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    int  agentot=0, agentot2=0, agentdif=0, cont=0, nagent;
  43    unsigned long totregsl=0;
  44    int cstatus;
  45    char ftime[128];
  46    struct getwordstruct gwarea;
  47
  48    ip[0]='\0';
  49    data[0]='\0';
  50    agent[0]='\0';
  51    user[0]='\0';
  52    user_old[0]='\0';
  53    agent_old[0]='\0';
  54    ipbefore[0]='\0';
  55    namebefore[0]='\0';
  56
  57    sprintf(tmp3,"%s/squagent.unsort",TempDir);
  58    sprintf(tmp2,"%s/squagent.log",TempDir);
  59
  60    if((fp_in=fopen(UserAgentLog,"r"))==NULL) {
  61       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],UserAgentLog);
  62       exit(1);
  63    }
  64
  65    if((fp_ou=fopen(tmp3,"w"))==NULL) {
  66       fprintf(stderr, "SARG: (email) %s: %s\n",text[45],tmp3);
  67       exit(1);
  68    }
  69
  70    if(debug) {
  71       debuga("%s: %s",text[66],UserAgentLog);
  72    }
  73
  74    while(fgets(buf,sizeof(buf),fp_in)!=NULL) {
  75       totregsl++;
  76       getword_start(&gwarea,buf);
  77       if (getword(ip,sizeof(ip),&gwarea,' ')<0 || getword_skip(MAXLEN,&gwarea,'[')<0 ||
  78           getword(data,sizeof(data),&gwarea,' ')<0) {
  79          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",UserAgentLog);
  80          exit(1);
  81       }
  82       if(totregsl == 1)
  83          strcpy(idate,data);
  84       if (getword_skip(MAXLEN,&gwarea,'"')<0 || getword(agent,sizeof(agent),&gwarea,'"')<0) {
  85          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",UserAgentLog);
  86          exit(1);
  87       }
  88       strcpy(warea,agent);
  89       strup(warea);
  90       if(strstr(warea,"SCRIPT") != 0 || strstr(warea,"ONLOAD") != 0)
  91          baddata();
  92
  93       if(gwarea.current[0]!='\0') {
  94          if (getword_skip(MAXLEN,&gwarea,' ')<0 || getword(user,sizeof(user),&gwarea,'\n')<0) {
  95             printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",UserAgentLog);
  96             exit(1);
  97          }
  98       }
  99
 100       if(user[0] == '-')
 101          strcpy(user,ip);
 102       if(user[0] == '\0')
 103          strcpy(user,ip);
 104
 105       fprintf(fp_ou,"%s\t%s\t%s\n",ip,agent,user);
 106       user[0]='\0';
 107    }
 108
 109    strcpy(fdate,data);
 110
 111    if(debug) {
 112       debuga("   %s: %ld",text[10],totregsl);
 113    }
 114
 115    fclose(fp_in);
 116    fclose(fp_ou);
 117
 118    if(debug) {
 119       debuga("%s: %s",text[54],tmp2);
 120    }
 121
 122    sprintf(csort,"sort -n -t \"\t\" -k 3,3 -k 2,2 -k 1,1 -o \"%s\" \"%s\"",tmp2,tmp3);
 123    cstatus=system(csort);
 124    if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 125       fprintf(stderr, "SARG: sort command return status %d\n",WEXITSTATUS(cstatus));
 126       fprintf(stderr, "SARG: sort command: %s\n",csort);
 127       exit(1);
 128    }
 129    if((fp_in=fopen(tmp2,"r"))==NULL) {
 130       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],tmp2);
 131       fprintf(stderr, "SARG: sort command: %s\n",csort);
 132       exit(1);
 133    }
 134
 135    unlink(tmp3);
 136
 137    sprintf(hfile,"%s/%s/useragent.html", outdir,period);
 138    if((fp_ht=fopen(hfile,"w"))==NULL) {
 139       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],hfile);
 140       exit(1);
 141    }
 142
 143    if(debug)
 144       debuga("%s",text[72]);
 145
 146    fprintf(fp_ht, "<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 4.01 Transitional//EN\" \"http://www.w3.org/TR/html4/loose.dtd\">\n<html>\n<head>\n  <meta http-equiv=\"Content-Type\" content=\"text/html; charset=%s\">\n",CharSet);
 147    css(fp_ht);
 148    fputs("</head>\n",fp_ht);
 149    //if(FontFace[0] != '\0') fprintf(fp_ht,"<font face=\"%s\">\n",FontFace);
 150    fprintf(fp_ht,"<body bgcolor=\"%s\" text=\"%s\" background=\"%s\">\n",BgColor,TxColor,BgImage);
 151    write_logo_image(fp_ht);
 152
 153    if(IndexTree == INDEX_TREE_DATE)
 154       show_sarg(fp_ht, "../../..");
 155    else
 156       show_sarg(fp_ht,"..");
 157    fputs("<div align=\"center\"><table cellpadding=\"0\" cellspacing=\"0\">\n",fp_ht);
 158    fprintf(fp_ht,"<tr><th align=\"center\" colspan=\"2\"><b><font color=\"%s\" size=\"+1\">%s</font></b></th></tr>\n",TiColor,text[105]);
 159
 160    fputs("<tr><td></td></tr><tr><td></td></tr>",fp_ht);
 161    fputs("</table></div>\n",fp_ht);
 162
 163    fputs("<div align=\"center\"><table cellpadding=\"0\" cellspacing=\"0\">\n",fp_ht);
 164
 165    fprintf(fp_ht,"<tr><td align=\"right\" bgcolor=\"%s\">%s:<td align=\"left\" bgcolor=\"%s\">%s - %s</td></td></tr>\n",HeaderBgColor,text[89],TxBgColor,idate,fdate);
 166
 167    fputs("</table></div>\n",fp_ht);
 168    fputs("<div align=\"center\"><table cellpadding=\"0\" cellspacing=\"0\">\n",fp_ht);
 169    fputs("<tr><td></td><td></td></tr>",fp_ht);
 170
 171    fprintf(fp_ou,"<tr><th align=\"left\" bgcolor=\"%s\"><font color=\"%s\">%s</font></th><th bgcolor=\"%s\" align=\"left\"><font color=\"%s\">%s</font></th></tr>\n",HeaderBgColor,HeaderColor,text[98],HeaderBgColor,HeaderColor,text[106]);
 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          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",tmp2);
 177          exit(1);
 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           getword(user,sizeof(user),&gwarea,'\t')<0) {
 190          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",tmp2);
 191          exit(1);
 192       }
 193
 194       if(strcmp(user,user_old) != 0) {
 195          fprintf(fp_ht,"<tr><td align=\"left\" bgcolor=\"%s\">%s</td><td align=\"left\" bgcolor=\"%s\">%s</td></tr>\n",TxBgColor,user,TxBgColor,agent);
 196          strcpy(user_old,user);
 197          strcpy(agent_old,agent);
 198       } else {
 199          if(strcmp(agent,agent_old) != 0) {
 200             fprintf(fp_ht,"<tr><td></td><td align=\"left\" bgcolor=\"%s\">%s</td></tr>\n",TxBgColor,agent);
 201             strcpy(agent_old,agent);
 202          }
 203       }
 204    }
 205
 206    fputs("</table>\n",fp_ht);
 207    fclose(fp_in);
 208    fclose(fp_ou);
 209
 210    sprintf(csort,"sort -t \"\t\" -k 2,2 -o \"%s\" \"%s\"",tmp3,tmp2);
 211    cstatus=system(csort);
 212    if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 213       fprintf(stderr, "SARG: sort command return status %d\n",WEXITSTATUS(cstatus));
 214       fprintf(stderr, "SARG: sort command: %s\n",csort);
 215       exit(1);
 216    }
 217    if((fp_in=fopen(tmp3,"r"))==NULL) {
 218       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],tmp3);
 219       fprintf(stderr, "SARG: sort command: %s\n",csort);
 220       exit(1);
 221    }
 222
 223    unlink(tmp2);
 224
 225    if((fp_ou=fopen(tmp2,"w"))==NULL) {
 226       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],tmp2);
 227       exit(1);
 228    }
 229
 230    agent_old[0]='\0';
 231    cont=0;
 232
 233    while(fgets(buf,sizeof(buf),fp_in)!=NULL) {
 234       getword_start(&gwarea,buf);
 235       if (getword(ip,sizeof(ip),&gwarea,'\t')<0 ||
 236           getword(agent,sizeof(agent),&gwarea,'\t')<0) {
 237          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",tmp);
 238          exit(1);
 239       }
 240
 241       if(!cont) {
 242          cont++;
 243          strcpy(agent_old,agent);
 244       }
 245
 246       if(strcmp(agent,agent_old) != 0) {
 247          agentdif++;
 248          fprintf(fp_ou,"%06d %s\n",agentot,agent_old);
 249          strcpy(agent_old,agent);
 250          agentot2+=agentot;
 251          agentot=0;
 252       }
 253       agentot++;
 254    }
 255    agentdif++;
 256    fprintf(fp_ou,"%06d %s\n",agentot,agent);
 257    agentot2+=agentot;
 258
 259    fclose(fp_in);
 260    fclose(fp_ou);
 261
 262    unlink(tmp3);
 263
 264    sprintf(csort,"sort -n -r -k 1,1 -o \"%s\" \"%s\"",tmp3,tmp2);
 265    cstatus=system(csort);
 266    if (!WIFEXITED(cstatus) || WEXITSTATUS(cstatus)) {
 267       fprintf(stderr, "SARG: sort command return status %d\n",WEXITSTATUS(cstatus));
 268       fprintf(stderr, "SARG: sort command: %s\n",csort);
 269       exit(1);
 270    }
 271    if((fp_in=fopen(tmp3,"r"))==NULL) {
 272       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],tmp3);
 273       fprintf(stderr, "SARG: sort command: %s\n",csort);
 274       exit(1);
 275    }
 276
 277    unlink(tmp2);
 278
 279    if((fp_ht=fopen(hfile,"a"))==NULL) {
 280       fprintf(stderr, "SARG: (useragent) %s: %s\n",text[45],hfile);
 281       exit(1);
 282    }
 283
 284    fputs("<br><br>\n",fp_ht);
 285
 286    fputs("<div align=\"center\"><table cellpadding=\"0\" cellspacing=\"0\">\n",fp_ht);
 287    fprintf(fp_ht,"<tr><th align=\"left\" bgcolor=\"%s\"><font color=\"%s\">%s</font></th><th bgcolor=\"%s\" align=\"left\"><font color=\"%s\">%s</font></th><th align=\"center\" bgcolor=\"%s\"><font color=\"%s\">%%</font></th></tr>\n",HeaderBgColor,HeaderColor,text[106],HeaderBgColor,HeaderColor,text[107],HeaderBgColor,HeaderColor);
 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          printf("SARG: Maybe you have a broken record or garbage in your %s file.\n",tmp);
 295          exit(1);
 296       }
 297       nagent=atoi(tagent);
 298       if(agentot2>0) {
 299          perc=nagent * 100;
 300          perc=perc / agentot2;
 301       }
 302
 303       fprintf(fp_ht,"<tr><td align=\"left\" bgcolor=\"%s\">%s</td><td align=\"right\" bgcolor=\"%s\">%d</td><td align=\"right\" bgcolor=\"%s\">%3.2f</td></tr>\n",TxBgColor,gwarea.current,TxBgColor,nagent,TxBgColor,perc);
 304    }
 305
 306    fputs("</table></html>\n",fp_ht);
 307    if(ShowSargInfo) {
 308       zdate(ftime, sizeof(ftime), DateFormat);
 309       fprintf(fp_ht,"<br><br><div align=\"center\"><font size=\"-2\">%s <a href=\"%s\">%s-%s</a> %s %s</font></div>\n",text[108],URL,PGM,VERSION,text[109],ftime);
 310    }
 311
 312    fputs("</table>\n</body>\n</html>\n",fp_ht);
 313    fclose(fp_in);
 314    fclose(fp_ht);
 315
 316    unlink(tmp3);
 317
 318    return;
 319
 320 }