useragent.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 useragent(void)
  31 {
  32         FILE *fp_in = NULL, *fp_ou = NULL, *fp_ht = NULL;
  33         char buf[MAXLEN];
  34         char ip[MAXLEN], data[MAXLEN], agent[MAXLEN], user[MAXLEN];
  35         char ipbefore[MAXLEN], namebefore[MAXLEN];
  36         char tagent[MAXLEN];
  37         char user_old[MAXLEN]="$#%0a3bc6";
  38         char agent_old[MAXLEN]="$#%0a3bc6";
  39         char hfile[MAXLEN];
  40         char idate[MAXLEN], fdate[MAXLEN];
  41         char tmp2[MAXLEN];
  42         char tmp3[MAXLEN];
  43         char day[4],month[5],year[5], wdate[20];
  44         char csort[MAXLEN];
  45         int  agentot=0, agentot2=0, agentdif=0, cont=0, nagent;
  46         unsigned long totregsl=0;
  47         int cstatus;
  48         int ndate;
  49         double perc;
  50         struct getwordstruct gwarea, gwarea1;
  51
  52         ip[0]='\0';
  53         data[0]='\0';
  54         agent[0]='\0';
  55         user[0]='\0';
  56         user_old[0]='\0';
  57         agent_old[0]='\0';
  58         ipbefore[0]='\0';
  59         namebefore[0]='\0';
  60
  61         sprintf(tmp3,"%s/squagent.unsort",tmp);
  62         sprintf(tmp2,"%s/squagent.log",tmp);
  63
  64         if((fp_in=fopen(UserAgentLog,"r"))==NULL) {
  65                 debuga(_("(useragent) Cannot open file %s\n"),UserAgentLog);
  66                 exit(EXIT_FAILURE);
  67         }
  68
  69         if((fp_ou=fopen(tmp3,"w"))==NULL) {
  70                 debuga(_("(useragent) Cannot open file %s\n"),tmp3);
  71                 exit(EXIT_FAILURE);
  72         }
  73
  74         if(debug) {
  75                 debuga(_("Reading useragent log: %s\n"),UserAgentLog);
  76         }
  77
  78         while(fgets(buf,sizeof(buf),fp_in)!=NULL) {
  79                 totregsl++;
  80                 getword_start(&gwarea,buf);
  81                 if (getword(ip,sizeof(ip),&gwarea,' ')<0 || getword_skip(MAXLEN,&gwarea,'[')<0 ||
  82                     getword(data,sizeof(data),&gwarea,' ')<0) {
  83                         debuga(_("Maybe you have a broken record or garbage in your %s file\n"),UserAgentLog);
  84                         exit(EXIT_FAILURE);
  85                 }
  86                 getword_start(&gwarea1,data);
  87                 if (getword(day,sizeof(day),&gwarea1,'/')<0 || getword(month,sizeof(month),&gwarea1,'/')<0 ||
  88                     getword(year,sizeof(year),&gwarea1,':')<0) {
  89                         debuga(_("Maybe you have a broken date in your %s file\n"),UserAgentLog);
  90                         exit(EXIT_FAILURE);
  91                 }
  92                 buildymd(day,month,year,wdate);
  93                 ndate=atoi(wdate);
  94                 if (ndate<dfrom) continue;
  95                 if (ndate>duntil) break;
  96                 if(totregsl == 1)
  97                         strcpy(idate,data);
  98                 strcpy(fdate,data);
  99                 if (getword_skip(MAXLEN,&gwarea,'"')<0 || getword(agent,sizeof(agent),&gwarea,'"')<0) {
 100                         debuga(_("Maybe you have a broken useragent entry in your %s file\n"),UserAgentLog);
 101                         exit(EXIT_FAILURE);
 102                 }
 103
 104                 if(gwarea.current[0]!='\0') {
 105                         if (getword_skip(MAXLEN,&gwarea,' ')<0 || getword(user,sizeof(user),&gwarea,'\n')<0) {
 106                                 debuga(_("Maybe you have a broken record or garbage in your %s file\n"),UserAgentLog);
 107                                 exit(EXIT_FAILURE);
 108                         }
 109                         if(user[0] == '-')
 110                                 strcpy(user,ip);
 111                         if(user[0] == '\0')
 112                                 strcpy(user,ip);
 113                 } else {
 114                         strcpy(user,ip);
 115                 }
 116
 117                 fprintf(fp_ou,"%s\t%s\t%s\n",ip,agent,user);
 118                 useragent_count++;
 119         }
 120
 121         if(debug) {
 122                 debuga(_("   Records read: %ld\n"),totregsl);
 123         }
 124
 125         fclose(fp_in);
 126         if (fclose(fp_ou)==EOF) {
 127                 debuga(_("Failed to close file %s - %s\n"),tmp3,strerror(errno));
 128                 exit(EXIT_FAILURE);
 129         }
 130
 131         if(debug) {
 132                 debuga(_("Sorting file: %s\n"),tmp2);
 133         }
 134
 135         sprintf(csort,"sort -n -t \"\t\" -k 3,3 -k 2,2 -k 1,1 -o \"%s\" \"%s\"",tmp2,tmp3);
 136         cstatus=system(csort);
 137         if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 138                 debuga(_("sort command return status %d\n"),WEXITSTATUS(cstatus));
 139                 debuga(_("sort command: %s\n"),csort);
 140                 exit(EXIT_FAILURE);
 141         }
 142         if((fp_in=fopen(tmp2,"r"))==NULL) {
 143                 debuga(_("(useragent) Cannot open file %s\n"),tmp2);
 144                 debuga(_("sort command: %s\n"),csort);
 145                 exit(EXIT_FAILURE);
 146         }
 147
 148         unlink(tmp3);
 149
 150         snprintf(hfile,sizeof(hfile),"%s/useragent.html", outdirname);
 151         if((fp_ht=fopen(hfile,"w"))==NULL) {
 152                 debuga(_("(useragent) Cannot open file %s\n"),hfile);
 153                 exit(EXIT_FAILURE);
 154         }
 155
 156         if(debug)
 157                 debuga(_("Making Useragent report\n"));
 158
 159         write_html_header(fp_ht,(IndexTree == INDEX_TREE_DATE) ? 3 : 1,_("Squid Useragent's Report"),HTML_JS_NONE);
 160         fprintf(fp_ht,"<tr><th class=\"header_c\">%s</th></tr>\n",_("Squid Useragent's Report"));
 161         fprintf(fp_ht,"<tr><td class=\"header_c\">%s: %s - %s</td></tr>\n",_("Period"),idate,fdate);
 162         close_html_header(fp_ht);
 163
 164         fputs("<br><br>\n",fp_ht);
 165
 166         fputs("<div class=\"report\"><table cellpadding=\"0\" cellspacing=\"0\">\n",fp_ht);
 167         fputs("<tr><td>&nbsp;</td><td>&nbsp;</td></tr>",fp_ht);
 168
 169         fprintf(fp_ht,"<tr><th class=\"header_l\">%s</th><th class=\"header_l\">%s</th></tr>\n",_("USERID"),_("AGENT"));
 170
 171         while(fgets(buf,sizeof(buf),fp_in)!=NULL) {
 172                 getword_start(&gwarea,buf);
 173                 if (getword(ip,sizeof(ip),&gwarea,'\t')<0) {
 174                         debuga(_("There is an invalid IP address in file %s\n"),tmp2);
 175                         exit(EXIT_FAILURE);
 176                 }
 177
 178                 if(Ip2Name) {
 179                         if(strcmp(ip,ipbefore) != 0) {
 180                                 strcpy(ipbefore,ip);
 181                                 ip2name(ip,sizeof(ip));
 182                                 strcpy(namebefore,ip);
 183                         } else strcpy(ip,namebefore);
 184                 }
 185
 186                 if (getword(agent,sizeof(agent),&gwarea,'\t')<0) {
 187                         debuga(_("There is an invalid useragent in file %s\n"),tmp2);
 188                         exit(EXIT_FAILURE);
 189                 }
 190                 if (getword(user,sizeof(user),&gwarea,'\t')<0) {
 191                         debuga(_("There is an invalid user ID in file %s\n"),tmp2);
 192                         exit(EXIT_FAILURE);
 193                 }
 194
 195                 if(strcmp(user,user_old) != 0) {
 196                         fprintf(fp_ht,"<tr><td class=\"data2\">%s</td><td class=\"data2\">",user);
 197                         output_html_string(fp_ht,agent,250);
 198                         fputs("</td></tr>\n",fp_ht);
 199                         strcpy(user_old,user);
 200                         strcpy(agent_old,agent);
 201                 } else if(strcmp(agent,agent_old) != 0) {
 202                         fputs("<tr><td></td><td class=\"data2\">",fp_ht);
 203                         output_html_string(fp_ht,agent,250);
 204                         fputs("</td></tr>\n",fp_ht);
 205                         strcpy(agent_old,agent);
 206                 }
 207         }
 208
 209         fputs("</table>\n",fp_ht);
 210         fclose(fp_in);
 211
 212         sprintf(csort,"sort -t \"\t\" -k 2,2 -o \"%s\" \"%s\"",tmp3,tmp2);
 213         cstatus=system(csort);
 214         if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 215                 debuga(_("sort command return status %d\n"),WEXITSTATUS(cstatus));
 216                 debuga(_("sort command: %s\n"),csort);
 217                 exit(EXIT_FAILURE);
 218         }
 219         if((fp_in=fopen(tmp3,"r"))==NULL) {
 220                 debuga(_("(useragent) Cannot open file %s\n"),tmp3);
 221                 debuga(_("sort command: %s\n"),csort);
 222                 exit(EXIT_FAILURE);
 223         }
 224
 225         unlink(tmp2);
 226
 227         if((fp_ou=fopen(tmp2,"w"))==NULL) {
 228                 debuga(_("(useragent) Cannot open file %s\n"),tmp2);
 229                 exit(EXIT_FAILURE);
 230         }
 231
 232         agent_old[0]='\0';
 233         cont=0;
 234
 235         while(fgets(buf,sizeof(buf),fp_in)!=NULL) {
 236                 getword_start(&gwarea,buf);
 237                 if (getword(ip,sizeof(ip),&gwarea,'\t')<0) {
 238                         debuga(_("There is an invalid IP address in file %s\n"),tmp3);
 239                         exit(EXIT_FAILURE);
 240                 }
 241                 if (getword(agent,sizeof(agent),&gwarea,'\t')<0) {
 242                         debuga(_("There is an invalid useragent in file %s\n"),tmp3);
 243                         exit(EXIT_FAILURE);
 244                 }
 245
 246                 if(!cont) {
 247                         cont++;
 248                         strcpy(agent_old,agent);
 249                 }
 250
 251                 if(strcmp(agent,agent_old) != 0) {
 252                         agentdif++;
 253                         fprintf(fp_ou,"%06d %s\n",agentot,agent_old);
 254                         strcpy(agent_old,agent);
 255                         agentot2+=agentot;
 256                         agentot=0;
 257                 }
 258                 agentot++;
 259         }
 260         agentdif++;
 261         fprintf(fp_ou,"%06d %s\n",agentot,agent);
 262         agentot2+=agentot;
 263
 264         fclose(fp_in);
 265         if (fclose(fp_ou)==EOF) {
 266                 debuga(_("Failed to close file %s - %s\n"),tmp3,strerror(errno));
 267                 exit(EXIT_FAILURE);
 268         }
 269
 270         unlink(tmp3);
 271
 272         sprintf(csort,"sort -n -r -k 1,1 -o \"%s\" \"%s\"",tmp3,tmp2);
 273         cstatus=system(csort);
 274         if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 275                 debuga(_("sort command return status %d\n"),WEXITSTATUS(cstatus));
 276                 debuga(_("sort command: %s\n"),csort);
 277                 exit(EXIT_FAILURE);
 278         }
 279         if((fp_in=fopen(tmp3,"r"))==NULL) {
 280                 debuga(_("(useragent) Cannot open file %s\n"),tmp3);
 281                 debuga(_("sort command: %s\n"),csort);
 282                 exit(EXIT_FAILURE);
 283         }
 284
 285         unlink(tmp2);
 286
 287         fputs("<br><br>\n",fp_ht);
 288
 289         fputs("<table cellpadding=\"0\" cellspacing=\"0\">\n",fp_ht);
 290         fprintf(fp_ht,"<tr><th class=\"header_l\">%s</th><th class=\"header_l\">%s</th><th class=\"header_c\">%%</th></tr>\n",_("AGENT"),_("TOTAL"));
 291
 292         perc=0.;
 293         while(fgets(buf,sizeof(buf),fp_in)!=NULL) {
 294                 fixendofline(buf);
 295                 getword_start(&gwarea,buf);
 296                 if (getword(tagent,sizeof(tagent),&gwarea,' ')<0) {
 297                         debuga(_("There is an invalid useragent in file %s\n"),tmp3);
 298                         exit(EXIT_FAILURE);
 299                 }
 300                 nagent=atoi(tagent);
 301                 perc=(agentot2>0) ? nagent * 100. / agentot2 : 0.;
 302
 303                 fputs("<tr><td class=\"data2\">",fp_ht);
 304                 output_html_string(fp_ht,gwarea.current,250);
 305                 fprintf(fp_ht,"</td><td class=\"data\">%d</td><td class=\"data\">%3.2lf</td></tr>\n",nagent,perc);
 306         }
 307         fclose(fp_in);
 308
 309         fputs("</table></div>\n",fp_ht);
 310         if (write_html_trailer(fp_ht)<0)
 311                 debuga(_("Write error in file %s\n"),hfile);
 312         if (fclose(fp_ht)==EOF)
 313                 debuga(_("Failed to close file %s - %s\n"),hfile,strerror(errno));
 314
 315         unlink(tmp3);
 316
 317         return;
 318 }