp_ou);
 208         fprintf(fp_ou,"<thead><tr><th class=\"header_l\">%s</th><th class=\"header_l",_("NUM"));
 209         if (SortTableJs[0]) fputs(" sorttable_alpha",fp_ou);
 210         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"));
 211
 212         regs=0;
 213         ntopsites = 0;
 214
 215         if ((line=longline_create())==NULL) {
 216                 debuga(_("Not enough memory to read file %s\n"),sites);
 217                 exit(EXIT_FAILURE);
 218         }
 219
 220         while(regs<TopSitesNum && (buf=longline_read(fp_in,line))!=NULL) {
 221                 getword_start(&gwarea,buf);
 222                 if (getword_atoll(&nacc,&gwarea,'\t')<0) {
 223                         debuga(_("Maybe you have a broken record or garbage in your %s file\n"),sites);
 224                         exit(EXIT_FAILURE);
 225                 }
 226                 if (nacc == 0) continue;
 227                 if (getword_atoll(&nbytes,&gwarea,'\t')<0 || getword_atoll(&ntime,&gwarea,'\t')<0) {
 228                         debuga(_("Maybe you have a broken record or garbage in your %s file\n"),sites);
 229                         exit(EXIT_FAILURE);
 230                 }
 231                 if (getword_ptr(buf,&url,&gwarea,'\t')<0) {
 232                         debuga(_("The url is invalid in file %s\n"),sites);
 233                         exit(EXIT_FAILURE);
 234                 }
 235
 236                 twork1=nacc;
 237                 twork2=nbytes;
 238                 twork3=ntime;
 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                 fputs("</a></td><td class=\"data\"",fp_ou);
 253                 if (SortTableJs[0]) fprintf(fp_ou," sorttable_customkey=\"%"PRId64"\"",(uint64_t)twork1);
 254                 fprintf(fp_ou,">%s</td>",fixnum(twork1,1));
 255                 fputs("<td class=\"data\"",fp_ou);
 256                 if (SortTableJs[0]) fprintf(fp_ou," sorttable_customkey=\"%"PRId64"\"",(uint64_t)twork2);
 257                 fprintf(fp_ou,">%s</td>",fixnum(twork2,1));
 258                 fputs("<td class=\"data\"",fp_ou);
 259                 if (SortTableJs[0]) fprintf(fp_ou," sorttable_customkey=\"%"PRId64"\"",(uint64_t)twork3);
 260                 fprintf(fp_ou,">%s</td></tr>\n",fixtime(twork3));
 261         }
 262         fclose(fp_in);
 263         longline_destroy(&line);
 264
 265         fputs("</table></div>\n",fp_ou);
 266         if (write_html_trailer(fp_ou)<0)
 267                 debuga(_("Write error in file %s\n"),report);
 268         if (fclose(fp_ou)==EOF)
 269                 debuga(_("Failed to close file %s - %s\n"),report,strerror(errno));
 270
 271         return;
 272 }