download.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 /*@null@*/static char *DownloadSuffix=NULL;
  31 /*@null@*/static char **DownloadSuffixIndex=NULL;
  32 static int NDownloadSuffix=0;
  33
  34 void download_report(void)
  35 {
  36
  37    FILE *fp_in = NULL, *fp_ou = NULL;
  38
  39    char *buf;
  40    char *url;
  41    char report_in[MAXLEN];
  42    char report[MAXLEN];
  43    char ip[MAXLEN];
  44    char oip[MAXLEN];
  45    char user[MAXLEN];
  46    char ouser[MAXLEN];
  47    char ouser2[MAXLEN];
  48    char data[15];
  49    char hora[15];
  50    int  z=0;
  51    int  count=0;
  52    int i;
  53    int day,month,year;
  54    bool new_user;
  55    struct getwordstruct gwarea;
  56    longline line;
  57    struct userinfostruct *uinfo;
  58    struct tm t;
  59
  60    ouser[0]='\0';
  61    ouser2[0]='\0';
  62
  63    snprintf(report_in,sizeof(report_in),"%s/download.log",tmp);
  64    if(access(report_in, R_OK) != 0)
  65       return;
  66
  67    snprintf(report,sizeof(report),"%s/download.html",outdirname);
  68
  69    if((fp_in=MY_FOPEN(report_in,"r"))==NULL) {
  70       debuga(_("(download) Cannot open log file %s\n"),report_in);
  71       exit(EXIT_FAILURE);
  72    }
  73
  74    if((fp_ou=MY_FOPEN(report,"w"))==NULL) {
  75       debuga(_("(download) Cannot open log file %s\n"),report);
  76       exit(EXIT_FAILURE);
  77    }
  78
  79    write_html_header(fp_ou,(IndexTree == INDEX_TREE_DATE) ? 3 : 1,_("Downloads"),HTML_JS_NONE);
  80    fputs("<tr><td class=\"header_c\">",fp_ou);
  81    fprintf(fp_ou,_("Period: %s"),period.html);
  82    fputs("</td></tr>\n",fp_ou);
  83    fprintf(fp_ou,"<tr><th class=\"header_c\">%s</th></tr>\n",_("Downloads"));
  84    close_html_header(fp_ou);
  85
  86    fputs("<div class=\"report\"><table cellpadding=\"0\" cellspacing=\"2\">\n",fp_ou);
  87    fprintf(fp_ou,"<tr><th class=\"header_l\">%s</th><th class=\"header_l\">%s</th><th class=\"header_l\">%s</th><th class=\"header_l\">%s</th></tr>\n",_("USERID"),_("IP/NAME"),_("DATE/TIME"),_("ACCESSED SITE"));
  88
  89    if ((line=longline_create())==NULL) {
  90       debuga(_("Not enough memory to read the downloaded files\n"));
  91       exit(EXIT_FAILURE);
  92    }
  93
  94    while((buf=longline_read(fp_in,line))!=NULL) {
  95       getword_start(&gwarea,buf);
  96       if (getword(data,sizeof(data),&gwarea,'\t')<0 || getword(hora,sizeof(hora),&gwarea,'\t')<0 ||
  97           getword(user,sizeof(user),&gwarea,'\t')<0 || getword(ip,sizeof(ip),&gwarea,'\t')<0) {
  98          debuga(_("There is a broken record or garbage in file %s\n"),report_in);
  99          exit(EXIT_FAILURE);
 100       }
 101       if (getword_ptr(buf,&url,&gwarea,'\t')<0) {
 102          debuga(_("There is a broken url in file %s\n"),report_in);
 103          exit(EXIT_FAILURE);
 104       }
 105       if (sscanf(data,"%d/%d/%d",&day,&month,&year)!=3) continue;
 106       computedate(year,month,day,&t);
 107       strftime(data,sizeof(data),"%x",&t);
 108
 109       uinfo=userinfo_find_from_id(user);
 110       if (!uinfo) {
 111          debuga(_("Unknown user ID %s in file %s\n"),user,report_in);
 112          exit(EXIT_FAILURE);
 113       }
 114       new_user=false;
 115       if(!z) {
 116          strcpy(ouser,user);
 117          strcpy(oip,ip);
 118          z++;
 119          new_user=true;
 120       } else {
 121          if(strcmp(ouser,user) != 0) {
 122             strcpy(ouser,user);
 123             new_user=true;
 124          }
 125          if(strcmp(oip,ip) != 0) {
 126             strcpy(oip,ip);
 127             new_user=true;
 128          }
 129       }
 130
 131       if(DownloadReportLimit) {
 132          if(strcmp(ouser2,uinfo->label) == 0) {
 133             count++;
 134          } else {
 135             count=1;
 136             strcpy(ouser2,uinfo->label);
 137          }
 138          if(count >= DownloadReportLimit)
 139             continue;
 140       }
 141
 142       for (i=strlen(url)-1 ; i>=0 && (unsigned char)url[i]<' ' ; i--) url[i]=0;
 143
 144       fputs("<tr>",fp_ou);
 145       if (new_user)
 146          fprintf(fp_ou,"<td class=\"data\">%s</td><td class=\"data\">%s</td>",uinfo->label,ip);
 147       else
 148          fputs("<td class=\"data\"></td><td class=\"data\"></td>",fp_ou);
 149       fprintf(fp_ou,"<td class=\"data\">%s-%s</td><td class=\"data2\">",data,hora);
 150       if(BlockIt[0]!='\0') {
 151          fprintf(fp_ou,"<a href=\"%s%s?url=\"",wwwDocumentRoot,BlockIt);
 152          output_html_url(fp_ou,url);
 153          fprintf(fp_ou,"\"><img src=\"%s/sarg-squidguard-block.png\"></a>&nbsp;",ImageFile);
 154       }
 155       fputs("<a href=\"http://",fp_ou);
 156       output_html_url(fp_ou,url);
 157       fputs("\">http://",fp_ou);
 158       output_html_string(fp_ou,url,100);
 159       fputs("</a></td></tr>\n",fp_ou);
 160    }
 161    fclose(fp_in);
 162    longline_destroy(&line);
 163
 164    fputs("</table></div>\n",fp_ou);
 165    if (write_html_trailer(fp_ou)<0)
 166       debuga(_("Write error in file %s\n"),report);
 167    if (fclose(fp_ou)==EOF)
 168       debuga(_("Failed to close file %s - %s\n"),report,strerror(errno));
 169
 170    unlink(report_in);
 171
 172    return;
 173 }
 174
 175 void free_download(void)
 176 {
 177    if (DownloadSuffix) {
 178       free(DownloadSuffix);
 179       DownloadSuffix=NULL;
 180    }
 181    if (DownloadSuffixIndex) {
 182       free(DownloadSuffixIndex);
 183       DownloadSuffixIndex=NULL;
 184    }
 185    NDownloadSuffix=0;
 186 }
 187
 188 void set_download_suffix(const char *list)
 189 {
 190    char *str;
 191    int i, j, k;
 192    int cmp;
 193
 194    free_download();
 195
 196    DownloadSuffix=strdup(list);
 197    if (!DownloadSuffix) {
 198       debuga(_("Download suffix list too long\n"));
 199       exit(EXIT_FAILURE);
 200    }
 201    j = 1;
 202    for (i=0 ; list[i] ; i++)
 203       if (list[i] == ',') j++;
 204    DownloadSuffixIndex=malloc(j*sizeof(char *));
 205    if (!DownloadSuffixIndex) {
 206       debuga(_("Too many download suffixes\n"));
 207       exit(EXIT_FAILURE);
 208    }
 209
 210    str = DownloadSuffix;
 211    for (i=0 ; DownloadSuffix[i] ; i++) {
 212       if (DownloadSuffix[i] == ',') {
 213          DownloadSuffix[i] = '\0';
 214          if (*str) {
 215             cmp = -1;
 216             for (j=0 ; j<NDownloadSuffix && (cmp=strcasecmp(str,DownloadSuffixIndex[j]))>0 ; j++);
 217             if (cmp != 0) {
 218                for (k=NDownloadSuffix ; k>j ; k--)
 219                   DownloadSuffixIndex[k]=DownloadSuffixIndex[k-1];
 220                NDownloadSuffix++;
 221                DownloadSuffixIndex[j]=str;
 222             }
 223          }
 224          str=DownloadSuffix+i+1;
 225       }
 226    }
 227
 228    if (*str) {
 229       cmp = -1;
 230       for (j=0 ; j<NDownloadSuffix && (cmp=strcasecmp(str,DownloadSuffixIndex[j]))>0 ; j++);
 231       if (cmp != 0) {
 232          for (k=NDownloadSuffix ; k>j ; k--)
 233             DownloadSuffixIndex[k]=DownloadSuffixIndex[k-1];
 234          NDownloadSuffix++;
 235          DownloadSuffixIndex[j]=str;
 236       }
 237    }
 238 }
 239
 240 bool is_download_suffix(const char *url)
 241 {
 242    int urllen;
 243    int i;
 244    int down, up, center;
 245    const char *suffix;
 246    int cmp;
 247    const int max_suffix=10;
 248
 249    if (DownloadSuffix == NULL || NDownloadSuffix == 0) return(false);
 250
 251    urllen=strlen(url)-1;
 252    if (urllen<=0) return(false);
 253    if (url[urllen] == '.') return(false); //reject a single trailing dot
 254    for (i=0 ; i<urllen && (url[i]!='/' || url[i+1]=='/') && url[i]!='?' ; i++);
 255    if (i>=urllen) return(false); // url is a hostname without any path or file to download
 256
 257    for (i=0 ; i<=max_suffix && i<urllen && url[urllen-i]!='.' ; i++)
 258       if (url[urllen-i] == '/' || url[urllen-i] == '?') return(false);
 259    if (i>max_suffix || i>=urllen) return(false);
 260
 261    suffix=url+urllen-i+1;
 262    down=0;
 263    up=NDownloadSuffix-1;
 264    while (down<=up) {
 265       center=(down+up)/2;
 266       cmp=strcasecmp(suffix,DownloadSuffixIndex[center]);
 267       if (cmp == 0) return(true);
 268       if (cmp < 0)
 269          up = center-1;
 270       else
 271          down = center+1;
 272    }
 273    return(false);
 274 }
 275