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/sarg/squagent.unsort",TempDir);
  63    sprintf(tmp2,"%s/sarg/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
 105       if(gwarea.current[0]!='\0') {
 106          if (getword_skip(MAXLEN,&gwarea,' ')<0 || getword(user,sizeof(user),&gwarea,'\n')<0) {
 107             debuga(_("Maybe you have a broken record or garbage in your %s file\n"),UserAgentLog);
 108             exit(EXIT_FAILURE);
 109          }
 110          if(user[0] == '-')
 111             strcpy(user,ip);
 112          if(user[0] == '\0')
 113             strcpy(user,ip);
 114       } else {
 115          strcpy(user,ip);
 116       }
 117
 118       fprintf(fp_ou,"%s\t%s\t%s\n",ip,agent,user);
 119       useragent_count++;
 120    }
 121
 122    if(debug) {
 123       debuga(_("   Records read: %ld\n"),totregsl);
 124    }
 125
 126    fclose(fp_in);
 127    if (fclose(fp_ou)==EOF) {
 128       debuga(_("Failed to close file %s - %s\n"),tmp3,strerror(errno));
 129       exit(EXIT_FAILURE);
 130    }
 131
 132    if(debug) {
 133       debuga(_("Sorting file: %s\n"),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       debuga(_("sort command return status %d\n"),WEXITSTATUS(cstatus));
 140       debuga(_("sort command: %s\n"),csort);
 141       exit(EXIT_FAILURE);
 142    }
 143    if((fp_in=fopen(tmp2,"r"))==NULL) {
 144       debuga(_("(useragent) Cannot open file %s\n"),tmp2);
 145       debuga(_("sort command: %s\n"),csort);
 146       exit(EXIT_FAILURE);
 147    }
 148
 149    unlink(tmp3);
 150
 151    snprintf(hfile,sizeof(hfile),"%s/useragent.html", outdirname);
 152    if((fp_ht=fopen(hfile,"w"))==NULL) {
 153       debuga(_("(useragent) Cannot open file %s\n"),hfile);
 154       exit(EXIT_FAILURE);
 155    }
 156
 157    if(debug)
 158       debuga(_("Making Useragent report\n"));
 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",_("Squid Useragent's Report"));
 162    fprintf(fp_ht,"<tr><td class=\"header_c\">%s: %s - %s</td></tr>\n",_("Period"),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",_("USERID"),_("AGENT"));
 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          debuga(_("There is an invalid IP address in file %s\n"),tmp2);
 176          exit(EXIT_FAILURE);
 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          debuga(_("There is an invalid useragent in file %s\n"),tmp2);
 189          exit(EXIT_FAILURE);
 190       }
 191       if (getword(user,sizeof(user),&gwarea,'\t')<0) {
 192          debuga(_("There is an invalid user ID in file %s\n"),tmp2);
 193          exit(EXIT_FAILURE);
 194       }
 195
 196       if(strcmp(user,user_old) != 0) {
 197          fprintf(fp_ht,"<tr><td class=\"data2\">%s</td><td class=\"data2\">",user);
 198          output_html_string(fp_ht,agent,250);
 199          fputs("</td></tr>\n",fp_ht);
 200          strcpy(user_old,user);
 201          strcpy(agent_old,agent);
 202       } else if(strcmp(agent,agent_old) != 0) {
 203          fputs("<tr><td></td><td class=\"data2\">",fp_ht);
 204          output_html_string(fp_ht,agent,250);
 205          fputs("</td></tr>\n",fp_ht);
 206          strcpy(agent_old,agent);
 207       }
 208    }
 209
 210    fputs("</table>\n",fp_ht);
 211    fclose(fp_in);
 212
 213    sprintf(csort,"sort -t \"\t\" -k 2,2 -o \"%s\" \"%s\"",tmp3,tmp2);
 214    cstatus=system(csort);
 215    if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 216       debuga(_("sort command return status %d\n"),WEXITSTATUS(cstatus));
 217       debuga(_("sort command: %s\n"),csort);
 218       exit(EXIT_FAILURE);
 219    }
 220    if((fp_in=fopen(tmp3,"r"))==NULL) {
 221       debuga(_("(useragent) Cannot open file %s\n"),tmp3);
 222       debuga(_("sort command: %s\n"),csort);
 223       exit(EXIT_FAILURE);
 224    }
 225
 226    unlink(tmp2);
 227
 228    if((fp_ou=fopen(tmp2,"w"))==NULL) {
 229       debuga(_("(useragent) Cannot open file %s\n"),tmp2);
 230       exit(EXIT_FAILURE);
 231    }
 232
 233    agent_old[0]='\0';
 234    cont=0;
 235
 236    while(fgets(buf,sizeof(buf),fp_in)!=NULL) {
 237       getword_start(&gwarea,buf);
 238       if (getword(ip,sizeof(ip),&gwarea,'\t')<0) {
 239          debuga(_("There is an invalid IP address in file %s\n"),tmp3);
 240          exit(EXIT_FAILURE);
 241       }
 242       if (getword(agent,sizeof(agent),&gwarea,'\t')<0) {
 243          debuga(_("There is an invalid useragent in file %s\n"),tmp3);
 244          exit(EXIT_FAILURE);
 245       }
 246
 247       if(!cont) {
 248          cont++;
 249          strcpy(agent_old,agent);
 250       }
 251
 252       if(strcmp(agent,agent_old) != 0) {
 253          agentdif++;
 254          fprintf(fp_ou,"%06d %s\n",agentot,agent_old);
 255          strcpy(agent_old,agent);
 256          agentot2+=agentot;
 257          agentot=0;
 258       }
 259       agentot++;
 260    }
 261    agentdif++;
 262    fprintf(fp_ou,"%06d %s\n",agentot,agent);
 263    agentot2+=agentot;
 264
 265    fclose(fp_in);
 266    if (fclose(fp_ou)==EOF) {
 267       debuga(_("Failed to close file %s - %s\n"),tmp3,strerror(errno));
 268       exit(EXIT_FAILURE);
 269    }
 270
 271    unlink(tmp3);
 272
 273    sprintf(csort,"sort -n -r -k 1,1 -o \"%s\" \"%s\"",tmp3,tmp2);
 274    cstatus=system(csort);
 275    if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 276       debuga(_("sort command return status %d\n"),WEXITSTATUS(cstatus));
 277       debuga(_("sort command: %s\n"),csort);
 278       exit(EXIT_FAILURE);
 279    }
 280    if((fp_in=fopen(tmp3,"r"))==NULL) {
 281       debuga(_("(useragent) Cannot open file %s\n"),tmp3);
 282       debuga(_("sort command: %s\n"),csort);
 283       exit(EXIT_FAILURE);
 284    }
 285
 286    unlink(tmp2);
 287
 288    fputs("<br><br>\n",fp_ht);
 289
 290    fputs("<table cellpadding=\"0\" cellspacing=\"0\">\n",fp_ht);
 291    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"));
 292
 293    perc=0.;
 294    while(fgets(buf,sizeof(buf),fp_in)!=NULL) {
 295       fixendofline(buf);
 296       getword_start(&gwarea,buf);
 297       if (getword(tagent,sizeof(tagent),&gwarea,' ')<0) {
 298          debuga(_("There is an invalid useragent in file %s\n"),tmp3);
 299          exit(EXIT_FAILURE);
 300       }
 301       nagent=atoi(tagent);
 302       perc=(agentot2>0) ? nagent * 100. / agentot2 : 0.;
 303
 304       fputs("<tr><td class=\"data2\">",fp_ht);
 305       output_html_string(fp_ht,gwarea.current,250);
 306       fprintf(fp_ht,"</td><td class=\"data\">%d</td><td class=\"data\">%3.2lf</td></tr>\n",nagent,perc);
 307    }
 308    fclose(fp_in);
 309
 310    fputs("</table></div>\n",fp_ht);
 311    if (write_html_trailer(fp_ht)<0)
 312       debuga(_("Write error in file %s\n"),hfile);
 313    if (fclose(fp_ht)==EOF)
 314       debuga(_("Failed to close file %s - %s\n"),hfile,strerror(errno));
 315
 316    unlink(tmp3);
 317
 318    return;
 319
 320 }