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 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 day[4],month[5],year[5], wdate[20];
  43    int  agentot=0, agentot2=0, agentdif=0, cont=0, nagent;
  44    unsigned long totregsl=0;
  45    int cstatus;
  46    int ndate;
  47    double perc;
  48    struct getwordstruct gwarea, gwarea1;
  49
  50    ip[0]='\0';
  51    data[0]='\0';
  52    agent[0]='\0';
  53    user[0]='\0';
  54    user_old[0]='\0';
  55    agent_old[0]='\0';
  56    ipbefore[0]='\0';
  57    namebefore[0]='\0';
  58
  59    sprintf(tmp3,"%s/squagent.unsort",TempDir);
  60    sprintf(tmp2,"%s/squagent.log",TempDir);
  61
  62    if((fp_in=fopen(UserAgentLog,"r"))==NULL) {
  63       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],UserAgentLog);
  64       exit(1);
  65    }
  66
  67    if((fp_ou=fopen(tmp3,"w"))==NULL) {
  68       fprintf(stderr, "SARG: (email) %s: %s\n",text[45],tmp3);
  69       exit(1);
  70    }
  71
  72    if(debug) {
  73       debuga("%s: %s",text[66],UserAgentLog);
  74    }
  75
  76    while(fgets(buf,sizeof(buf),fp_in)!=NULL) {
  77       totregsl++;
  78       getword_start(&gwarea,buf);
  79       if (getword(ip,sizeof(ip),&gwarea,' ')<0 || getword_skip(MAXLEN,&gwarea,'[')<0 ||
  80           getword(data,sizeof(data),&gwarea,' ')<0) {
  81          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",UserAgentLog);
  82          exit(1);
  83       }
  84       getword_start(&gwarea1,data);
  85       if (getword(day,sizeof(day),&gwarea1,'/')<0 || getword(month,sizeof(month),&gwarea1,'/')<0 ||
  86           getword(year,sizeof(year),&gwarea1,':')<0) {
  87          printf("SARG: Maybe you have a broken date in your %s file.\n",UserAgentLog);
  88          exit(1);
  89       }
  90       buildymd(day,month,year,wdate);
  91       ndate=atoi(wdate);
  92       if (ndate<dfrom) continue;
  93       if (ndate>duntil) break;
  94       if(totregsl == 1)
  95          strcpy(idate,data);
  96       strcpy(fdate,data);
  97       if (getword_skip(MAXLEN,&gwarea,'"')<0 || getword(agent,sizeof(agent),&gwarea,'"')<0) {
  98          printf("SARG: Maybe you have a broken useragent entry in your %s file.\n",UserAgentLog);
  99          exit(1);
 100       }
 101       strcpy(warea,agent);
 102       strup(warea);
 103       if(strstr(warea,"SCRIPT") != 0 || strstr(warea,"ONLOAD") != 0)
 104          baddata();
 105
 106       if(gwarea.current[0]!='\0') {
 107          if (getword_skip(MAXLEN,&gwarea,' ')<0 || getword(user,sizeof(user),&gwarea,'\n')<0) {
 108             printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",UserAgentLog);
 109             exit(1);
 110          }
 111       }
 112
 113       if(user[0] == '-')
 114          strcpy(user,ip);
 115       if(user[0] == '\0')
 116          strcpy(user,ip);
 117
 118       fprintf(fp_ou,"%s\t%s\t%s\n",ip,agent,user);
 119       user[0]='\0';
 120       useragent_count++;
 121    }
 122
 123    if(debug) {
 124       debuga("   %s: %ld",text[10],totregsl);
 125    }
 126
 127    fclose(fp_in);
 128    fclose(fp_ou);
 129
 130    if(debug) {
 131       debuga("%s: %s",text[54],tmp2);
 132    }
 133
 134    sprintf(csort,"sort -n -t \"\t\" -k 3,3 -k 2,2 -k 1,1 -o \"%s\" \"%s\"",tmp2,tmp3);
 135    cstatus=system(csort);
 136    if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 137       fprintf(stderr, "SARG: sort command return status %d\n",WEXITSTATUS(cstatus));
 138       fprintf(stderr, "SARG: sort command: %s\n",csort);
 139       exit(1);
 140    }
 141    if((fp_in=fopen(tmp2,"r"))==NULL) {
 142       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],tmp2);
 143       fprintf(stderr, "SARG: sort command: %s\n",csort);
 144       exit(1);
 145    }
 146
 147    unlink(tmp3);
 148
 149    sprintf(hfile,"%s/%s/useragent.html", outdir,period);
 150    if((fp_ht=fopen(hfile,"w"))==NULL) {
 151       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],hfile);
 152       exit(1);
 153    }
 154
 155    if(debug)
 156       debuga("%s",text[72]);
 157
 158    write_html_header(fp_ht,(IndexTree == INDEX_TREE_DATE) ? 3 : 1,_("Squid Useragent's Report"));
 159    fprintf(fp_ht,"<tr><th class=\"header_c\">%s</th></tr>\n",text[105]);
 160    fprintf(fp_ht,"<tr><td class=\"header_c\">%s: %s - %s</td></tr>\n",text[89],idate,fdate);
 161    close_html_header(fp_ht);
 162
 163    fputs("<br><br>\n",fp_ht);
 164
 165    fputs("<div class=\"report\"><table cellpadding=\"0\" cellspacing=\"0\">\n",fp_ht);
 166    fputs("<tr><td>&nbsp;</td><td>&nbsp;</td></tr>",fp_ht);
 167
 168    fprintf(fp_ht,"<tr><th class=\"header_l\">%s</th><th class=\"header_l\">%s</th></tr>\n",text[98],text[106]);
 169
 170    while(fgets(buf,sizeof(buf),fp_in)!=NULL) {
 171       getword_start(&gwarea,buf);
 172       if (getword(ip,sizeof(ip),&gwarea,'\t')<0) {
 173          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",tmp2);
 174          exit(1);
 175       }
 176
 177       if(Ip2Name) {
 178          if(strcmp(ip,ipbefore) != 0) {
 179             strcpy(ipbefore,ip);
 180             ip2name(ip,sizeof(ip));
 181             strcpy(namebefore,ip);
 182          } else strcpy(ip,namebefore);
 183       }
 184
 185       if (getword(agent,sizeof(agent),&gwarea,'\t')<0 ||
 186           getword(user,sizeof(user),&gwarea,'\t')<0) {
 187          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",tmp2);
 188          exit(1);
 189       }
 190
 191       if(strcmp(user,user_old) != 0) {
 192          fprintf(fp_ht,"<tr><td class=\"data2\">%s</td><td class=\"data2\">%s</td></tr>\n",user,agent);
 193          strcpy(user_old,user);
 194          strcpy(agent_old,agent);
 195       } else {
 196          if(strcmp(agent,agent_old) != 0) {
 197             fprintf(fp_ht,"<tr><td></td><td class=\"data2\">%s</td></tr>\n",agent);
 198             strcpy(agent_old,agent);
 199          }
 200       }
 201    }
 202
 203    fputs("</table>\n",fp_ht);
 204    fclose(fp_in);
 205
 206    sprintf(csort,"sort -t \"\t\" -k 2,2 -o \"%s\" \"%s\"",tmp3,tmp2);
 207    cstatus=system(csort);
 208    if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 209       fprintf(stderr, "SARG: sort command return status %d\n",WEXITSTATUS(cstatus));
 210       fprintf(stderr, "SARG: sort command: %s\n",csort);
 211       exit(1);
 212    }
 213    if((fp_in=fopen(tmp3,"r"))==NULL) {
 214       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],tmp3);
 215       fprintf(stderr, "SARG: sort command: %s\n",csort);
 216       exit(1);
 217    }
 218
 219    unlink(tmp2);
 220
 221    if((fp_ou=fopen(tmp2,"w"))==NULL) {
 222       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],tmp2);
 223       exit(1);
 224    }
 225
 226    agent_old[0]='\0';
 227    cont=0;
 228
 229    while(fgets(buf,sizeof(buf),fp_in)!=NULL) {
 230       getword_start(&gwarea,buf);
 231       if (getword(ip,sizeof(ip),&gwarea,'\t')<0 ||
 232           getword(agent,sizeof(agent),&gwarea,'\t')<0) {
 233          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",tmp);
 234          exit(1);
 235       }
 236
 237       if(!cont) {
 238          cont++;
 239          strcpy(agent_old,agent);
 240       }
 241
 242       if(strcmp(agent,agent_old) != 0) {
 243          agentdif++;
 244          fprintf(fp_ou,"%06d %s\n",agentot,agent_old);
 245          strcpy(agent_old,agent);
 246          agentot2+=agentot;
 247          agentot=0;
 248       }
 249       agentot++;
 250    }
 251    agentdif++;
 252    fprintf(fp_ou,"%06d %s\n",agentot,agent);
 253    agentot2+=agentot;
 254
 255    fclose(fp_in);
 256    fclose(fp_ou);
 257
 258    unlink(tmp3);
 259
 260    sprintf(csort,"sort -n -r -k 1,1 -o \"%s\" \"%s\"",tmp3,tmp2);
 261    cstatus=system(csort);
 262    if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 263       fprintf(stderr, "SARG: sort command return status %d\n",WEXITSTATUS(cstatus));
 264       fprintf(stderr, "SARG: sort command: %s\n",csort);
 265       exit(1);
 266    }
 267    if((fp_in=fopen(tmp3,"r"))==NULL) {
 268       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],tmp3);
 269       fprintf(stderr, "SARG: sort command: %s\n",csort);
 270       exit(1);
 271    }
 272
 273    unlink(tmp2);
 274
 275    fputs("<br><br>\n",fp_ht);
 276
 277    fputs("<table cellpadding=\"0\" cellspacing=\"0\">\n",fp_ht);
 278    fprintf(fp_ht,"<tr><th class=\"header_l\">%s</th><th class=\"header_l\">%s</th><th class=\"header_c\">%%</th></tr>\n",text[106],text[107]);
 279
 280    perc=0.;
 281    while(fgets(buf,sizeof(buf),fp_in)!=NULL) {
 282       fixendofline(buf);
 283       getword_start(&gwarea,buf);
 284       if (getword(tagent,sizeof(tagent),&gwarea,' ')<0) {
 285          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",tmp);
 286          exit(1);
 287       }
 288       nagent=atoi(tagent);
 289       perc=(agentot2>0) ? nagent * 100. / agentot2 : 0.;
 290
 291       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);
 292    }
 293    fclose(fp_in);
 294
 295    fputs("</table></div>\n",fp_ht);
 296    write_html_trailer(fp_ht);
 297    fclose(fp_ht);
 298
 299    unlink(tmp3);
 300
 301    return;
 302
 303 }