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 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",TempDir);
  62    sprintf(tmp2,"%s/squagent.log",TempDir);
  63
  64    if((fp_in=fopen(UserAgentLog,"r"))==NULL) {
  65       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],UserAgentLog);
  66       exit(1);
  67    }
  68
  69    if((fp_ou=fopen(tmp3,"w"))==NULL) {
  70       fprintf(stderr, "SARG: (email) %s: %s\n",text[45],tmp3);
  71       exit(1);
  72    }
  73
  74    if(debug) {
  75       debuga("%s: %s",text[66],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          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",UserAgentLog);
  84          exit(1);
  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          printf("SARG: Maybe you have a broken date in your %s file.\n",UserAgentLog);
  90          exit(1);
  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          printf("SARG: Maybe you have a broken useragent entry in your %s file.\n",UserAgentLog);
 101          exit(1);
 102       }
 103       strcpy(warea,agent);
 104       strup(warea);
 105       if(strstr(warea,"SCRIPT") != 0 || strstr(warea,"ONLOAD") != 0)
 106          baddata();
 107
 108       if(gwarea.current[0]!='\0') {
 109          if (getword_skip(MAXLEN,&gwarea,' ')<0 || getword(user,sizeof(user),&gwarea,'\n')<0) {
 110             printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",UserAgentLog);
 111             exit(1);
 112          }
 113       }
 114
 115       if(user[0] == '-')
 116          strcpy(user,ip);
 117       if(user[0] == '\0')
 118          strcpy(user,ip);
 119
 120       fprintf(fp_ou,"%s\t%s\t%s\n",ip,agent,user);
 121       user[0]='\0';
 122       useragent_count++;
 123    }
 124
 125    if(debug) {
 126       debuga("   %s: %ld",text[10],totregsl);
 127    }
 128
 129    fclose(fp_in);
 130    fclose(fp_ou);
 131
 132    if(debug) {
 133       debuga("%s: %s",text[54],tmp2);
 134    }
 135
 136    sprintf(csort,"sort -n -t \"\t\" -k 3,3 -k 2,2 -k 1,1 -o \"%s\" \"%s\"",tmp2,tmp3);
 137    cstatus=system(csort);
 138    if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 139       fprintf(stderr, "SARG: sort command return status %d\n",WEXITSTATUS(cstatus));
 140       fprintf(stderr, "SARG: sort command: %s\n",csort);
 141       exit(1);
 142    }
 143    if((fp_in=fopen(tmp2,"r"))==NULL) {
 144       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],tmp2);
 145       fprintf(stderr, "SARG: sort command: %s\n",csort);
 146       exit(1);
 147    }
 148
 149    unlink(tmp3);
 150
 151    sprintf(hfile,"%s/useragent.html", dirname);
 152    if((fp_ht=fopen(hfile,"w"))==NULL) {
 153       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],hfile);
 154       exit(1);
 155    }
 156
 157    if(debug)
 158       debuga("%s",text[72]);
 159
 160    write_html_header(fp_ht,(IndexTree == INDEX_TREE_DATE) ? 3 : 1,_("Squid Useragent's Report"));
 161    fprintf(fp_ht,"<tr><th class=\"header_c\">%s</th></tr>\n",text[105]);
 162    fprintf(fp_ht,"<tr><td class=\"header_c\">%s: %s - %s</td></tr>\n",text[89],idate,fdate);
 163    close_html_header(fp_ht);
 164
 165    fputs("<br><br>\n",fp_ht);
 166
 167    fputs("<div class=\"report\"><table cellpadding=\"0\" cellspacing=\"0\">\n",fp_ht);
 168    fputs("<tr><td>&nbsp;</td><td>&nbsp;</td></tr>",fp_ht);
 169
 170    fprintf(fp_ht,"<tr><th class=\"header_l\">%s</th><th class=\"header_l\">%s</th></tr>\n",text[98],text[106]);
 171
 172    while(fgets(buf,sizeof(buf),fp_in)!=NULL) {
 173       getword_start(&gwarea,buf);
 174       if (getword(ip,sizeof(ip),&gwarea,'\t')<0) {
 175          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",tmp2);
 176          exit(1);
 177       }
 178
 179       if(Ip2Name) {
 180          if(strcmp(ip,ipbefore) != 0) {
 181             strcpy(ipbefore,ip);
 182             ip2name(ip,sizeof(ip));
 183             strcpy(namebefore,ip);
 184          } else strcpy(ip,namebefore);
 185       }
 186
 187       if (getword(agent,sizeof(agent),&gwarea,'\t')<0 ||
 188           getword(user,sizeof(user),&gwarea,'\t')<0) {
 189          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",tmp2);
 190          exit(1);
 191       }
 192
 193       if(strcmp(user,user_old) != 0) {
 194          fprintf(fp_ht,"<tr><td class=\"data2\">%s</td><td class=\"data2\">%s</td></tr>\n",user,agent);
 195          strcpy(user_old,user);
 196          strcpy(agent_old,agent);
 197       } else {
 198          if(strcmp(agent,agent_old) != 0) {
 199             fprintf(fp_ht,"<tr><td></td><td class=\"data2\">%s</td></tr>\n",agent);
 200             strcpy(agent_old,agent);
 201          }
 202       }
 203    }
 204
 205    fputs("</table>\n",fp_ht);
 206    fclose(fp_in);
 207
 208    sprintf(csort,"sort -t \"\t\" -k 2,2 -o \"%s\" \"%s\"",tmp3,tmp2);
 209    cstatus=system(csort);
 210    if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 211       fprintf(stderr, "SARG: sort command return status %d\n",WEXITSTATUS(cstatus));
 212       fprintf(stderr, "SARG: sort command: %s\n",csort);
 213       exit(1);
 214    }
 215    if((fp_in=fopen(tmp3,"r"))==NULL) {
 216       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],tmp3);
 217       fprintf(stderr, "SARG: sort command: %s\n",csort);
 218       exit(1);
 219    }
 220
 221    unlink(tmp2);
 222
 223    if((fp_ou=fopen(tmp2,"w"))==NULL) {
 224       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],tmp2);
 225       exit(1);
 226    }
 227
 228    agent_old[0]='\0';
 229    cont=0;
 230
 231    while(fgets(buf,sizeof(buf),fp_in)!=NULL) {
 232       getword_start(&gwarea,buf);
 233       if (getword(ip,sizeof(ip),&gwarea,'\t')<0 ||
 234           getword(agent,sizeof(agent),&gwarea,'\t')<0) {
 235          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",tmp);
 236          exit(1);
 237       }
 238
 239       if(!cont) {
 240          cont++;
 241          strcpy(agent_old,agent);
 242       }
 243
 244       if(strcmp(agent,agent_old) != 0) {
 245          agentdif++;
 246          fprintf(fp_ou,"%06d %s\n",agentot,agent_old);
 247          strcpy(agent_old,agent);
 248          agentot2+=agentot;
 249          agentot=0;
 250       }
 251       agentot++;
 252    }
 253    agentdif++;
 254    fprintf(fp_ou,"%06d %s\n",agentot,agent);
 255    agentot2+=agentot;
 256
 257    fclose(fp_in);
 258    fclose(fp_ou);
 259
 260    unlink(tmp3);
 261
 262    sprintf(csort,"sort -n -r -k 1,1 -o \"%s\" \"%s\"",tmp3,tmp2);
 263    cstatus=system(csort);
 264    if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 265       fprintf(stderr, "SARG: sort command return status %d\n",WEXITSTATUS(cstatus));
 266       fprintf(stderr, "SARG: sort command: %s\n",csort);
 267       exit(1);
 268    }
 269    if((fp_in=fopen(tmp3,"r"))==NULL) {
 270       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],tmp3);
 271       fprintf(stderr, "SARG: sort command: %s\n",csort);
 272       exit(1);
 273    }
 274
 275    unlink(tmp2);
 276
 277    fputs("<br><br>\n",fp_ht);
 278
 279    fputs("<table cellpadding=\"0\" cellspacing=\"0\">\n",fp_ht);
 280    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]);
 281
 282    perc=0.;
 283    while(fgets(buf,sizeof(buf),fp_in)!=NULL) {
 284       fixendofline(buf);
 285       getword_start(&gwarea,buf);
 286       if (getword(tagent,sizeof(tagent),&gwarea,' ')<0) {
 287          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",tmp);
 288          exit(1);
 289       }
 290       nagent=atoi(tagent);
 291       perc=(agentot2>0) ? nagent * 100. / agentot2 : 0.;
 292
 293       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);
 294    }
 295    fclose(fp_in);
 296
 297    fputs("</table></div>\n",fp_ht);
 298    write_html_trailer(fp_ht);
 299    fclose(fp_ht);
 300
 301    unlink(tmp3);
 302
 303    return;
 304
 305 }