topsites.c

   1 /*
   2  * SARG Squid Analysis Report Generator      http://sarg.sourceforge.net
   3  *                                                            1998, 2012
   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 #ifdef ENABLE_DOUBLE_CHECK_DATA
  31 extern struct globalstatstruct globstat;
  32 #endif
  33
  34 void topsites(void)
  35 {
  36         FILE *fp_in, *fp_ou;
  37
  38         char *buf;
  39         char *url;
  40         char *ourl=NULL;
  41         char csort[255];
  42         char general[MAXLEN];
  43         char general2[MAXLEN];
  44         char general3[MAXLEN];
  45         char sites[MAXLEN];
  46         char report[MAXLEN];
  47         const char *sortf;
  48         const char *sortt;
  49         long long int nacc;
  50         long long int nbytes;
  51         long long int ntime;
  52         long long int tnacc=0;
  53         long long int tnbytes=0;
  54         long long int tntime=0;
  55         long long int twork1=0, twork2=0, twork3=0;
  56 #ifdef ENABLE_DOUBLE_CHECK_DATA
  57         long long int ttnacc=0;
  58         long long int ttnbytes=0;
  59         long long int ttntime=0;
  60 #endif
  61         int regs=0;
  62         int cstatus;
  63         int url_len;
  64         int ourl_size=0;
  65         struct getwordstruct gwarea;
  66         longline line;
  67         struct generalitemstruct item;
  68
  69         if(Privacy) {
  70                 if (debugz) debugaz(_("Top sites report not produced because privacy option is on\n"));
  71                 return;
  72         }
  73
  74         sprintf(general,"%s/sarg-general",outdirname);
  75         sprintf(sites,"%s/sarg-sites",outdirname);
  76         sprintf(general2,"%s/sarg-general2",outdirname);
  77         sprintf(general3,"%s/sarg-general3",outdirname);
  78
  79         sprintf(report,"%s/topsites.html",outdirname);
  80
  81         if (snprintf(csort,sizeof(csort),"sort -t \"\t\" -k 4,4 -o \"%s\" \"%s\"",general2,general)>=sizeof(csort)) {
  82                 debuga(_("Sort command too long when sorting file \"%s\" to \"%s\"\n"),general,general2);
  83                 exit(EXIT_FAILURE);
  84         }
  85         cstatus=system(csort);
  86         if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
  87                 debuga(_("sort command return status %d\n"),WEXITSTATUS(cstatus));
  88                 debuga(_("sort command: %s\n"),csort);
  89                 exit(EXIT_FAILURE);
  90         }
  91
  92         if((fp_in=fopen(general2,"r"))==NULL) {
  93                 debuga(_("(topsites) Cannot open log file %s\n"),general2);
  94                 debuga(_("sort command: %s\n"),csort);
  95                 exit(EXIT_FAILURE);
  96         }
  97
  98         if((fp_ou=fopen(general3,"w"))==NULL) {
  99                 debuga(_("(topsites) Cannot open log file %s\n"),general3);
 100                 exit(EXIT_FAILURE);
 101         }
 102
 103         if ((line=longline_create())==NULL) {
 104                 debuga(_("Not enough memory to read file %s\n"),general2);
 105                 exit(EXIT_FAILURE);
 106         }
 107
 108         while((buf=longline_read(fp_in,line))!=NULL) {
 109                 ger_read(buf,&item,general2);
 110                 if(item.total) continue;
 111
 112                 if(!regs) {
 113                         url_len=strlen(item.url);
 114                         if (!ourl || url_len>=ourl_size) {
 115                                 ourl_size=url_len+1;
 116                                 ourl=realloc(ourl,ourl_size);
 117                                 if (!ourl) {
 118                                         debuga(_("Not enough memory to store the url\n"));
 119                                         exit(EXIT_FAILURE);
 120                                 }
 121                         }
 122                         strcpy(ourl,item.url);
 123                         regs++;
 124                 }
 125
 126                 if(strcmp(item.url,ourl) != 0) {
 127                         /*
 128                         This complicated printf is due to Microsoft's inability to comply with any standard. Msvcrt is unable
 129                         to print a long long int unless it is exactly 64-bits long.
 130                         */
 131                         fprintf(fp_ou,"%"PRIu64"\t%"PRIu64"\t%"PRIu64"\t%s\n",(uint64_t)tnacc,(uint64_t)tnbytes,(uint64_t)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\n"));
 138                                         exit(EXIT_FAILURE);
 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 #ifdef ENABLE_DOUBLE_CHECK_DATA
 151                 ttnacc+=item.nacc;
 152                 ttnbytes+=item.nbytes;
 153                 ttntime+=item.nelap;
 154 #endif
 155         }
 156         fclose(fp_in);
 157         longline_destroy(&line);
 158
 159         if (ourl) {
 160                 /*
 161                 This complicated printf is due to Microsoft's inability to comply with any standard. Msvcrt is unable
 162                 to print a long long int unless it is exactly 64-bits long.
 163                 */
 164                 fprintf(fp_ou,"%"PRIu64"\t%"PRIu64"\t%"PRIu64"\t%s\n",(uint64_t)tnacc,(uint64_t)tnbytes,(uint64_t)tntime,ourl);
 165                 free(ourl);
 166         }
 167
 168         fclose(fp_ou);
 169
 170 #ifdef ENABLE_DOUBLE_CHECK_DATA
 171         if (ttnacc!=globstat.nacc || ttnbytes!=globstat.nbytes || ttntime!=globstat.elap) {
 172                 debuga(_("Total statistics mismatch when reading %s to produce the top sites\n"),general2);
 173                 exit(EXIT_FAILURE);
 174         }
 175 #endif
 176
 177         if (!KeepTempLog && unlink(general2)) {
 178                 debuga(_("Cannot delete \"%s\": %s\n"),general2,strerror(errno));
 179                 exit(EXIT_FAILURE);
 180         }
 181
 182         if((TopsitesSort & TOPSITE_SORT_CONNECT) != 0) {
 183                 sortf="-k 1,1 -k 2,2";
 184         } else if((TopsitesSort & TOPSITE_SORT_BYTES) != 0) {
 185                 sortf="-k 2,2 -k 1,1";
 186         } else if((TopsitesSort & TOPSITE_SORT_TIME) != 0) {
 187                 sortf="-k 3,3";
 188         } else {
 189                 sortf="-k 2,2 -k 1,1"; //default is BYTES
 190         }
 191         if((TopsitesSort & TOPSITE_SORT_REVERSE) != 0) {
 192                 sortt="-r";
 193         } else {
 194                 sortt="";
 195         }
 196
 197         if (snprintf(csort,sizeof(csort),"sort -t \"\t\" %s -n %s -o \"%s\" \"%s\"",sortt,sortf,sites,general3)>=sizeof(csort)) {
 198                 debuga(_("Sort command too long when sorting file \"%s\" to \"%s\"\n"),general3,sites);
 199                 exit(EXIT_FAILURE);
 200         }
 201         cstatus=system(csort);
 202         if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 203                 debuga(_("sort command return status %d\n"),WEXITSTATUS(cstatus));
 204                 debuga(_("sort command: %s\n"),csort);
 205                 exit(EXIT_FAILURE);
 206         }
 207         if((fp_in=fopen(sites,"r"))==NULL) {
 208                 debuga(_("(topsites) Cannot open log file %s\n"),sites);
 209                 debuga(_("sort command: %s\n"),csort);
 210                 exit(EXIT_FAILURE);
 211         }
 212
 213         if (!KeepTempLog && unlink(general3)) {
 214                 debuga(_("Cannot delete \"%s\": %s\n"),general3,strerror(errno));
 215                 exit(EXIT_FAILURE);
 216         }
 217
 218         if((fp_ou=fopen(report,"w"))==NULL) {
 219                 debuga(_("(topsites) Cannot open log file %s\n"),report);
 220                 exit(EXIT_FAILURE);
 221         }
 222
 223         write_html_header(fp_ou,(IndexTree == INDEX_TREE_DATE) ? 3 : 1,_("Top sites"),HTML_JS_SORTTABLE);
 224         fputs("<tr><td class=\"header_c\">",fp_ou);
 225         fprintf(fp_ou,_("Period: %s"),period.html);
 226         fputs("</td></tr>\n",fp_ou);
 227         fputs("<tr><th class=\"header_c\">",fp_ou);
 228         fprintf(fp_ou,_("Top %d sites"),TopSitesNum);
 229         fputs("</th></tr>\n",fp_ou);
 230         close_html_header(fp_ou);
 231
 232         fputs("<div class=\"report\"><table cellpadding=\"1\" cellspacing=\"2\"",fp_ou);
 233         if (SortTableJs[0]) fputs(" class=\"sortable\"",fp_ou);
 234         fputs(">\n",fp_ou);
 235         fprintf(fp_ou,"<thead><tr><th class=\"header_l\">%s</th><th class=\"header_l",_("NUM"));
 236         if (SortTableJs[0]) fputs(" sorttable_alpha",fp_ou);
 237         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"));
 238
 239         regs=0;
 240         ntopsites = 0;
 241
 242         if ((line=longline_create())==NULL) {
 243                 debuga(_("Not enough memory to read file %s\n"),sites);
 244                 exit(EXIT_FAILURE);
 245         }
 246
 247         while(regs<TopSitesNum && (buf=longline_read(fp_in,line))!=NULL) {
 248                 getword_start(&gwarea,buf);
 249                 if (getword_atoll(&nacc,&gwarea,'\t')<0) {
 250                         debuga(_("Maybe you have a broken record or garbage in your %s file\n"),sites);
 251                         exit(EXIT_FAILURE);
 252                 }
 253                 if (nacc == 0) continue;
 254                 if (getword_atoll(&nbytes,&gwarea,'\t')<0 || getword_atoll(&ntime,&gwarea,'\t')<0) {
 255                         debuga(_("Maybe you have a broken record or garbage in your %s file\n"),sites);
 256                         exit(EXIT_FAILURE);
 257                 }
 258                 if (getword_ptr(buf,&url,&gwarea,'\t')<0) {
 259                         debuga(_("The url is invalid in file %s\n"),sites);
 260                         exit(EXIT_FAILURE);
 261                 }
 262
 263                 twork1=nacc;
 264                 twork2=nbytes;
 265                 twork3=ntime;
 266
 267                 fprintf(fp_ou,"<tr><td class=\"data\">%d</td><td class=\"data2\">",++regs);
 268
 269                 if(BlockIt[0] != '\0' && url[0]!=ALIAS_PREFIX) {
 270                         fprintf(fp_ou,"<a href=\"%s%s?url=\"",wwwDocumentRoot,BlockIt);
 271                         output_html_url(fp_ou,url);
 272                         fputs("\"><img src=\"../images/sarg-squidguard-block.png\"></a>&nbsp;",fp_ou);
 273                 }
 274
 275                 output_html_link(fp_ou,url,100);
 276                 fputs("</td><td class=\"data\"",fp_ou);
 277                 if (SortTableJs[0]) fprintf(fp_ou," sorttable_customkey=\"%"PRId64"\"",(uint64_t)twork1);
 278                 fprintf(fp_ou,">%s</td>",fixnum(twork1,1));
 279                 fputs("<td class=\"data\"",fp_ou);
 280                 if (SortTableJs[0]) fprintf(fp_ou," sorttable_customkey=\"%"PRId64"\"",(uint64_t)twork2);
 281                 fprintf(fp_ou,">%s</td>",fixnum(twork2,1));
 282                 fputs("<td class=\"data\"",fp_ou);
 283                 if (SortTableJs[0]) fprintf(fp_ou," sorttable_customkey=\"%"PRId64"\"",(uint64_t)twork3);
 284                 fprintf(fp_ou,">%s</td></tr>\n",fixtime(twork3));
 285         }
 286         fclose(fp_in);
 287         longline_destroy(&line);
 288
 289         fputs("</table></div>\n",fp_ou);
 290         if (write_html_trailer(fp_ou)<0)
 291                 debuga(_("Write error in file %s\n"),report);
 292         if (fclose(fp_ou)==EOF)
 293                 debuga(_("Failed to close file %s - %s\n"),report,strerror(errno));
 294
 295         return;
 296 }