From: eldy <>
Date: Thu, 10 Oct 2002 21:09:04 +0000 (+0000)
Subject: Added URLWithQueryWithoutFollowingParameters parameter to exclude some parameters... 
X-Git-Tag: AWSTATS_5_1_BETA~42
X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=bcff6f868a619b2187e31f9389b23dc8d985d553;p=thirdparty%2FAWStats.git

Added URLWithQueryWithoutFollowingParameters parameter to exclude some parameters from URL when URLWithQuery is on.
---

diff --git a/docs/awstats_changelog.txt b/docs/awstats_changelog.txt
index 163dbdd1..cd082d02 100644
--- a/docs/awstats_changelog.txt
+++ b/docs/awstats_changelog.txt
@@ -1,11 +1,15 @@
 AWStats Changelog
 -----------------
 
+
 5.1
 
 - New feature: ShowxxxStats parmaters accept codes to decide which columns to
   show in chart.
-- New parameter: Added SkipUserAgents parameter.
+- New parameter: Added SkipUserAgents parameter to exclude some user agent
+  from statistics.
+- New parameter: Added URLWithQueryWithoutFollowingParameters to exclude some
+  parameters from URL when URLWithQuery is on.
 - Fixed: Fixed pb in refreshing page when clicking on "Update Now" link (no
   need to force the refresh).
 - Fixed: Pb with SAMBAR server ('Expires' line appears at the top of pages).
diff --git a/docs/awstats_config.html b/docs/awstats_config.html
index d99b5a81..4b210fd8 100644
--- a/docs/awstats_config.html
+++ b/docs/awstats_config.html
@@ -81,6 +81,7 @@ when reading it), follow the example:<br>
 <li><a href="#NotPageList">NotPageList</a>
 <li><a href="#ValidHTTPCodes">ValidHTTPCodes</a>
 <li><a href="#URLWithQuery">URLWithQuery</a>
+<li><a href="#URLWithQueryWithoutFollowingParameters">URLWithQueryWithoutFollowingParameters</a>
 <li><a href="#WarningMessages">WarningMessages</a>
 <li><a href="#NbOfLinesForCorruptedLog">NbOfLinesForCorruptedLog</a>
 <li><a href="#SplitSearchString">SplitSearchString</a>
@@ -408,8 +409,8 @@ when reading it), follow the example:<br>
 <br># same time to prevent unknown conflicts problems and avoid DoS attacks when
 <br># AllowToUpdateStatsFromBrowser is set to 1.
 <br># Because, when you use lock file, you can experience sometimes problems in
-<br># lock file not correctly removed (when process is killed for example that
-<br># require that you remove the file manualy), this option is not enabled by
+<br># lock file not correctly removed (when process is killed for example, this
+<br># requires that you remove the file manualy), this option is not enabled by
 <br># default (Do not enable this option with no consol server access).
 <br># Possible values: 0 or 1
 <br># Default: 0
@@ -572,7 +573,7 @@ when reading it), follow the example:<br>
 <br># ask AWStats to save last non corrupted file in a .bak file. This file is
 <br># stored in "DirData" directory with other 'historic files'.
 <br># Possible values: 0 or 1
-<br># Default: 1
+<br># Default: 0
 <br>#
 <br>KeepBackupOfHistoricFiles=0
 
@@ -745,16 +746,36 @@ when reading it), follow the example:<br>
 <br># different pages.
 <br># Warning, when set to 1, memory required to run AWStats is dramatically
 <br># increased if you have a lot of changing URLs (for example URLs with a random
-<br># id inside). Such web sites should not set this option to 1.
+<br># id inside). Such web sites should not set this option to 1 or use seriously
+<br># the next parameter URLWithQueryWithoutFollowingParameters.
 <br># Possible values:
 <br># 0 - URLs are cleaned from the query string (ie: "/mypage.html")
-<br># 1 - Full URL with query string is used     (ie: "/mypage.html?x=y")
+<br># 1 - Full URL with query string is used     (ie: "/mypage.html?p=x&q=y")
 <br># Default: 0
 <br># 
 <br>URLWithQuery=0
 
 <br><br><hr>
 
+<a name="URLWithQueryWithoutFollowingParameters"><b>URLWithqueryWithoutFollowingParameters</b></a><br>
+<b>Version : </b>5.1 +<br>
+<br># When URLWithQuery is on, you will get the full URL with all parameters in
+<br># URL reports. But among thoose parameters, sometimes you don't need a
+<br># particular parameter because it does not identify the page or because it's
+<br># a random ID changing for each access even if URL points to same page. In
+<br># such cases, it is higly recommanded to ask AWStats to remove such parameters
+<br># from the URL before counting, manipulating and storing it. Enter here list
+<br># of all non wanted parameters. For example if you enter "id", one hit on
+<br># /mypage.cgi?p=abc&id=Yo4UomP9d and /mypage.cgi?p=abc&id=Mu8fdxl3r
+<br># will be reported as 2 hits on /mypage.cgi?p=abc
+<br># This parameter is not used when URLWithQuery is 0.
+<br># Example: "PHPSESSID"
+<br># Default: ""
+<br># 
+<br>URLWithQueryWithoutFollowingParameters=""
+
+<br><br><hr>
+
 <a name="WarningMessages"><b>WarningMessages</b></a><br>
 <b>Version : </b>1.0 +<br>
 <br># AWStats can detect setup problems or show you important informations to have
diff --git a/wwwroot/cgi-bin/awstats.model.conf b/wwwroot/cgi-bin/awstats.model.conf
index a96ad7cd..8309d1fc 100644
--- a/wwwroot/cgi-bin/awstats.model.conf
+++ b/wwwroot/cgi-bin/awstats.model.conf
@@ -207,8 +207,8 @@ AllowToUpdateStatsFromBrowser=0
 # same time to prevent unknown conflicts problems and avoid DoS attacks when
 # AllowToUpdateStatsFromBrowser is set to 1.
 # Because, when you use lock file, you can experience sometimes problems in
-# lock file not correctly removed (when process is killed for example that
-# require that you remove the file manualy), this option is not enabled by
+# lock file not correctly removed (when process is killed for example, this
+# requires that you remove the file manualy), this option is not enabled by
 # default (Do not enable this option with no consol server access).
 # Possible values: 0 or 1
 # Default: 0
@@ -340,7 +340,7 @@ ArchiveLogRecords=0
 # ask AWStats to save last non corrupted file in a .bak file. This file is
 # stored in "DirData" directory with other 'historic files'.
 # Possible values: 0 or 1
-# Default: 1
+# Default: 0
 #
 KeepBackupOfHistoricFiles=0
 
@@ -435,7 +435,8 @@ ValidHTTPCodes="200 304"
 # different pages.
 # Warning, when set to 1, memory required to run AWStats is dramatically
 # increased if you have a lot of changing URLs (for example URLs with a random
-# id inside). Such web sites should not set this option to 1.
+# id inside). Such web sites should not set this option to 1 or use seriously
+# the next parameter URLWithQueryWithoutFollowingParameters.
 # Possible values:
 # 0 - URLs are cleaned from the query string (ie: "/mypage.html")
 # 1 - Full URL with query string is used     (ie: "/mypage.html?p=x&q=y")
@@ -444,6 +445,22 @@ ValidHTTPCodes="200 304"
 URLWithQuery=0
 
 
+# When URLWithQuery is on, you will get the full URL with all parameters in
+# URL reports. But among thoose parameters, sometimes you don't need a
+# particular parameter because it does not identify the page or because it's
+# a random ID changing for each access even if URL points to same page. In
+# such cases, it is higly recommanded to ask AWStats to remove such parameters
+# from the URL before counting, manipulating and storing it. Enter here list
+# of all non wanted parameters. For example if you enter "id", one hit on
+# /mypage.cgi?p=abc&id=Yo4UomP9d and /mypage.cgi?p=abc&id=Mu8fdxl3r
+# will be reported as 2 hits on /mypage.cgi?p=abc
+# This parameter is not used when URLWithQuery is 0.
+# Example: "PHPSESSID"
+# Default: ""
+# 
+URLWithQueryWithoutFollowingParameters=""
+
+
 # AWStats can detect setup problems or show you important informations to have
 # a better use. Keep this to 1, except if AWStats says you can change it.
 # Possible values: 0 or 1
diff --git a/wwwroot/cgi-bin/awstats.pl b/wwwroot/cgi-bin/awstats.pl
index f6cc73ff..01b89c12 100644
--- a/wwwroot/cgi-bin/awstats.pl
+++ b/wwwroot/cgi-bin/awstats.pl
@@ -212,6 +212,7 @@ use vars qw/
 use vars qw/
 @SessionsRange @HostAliases @AllowAccessFromWebToFollowingAuthenticatedUsers
 @DefaultFile @OnlyFiles @SkipDNSLookupFor @SkipFiles @SkipHosts @SkipUserAgents
+@URLWithQueryWithoutFollowingParameters
 @PluginsToLoad @DOWIndex @RobotsSearchIDOrder
 @_from_p @_from_h @_time_p @_time_h @_time_k
 @keylist
@@ -220,6 +221,7 @@ use vars qw/
 @HostAliases=();
 @AllowAccessFromWebToFollowingAuthenticatedUsers=();
 @DefaultFile = @OnlyFiles = @SkipDNSLookupFor = @SkipFiles = @SkipHosts = @SkipUserAgents = ();
+@URLWithQueryWithoutFollowingParameters = ();
 @PluginsToLoad = @DOWIndex = @RobotsSearchIDOrder = ();
 @_from_p = @_from_h = ();
 @_time_p = @_time_h = @_time_k = ();
@@ -1095,7 +1097,11 @@ sub Parse_Config {
 			$FoundValidHTTPCodes=1;
 			next;
 			}
-		if ($param =~ /^URLWithQuery/)			{ $URLWithQuery=$value; next; }
+		if ($param =~ /^URLWithQuery$/)			{ $URLWithQuery=$value; next; }
+		if ($param =~ /^URLWithQueryWithoutFollowingParameters$/)	{
+			foreach my $elem (split(/\s+/,$value))	{ push @URLWithQueryWithoutFollowingParameters,$elem; }
+			next;
+			}
 		if ($param =~ /^WarningMessages/)       { $WarningMessages=$value; next; }
 		if ($param =~ /^NbOfLinesForCorruptedLog/) { $NbOfLinesForCorruptedLog=$value; next; }
 		if ($param =~ /^Expires/)               { $Expires=$value; next; }
@@ -4612,9 +4618,20 @@ if ($UpdateStats && $FrameName ne "index" && $FrameName ne "mainleft") {	# Updat
 		my $urlwithnoquery;
 		if ($URLWithQuery) {
 			$urlwithnoquery=$field[$pos_url];
-			$urlwithnoquery =~ s/\?.*//;
+			my $foundparam=($urlwithnoquery =~ s/\?.*//);
 			# We combine the URL and query strings (for some IIS setup).
-			if ($pos_query >=0 && $field[$pos_query] && $field[$pos_query] ne "-") { $field[$pos_url] .= "?" . $field[$pos_query]; }
+			if (! $foundparam && $pos_query >=0 && $field[$pos_query] && $field[$pos_query] ne "-") {
+				$foundparam=1;
+				$field[$pos_url] .= "?" . $field[$pos_query];
+			}
+ 			# Remove params that are marked to be ignored in URLWithQueryWithoutFollowingParameters
+			if ($foundparam && @URLWithQueryWithoutFollowingParameters) {
+ 				map {$field[$pos_url] =~ s/$_=[^&]*//;} @URLWithQueryWithoutFollowingParameters;
+ 				# Cut starting or trailing ? or &
+ 				$field[$pos_url] =~ tr/&/&/s;
+ 				$field[$pos_url] =~ s/\?&/\?/;
+ 				$field[$pos_url] =~ s/[\?&]$//;
+ 			}
 		}
 		else {
 			# Trunc CGI parameters in URL