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