readlog_common.c

   1 /*
   2  * SARG Squid Analysis Report Generator      http://sarg.sourceforge.net
   3  *                                                            1998, 2015
   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 #include "include/readlog.h"
  30
  31 /*!
  32 A new file is being read. The name of the file is \a FileName.
  33 */
  34 static void Common_NewFile(const char *FileName)
  35 {
  36 }
  37
  38 /*!
  39 Extract a column containing a long long int from \a Line.
  40
  41 The extracted value is stored in \a Value.
  42
  43 The pointer to the next byte just after the number is returned
  44 by the function.
  45 */
  46 static char *Common_GetLongLongInt(char *Line,long long int *Value)
  47 {
  48         *Value=0LL;
  49         if (*Line=='-') {
  50                 ++Line;
  51         } else {
  52                 while (isdigit(*Line)) *Value=*Value*10+(*Line++-'0');
  53         }
  54         return(Line);
  55 }
  56
  57 /*!
  58 Read one entry from a standard squid log format.
  59
  60 \param Line One line from the input log file.
  61 \param Entry Where to store the information parsed from the line.
  62
  63 \retval RLRC_NoError One valid entry is parsed.
  64 \retval RLRC_Unknown The line is invalid.
  65 \retval RLRC_InternalError An internal error was encountered.
  66 */
  67 static enum ReadLogReturnCodeEnum Common_ReadEntry(char *Line,struct ReadLogStruct *Entry)
  68 {
  69         const char *Begin;
  70         int IpLen;
  71         int HttpCodeLen;
  72         int UrlLen;
  73         int UserLen;
  74         int Day;
  75         char MonthName[4];
  76         int MonthNameLen;
  77         int Month;
  78         int Year;
  79         int Hour;
  80         int Minute;
  81         int Second;
  82         char *Ip;
  83
  84         // get IP address
  85         Entry->Ip=Ip=Line;
  86         for (IpLen=0 ; *Line && *Line!=' ' ; IpLen++) Line++;
  87         if (*Line!=' ' || IpLen==0) return(RLRC_Unknown);
  88
  89         if (!squid24) {
  90                 // squid version <= 2.4 store the user ID in the second column: skip the first column here
  91                 Begin=++Line;
  92                 while (*Line && *Line!=' ') Line++;
  93                 if (*Line!=' '|| Line==Begin) return(RLRC_Unknown);
  94         }
  95
  96         // the ID of the user or - if the user is unidentified
  97         Entry->User=++Line;
  98         for (UserLen=0 ; *Line && *Line!=' ' ; UserLen++) Line++;
  99         if (*Line!=' ' || UserLen==0) return(RLRC_Unknown);
 100
 101         if (squid24) {
 102                 // squid version > 2.4 store the user ID in the first column: skip the second column here
 103                 Begin=++Line;
 104                 while (*Line && *Line!=' ') Line++;
 105                 if (*Line!=' '|| Line==Begin) return(RLRC_Unknown);
 106         }
 107
 108         // get the date enclosed within square brackets
 109         ++Line;
 110         if (*Line!='[') return(RLRC_Unknown);
 111         ++Line;
 112         Day=0;
 113         while (isdigit(*Line)) Day=Day*10+(*Line++-'0');
 114         if (*Line!='/' || Day<1 || Day>31) return(RLRC_Unknown);
 115
 116         ++Line;
 117         for (MonthNameLen=0 ; MonthNameLen<sizeof(MonthName)-1 && isalpha(*Line) ; MonthNameLen++) MonthName[MonthNameLen]=*Line++;
 118         if (*Line!='/') return(RLRC_Unknown);
 119         MonthName[MonthNameLen]='\0';
 120         Month=month2num(MonthName);
 121         if (Month>=12) return(RLRC_Unknown);
 122
 123         ++Line;
 124         Year=0;
 125         while (isdigit(*Line)) Year=Year*10+(*Line++-'0');
 126         if (*Line!=':' || Year<1900 || Year>2200) return(RLRC_Unknown);
 127
 128         // get the time
 129         ++Line;
 130         Hour=0;
 131         while (isdigit(*Line)) Hour=Hour*10+(*Line++-'0');
 132         if (*Line!=':' || Hour>=24) return(RLRC_Unknown);
 133         ++Line;
 134         Minute=0;
 135         while (isdigit(*Line)) Minute=Minute*10+(*Line++-'0');
 136         if (*Line!=':' || Minute>=60) return(RLRC_Unknown);
 137         ++Line;
 138         Second=0;
 139         while (isdigit(*Line)) Second=Second*10+(*Line++-'0');
 140         if (*Line!=' ' || Second>60) return(RLRC_Unknown); //second can be 60 due to a leap second
 141
 142         // skip the timezone up to the closing ]
 143         while (*Line && *Line!=']') Line++;
 144         if (*Line!=']') return(RLRC_Unknown);
 145
 146         Entry->EntryTime.tm_year=Year-1900;
 147         Entry->EntryTime.tm_mon=Month;
 148         Entry->EntryTime.tm_mday=Day;
 149         Entry->EntryTime.tm_hour=Hour;
 150         Entry->EntryTime.tm_min=Minute;
 151         Entry->EntryTime.tm_sec=Second;
 152         Entry->EntryTime.tm_isdst=-1;
 153
 154         // the URL is enclosed between double qhotes
 155         ++Line;
 156         if (*Line!=' ') return(RLRC_Unknown);
 157         ++Line;
 158         if (*Line!='\"') return(RLRC_Unknown);
 159
 160         // skip the HTTP function
 161         Begin=++Line;
 162         while (isalpha(*Line)) Line++;
 163         if (*Line!=' ' || Line==Begin) return(RLRC_Unknown);
 164
 165         // get the URL
 166         Entry->Url=++Line;
 167         for (UrlLen=0 ; *Line && *Line!=' ' ; UrlLen++) Line++;
 168         if (*Line!=' ' || UrlLen==0) return(RLRC_Unknown);
 169
 170         // skip the HTTP/...
 171         ++Line;
 172         while (*Line && *Line!='\"') Line++;
 173         if (*Line!='\"') return(RLRC_Unknown);
 174         ++Line;
 175         if (*Line!=' ') return(RLRC_Unknown);
 176
 177         // get the HTTP code.
 178         Entry->HttpCode=++Line;
 179         for (HttpCodeLen=0 ; *Line && *Line!=' ' ; HttpCodeLen++) Line++;
 180         if (*Line!=' ' || HttpCodeLen==0) return(RLRC_Unknown);
 181
 182         // get the number of transfered bytes.
 183         Begin=++Line;
 184         Line=Common_GetLongLongInt(Line,&Entry->DataSize);
 185         // some log contains more columns
 186         if ((*Line && *Line!=' ') || Begin==Line) return(RLRC_Unknown);
 187
 188         // check the entry time
 189         if (mktime(&Entry->EntryTime)==-1) {
 190                 debuga(__FILE__,__LINE__,_("Invalid date or time found in the common log file\n"));
 191                 return(RLRC_InternalError);
 192         }
 193
 194         // it is safe to alter the line buffer now that we are returning a valid entry
 195         Ip[IpLen]='\0';
 196         Entry->HttpCode[HttpCodeLen]='\0';
 197         Entry->Url[UrlLen]='\0';
 198         Entry->User[UserLen]='\0';
 199
 200         return(RLRC_NoError);
 201 }
 202
 203 //! \brief Object to read a standard common log format.
 204 const struct ReadLogProcessStruct ReadCommonLog=
 205 {
 206         /* TRANSLATORS: This is the name of the log format displayed when this format is detected in an input log file. */
 207         N_("common log format"),
 208         Common_NewFile,
 209         Common_ReadEntry
 210 };