Package: webalizer
Version: 2.23.05-1
Severity: wishlist
Tags: upstream
Hello,
Trying to group the google sites with webalizer is hell as its matching
algorithm is primitive. For example:
www.google.co.uk, www.google.com and www.google.ie cannot be catched by using
something like www.google.*
The attached (quick) patch allows perl regular expressions for advanced
matching and allows rules like:
GroupReferrer @www\.google\..* Google
GroupReferrer @[a-zA-Z\.]*freecode\.com/.* Freecode
regards,
Nikos
diff -ur webalizer-2.23-05/configure.in webalizer-2.23-05.new/configure.in
--- webalizer-2.23-05/configure.in 2011-01-03 05:53:44.000000000 +0100
+++ webalizer-2.23-05.new/configure.in 2012-12-12 19:50:55.604735560 +0100
@@ -119,6 +119,11 @@
AC_MSG_ERROR(z library not found.. please install libz)
fi
+AC_CHECK_LIB(pcre, main, PCRE="yes"; LIBS="-lpcre ${LIBS}", PCRE="no")
+if test "${PCRE}" = "no"; then
+ AC_MSG_ERROR(libpcre was not found.. please install it!)
+fi
+
AC_CHECK_HEADER(zlib.h, HDR="yes", HDR="no")
if test "${HDR}" = "no"; then
AC_MSG_ERROR(zlib.h header not found.. please install)
diff -ur webalizer-2.23-05/linklist.c webalizer-2.23-05.new/linklist.c
--- webalizer-2.23-05/linklist.c 2011-01-03 05:51:20.000000000 +0100
+++ webalizer-2.23-05.new/linklist.c 2012-12-12 20:22:42.652680042 +0100
@@ -32,6 +32,7 @@
#include <unistd.h> /* normal stuff */
#include <ctype.h>
#include <sys/utsname.h>
+#include <pcre.h>
/* ensure sys/types */
#ifndef _SYS_TYPES_H
@@ -52,6 +53,8 @@
#include "lang.h"
#include "linklist.h"
+#define OVECCOUNT 30
+
/* internal function prototypes */
NLISTPTR new_nlist(char *); /* new list node */
@@ -274,9 +277,38 @@
int isinstr(char *str, char *cp)
{
char *cp1,*cp2;
+ char *pattern;
+ const char* error;
+ int erroffset, rc;
+ int ovector[OVECCOUNT];
+
+ pcre *re;
cp1=(cp+strlen(cp))-1;
- if (*cp=='*')
+
+ if (*cp=='@') /* perl regex */
+ {
+ pattern = ++cp;
+
+ re = pcre_compile(pattern, 0, &error, &erroffset, NULL);
+ if (re == NULL)
+ {
+ fprintf(stderr, "Error in perl regex at offset %d: %s\n", erroffset, error);
+ return 0;
+ }
+
+ rc = pcre_exec(re, NULL, str, strlen(str), 0, 0, ovector, OVECCOUNT);
+ if (rc < 0)
+ {
+ pcre_free(re);
+ return 0;
+ }
+
+ /* fprintf(stderr, "\nMatch succeeded (%s l: %s) at offset %d\n", pattern, str, ovector[0]); */
+ pcre_free(re);
+ return 1;
+ }
+ else if (*cp=='*')
{
/* if leading wildcard, start from end */
cp2=str+strlen(str)-1;
diff -ur webalizer-2.23-05/sample.conf webalizer-2.23-05.new/sample.conf
--- webalizer-2.23-05/sample.conf 2011-01-03 05:53:33.000000000 +0100
+++ webalizer-2.23-05.new/sample.conf 2012-12-12 20:28:57.340669134 +0100
@@ -572,6 +572,9 @@
#GroupSite *.compuserve.com
#GroupReferrer yahoo.com/ Yahoo!
+# note that patterns starting with '@' are perl regular expressions
+#GroupReferrer @www\.google\..* Google
+#GroupReferrer www.bing.com/ Bing
#GroupReferrer excite.com/ Excite
#GroupReferrer infoseek.com/ InfoSeek
#GroupReferrer webcrawler.com/ WebCrawler
