- Reworked the CXX checking; now, _any_ user input will be checked

and the request will be terminated when something suspicious is
detected.  This will be logged in the watchdog.  With help from Marco.

- Fixed translation issue in the archive module.  Patch by Gerhard.

- Removed dead parameter from variable_get().  Patch by Chris Johnson.
Fixes bug #2111.

- Improved input checking of taxonomy module.  Patch by Gerhard.
Fixes bug #2112.

12 files changed, 107 insertions, 96 deletions

diff --git a/includes/common.inc b/includes/common.inc
index 12a71bbda..2346f2646 100644
--- a/includes/common.inc
+++ b/includes/common.inc
@@ -179,7 +179,7 @@ function variable_init($conf = array()) {
   return $conf;
 }
 
-function variable_get($name, $default, $object = 0) {
+function variable_get($name, $default) {
   global $conf;
 
   return isset($conf[$name]) ? $conf[$name] : $default;
@@ -413,12 +413,6 @@ function search_type($type = 0, $action = 0, $keys = 0, $options = 0) {
 function drupal_goto($url) {
 
   /*
-  ** Check the URL to prevent XSS attacks:
-  */
-
-  $url = check_url($url);
-
-  /*
   ** Translate & to simply &
   */
 
@@ -475,12 +469,61 @@ function referer_load() {
   }
 }
 
-function check_url($uri) {
-  /*
-  ** We pipe the request URI through htmlspecialchars() to prevent
-  ** XSS attacks.
-  */
+function xss_check_input_data($data) {
+
+  if (is_array($data)) {
+    /*
+    ** Form data can contain a number of nested arrays.
+    */
+
+    foreach ($data as $key => $value) {
+      xss_check_input_data($value);
+    }
+  }
+  else {
+    /*
+    ** Detect evil input data.
+    */
+
+    // check attributes:
+    $match  = preg_match("/\Wstyle\s*=[^>]+?>/i", $data);
+    $match += preg_match("/\Wdynsrc\s*=[^>]+?>/i", $data);
+    $match += preg_match("/\Wdatasrc\s*=[^>]+?>/i", $data);
+    $match += preg_match("/\Wdata\s*=[^>]+?>/i", $data);
+    $match += preg_match("/\Wlowsrc\s*=[^>]+?>/i", $data);
+    $match += preg_match("/\Wstyle\s*=[^>]+?>/i", $data);
+    $match += preg_match("/\Won[a-z]+\s*=[^>]+?>/i", $data);
+    $match += preg_match("/\Wsrc\s*=[\s'\"]*javascript[^>]+?>/i", $data);
+    $match += preg_match("/\Whref\s*=[\s'\"]*javascript:[^>]+?>/i", $data);
+    $match += preg_match("/\Whref\s*=[\s'\"]*javascript:[^>]+?>/i", $data);
+
+    // check tags:
+    $match += preg_match("/<\s*applet/i", $data);
+    $match += preg_match("/<\s*script/i", $data);
+    $match += preg_match("/<\s*object/i", $data);
+    $match += preg_match("/<\s*style/i", $data);
+    $match += preg_match("/<\s*embed/i", $data);
+    $match += preg_match("/<\s*form/i", $data);
+    $match += preg_match("/<\s*blink/i", $data);
+    $match += preg_match("/<\s*meta/i", $data);
+    $match += preg_match("/<\s*font/i", $data);
+    $match += preg_match("/<\s*html/i", $data);
+    $match += preg_match("/<\s*frame/i", $data);
+    $match += preg_match("/<\s*iframe/i", $data);
+    $match += preg_match("/<\s*layer/i", $data);
+    $match += preg_match("/<\s*ilayer/i", $data);
+    $match += preg_match("/<\s*head/i", $data);
+    $match += preg_match("/<\s*frameset/i", $data);
+    $match += preg_match("/<\s*xml/i", $data);
+
+    if ($match) {
+      watchdog("warning", "terminated request because of suspicious input data: ". drupal_specialchars($data));
+      die("terminated request because of suspicious input data");
+    }
+  }
+}
 
+function check_url($uri) {
   $uri = htmlspecialchars($uri, ENT_QUOTES);
 
   /*
@@ -979,6 +1022,7 @@ function timer_start() {
 }
 
 function drupal_page_header() {
+
   if (variable_get("dev_timer", 0)) {
     timer_start();
   }
@@ -1036,6 +1080,9 @@ set_error_handler("error_handler");
 // spit out the correct charset http header
 header("Content-Type: text/html; charset=utf-8");
 
+// filter input data:
+xss_check_input_data($_REQUEST);
+
 // initialize installed modules:
 module_init();
 
@@ -1045,4 +1092,5 @@ $locale = locale_init();
 // initialize theme:
 $theme = theme_init();
 
+
 ?>
diff --git a/modules/aggregator.module b/modules/aggregator.module
index 78adefc7c..1d384e706 100644
--- a/modules/aggregator.module
+++ b/modules/aggregator.module
@@ -305,6 +305,9 @@ function import_refresh($feed) {
     }
     fclose($fp);
 
+    // filter the input data:
+    xss_check_input_data($data);
+
     // parse the data:
     $xml_parser = xml_parser_create();
     xml_set_element_handler($xml_parser, "import_element_start", "import_element_end");
@@ -320,14 +323,6 @@ function import_refresh($feed) {
     $tt = array_flip(get_html_translation_table(HTML_ENTITIES));
     $tt["&apos;"] = "'";
 
-    /*
-    ** Strip invalid tags and provide default values (if required):
-    */
-
-    foreach ($channel as $key => $value) {
-      $channel[$key] = node_filter(strtr(trim($value), $tt));
-    }
-
     db_query("UPDATE feed SET timestamp = %d, link = '%s', description = '%s' WHERE fid = %d", time(), $channel["LINK"], $channel["DESCRIPTION"], $feed["fid"]);
 
     /*
diff --git a/modules/aggregator/aggregator.module b/modules/aggregator/aggregator.module
index 78adefc7c..1d384e706 100644
--- a/modules/aggregator/aggregator.module
+++ b/modules/aggregator/aggregator.module
@@ -305,6 +305,9 @@ function import_refresh($feed) {
     }
     fclose($fp);
 
+    // filter the input data:
+    xss_check_input_data($data);
+
     // parse the data:
     $xml_parser = xml_parser_create();
     xml_set_element_handler($xml_parser, "import_element_start", "import_element_end");
@@ -320,14 +323,6 @@ function import_refresh($feed) {
     $tt = array_flip(get_html_translation_table(HTML_ENTITIES));
     $tt["&apos;"] = "'";
 
-    /*
-    ** Strip invalid tags and provide default values (if required):
-    */
-
-    foreach ($channel as $key => $value) {
-      $channel[$key] = node_filter(strtr(trim($value), $tt));
-    }
-
     db_query("UPDATE feed SET timestamp = %d, link = '%s', description = '%s' WHERE fid = %d", time(), $channel["LINK"], $channel["DESCRIPTION"], $feed["fid"]);
 
     /*
diff --git a/modules/archive.module b/modules/archive.module
index 206d90f76..b69d4feb1 100644
--- a/modules/archive.module
+++ b/modules/archive.module
@@ -87,7 +87,7 @@ function archive_calendar($original = 0) {
 
   $output .= " <tr class=\"header-week\">";
   for ($i = 0; $i < 7; $i++) {
-    $output .= "<td>". substr(ucfirst(t(date("l", $firstcolumn + $i * 86400))), 0, 1) ."</td>";
+    $output .= "<td>". t(substr(ucfirst(date("l", $firstcolumn + $i * 86400)), 0, 2)) ."</td>";
   }
   $output .= "</tr>\n";
 
diff --git a/modules/archive/archive.module b/modules/archive/archive.module
index 206d90f76..b69d4feb1 100644
--- a/modules/archive/archive.module
+++ b/modules/archive/archive.module
@@ -87,7 +87,7 @@ function archive_calendar($original = 0) {
 
   $output .= " <tr class=\"header-week\">";
   for ($i = 0; $i < 7; $i++) {
-    $output .= "<td>". substr(ucfirst(t(date("l", $firstcolumn + $i * 86400))), 0, 1) ."</td>";
+    $output .= "<td>". t(substr(ucfirst(date("l", $firstcolumn + $i * 86400)), 0, 2)) ."</td>";
   }
   $output .= "</tr>\n";
 
diff --git a/modules/import.module b/modules/import.module
index 78adefc7c..1d384e706 100644
--- a/modules/import.module
+++ b/modules/import.module
@@ -305,6 +305,9 @@ function import_refresh($feed) {
     }
     fclose($fp);
 
+    // filter the input data:
+    xss_check_input_data($data);
+
     // parse the data:
     $xml_parser = xml_parser_create();
     xml_set_element_handler($xml_parser, "import_element_start", "import_element_end");
@@ -320,14 +323,6 @@ function import_refresh($feed) {
     $tt = array_flip(get_html_translation_table(HTML_ENTITIES));
     $tt["&apos;"] = "'";
 
-    /*
-    ** Strip invalid tags and provide default values (if required):
-    */
-
-    foreach ($channel as $key => $value) {
-      $channel[$key] = node_filter(strtr(trim($value), $tt));
-    }
-
     db_query("UPDATE feed SET timestamp = %d, link = '%s', description = '%s' WHERE fid = %d", time(), $channel["LINK"], $channel["DESCRIPTION"], $feed["fid"]);
 
     /*
diff --git a/modules/node.module b/modules/node.module
index aad95c00f..ccdd264ba 100644
--- a/modules/node.module
+++ b/modules/node.module
@@ -499,11 +499,6 @@ function node_comment_mode($nid) {
 }
 
 function node_filter($text) {
-  $text = preg_replace("/\Wstyle\s*=[^>]+?>/i", ">", $text);
-  $text = preg_replace("/\Won[a-z]+\s*=[^>]+?>/i", ">", $text);
-  $text = preg_replace("/\Wsrc\s*=[\s'\"]*javascript[^>]+?>/i", ">", $text);
-  $text = preg_replace("/\Whref\s*=[\s'\"]*javascript:[^>]+?>/i", ">", $text);
-
   if (variable_get("filter_html", 0)) {
     $text = node_filter_html($text);
   }
diff --git a/modules/node/node.module b/modules/node/node.module
index aad95c00f..ccdd264ba 100644
--- a/modules/node/node.module
+++ b/modules/node/node.module
@@ -499,11 +499,6 @@ function node_comment_mode($nid) {
 }
 
 function node_filter($text) {
-  $text = preg_replace("/\Wstyle\s*=[^>]+?>/i", ">", $text);
-  $text = preg_replace("/\Won[a-z]+\s*=[^>]+?>/i", ">", $text);
-  $text = preg_replace("/\Wsrc\s*=[\s'\"]*javascript[^>]+?>/i", ">", $text);
-  $text = preg_replace("/\Whref\s*=[\s'\"]*javascript:[^>]+?>/i", ">", $text);
-
   if (variable_get("filter_html", 0)) {
     $text = node_filter_html($text);
   }
diff --git a/modules/search.module b/modules/search.module
index 8a332263b..f024f0d5e 100644
--- a/modules/search.module
+++ b/modules/search.module
@@ -364,15 +364,7 @@ function search_view($keys) {
         theme("box", t("Search Results"), $output);
       }
       else {
-        // no results. try a substring search
-        $output = search_data("*". $keys. "*");
-
-        if ($output) {
-          theme("box", t("Search Results"), $output);
-        }
-        else {
-          theme("box", t("Search Results"), t("Your search yielded no results."));
-        }
+        theme("box", t("Search Results"), t("Your search yielded no results."));
       }
     }
 
diff --git a/modules/search/search.module b/modules/search/search.module
index 8a332263b..f024f0d5e 100644
--- a/modules/search/search.module
+++ b/modules/search/search.module
@@ -364,15 +364,7 @@ function search_view($keys) {
         theme("box", t("Search Results"), $output);
       }
       else {
-        // no results. try a substring search
-        $output = search_data("*". $keys. "*");
-
-        if ($output) {
-          theme("box", t("Search Results"), $output);
-        }
-        else {
-          theme("box", t("Search Results"), t("Your search yielded no results."));
-        }
+        theme("box", t("Search Results"), t("Your search yielded no results."));
       }
     }
 
diff --git a/modules/taxonomy.module b/modules/taxonomy.module
index 89427d139..567a8e5e3 100644
--- a/modules/taxonomy.module
+++ b/modules/taxonomy.module
@@ -657,23 +657,25 @@ function _prepare_insert($data, $stage) {
 function taxonomy_select_nodes($taxonomy, $pager = 1) {
   global $user;
 
-  if ($taxonomy->operator == "or") {
-    $sql = "SELECT DISTINCT(n.nid), n.title, n.type, n.created, n.changed, n.uid, n.static, n.created, u.name FROM node n LEFT JOIN term_node r ON n.nid = r.nid LEFT JOIN users u ON n.uid = u.uid WHERE r.tid IN ($taxonomy->str_tids) AND n.status = '1' ORDER BY static DESC, created DESC";
-    $sql_count = "SELECT COUNT(DISTINCT(n.nid)) FROM node n LEFT JOIN term_node r ON n.nid = r.nid LEFT JOIN users u ON n.uid = u.uid WHERE r.tid IN ($taxonomy->str_tids) AND n.status = '1'";
-  }
-  else {
-    $sql = "SELECT n.nid, n.title, n.type, n.created, n.changed, n.uid, u.name FROM node n LEFT JOIN term_node r ON n.nid = r.nid LEFT JOIN users u ON n.uid = u.uid WHERE r.tid IN ($taxonomy->str_tids) AND n.status = '1' GROUP BY n.nid, n.title, n.type, n.created, n.changed, n.uid, u.name HAVING COUNT(n.nid) = ".count($taxonomy->tids)." ORDER BY static DESC, created DESC";
+  if ($taxonomy->str_tids) {
+    if ($taxonomy->operator == "or") {
+      $sql = "SELECT DISTINCT(n.nid), n.title, n.type, n.created, n.changed, n.uid, n.static, n.created, u.name FROM node n LEFT JOIN term_node r ON n.nid = r.nid LEFT JOIN users u ON n.uid = u.uid WHERE r.tid IN ($taxonomy->str_tids) AND n.status = '1' ORDER BY static DESC, created DESC";
+      $sql_count = "SELECT COUNT(DISTINCT(n.nid)) FROM node n LEFT JOIN term_node r ON n.nid = r.nid LEFT JOIN users u ON n.uid = u.uid WHERE r.tid IN ($taxonomy->str_tids) AND n.status = '1'";
+    }
+    else {
+      $sql = "SELECT n.nid, n.title, n.type, n.created, n.changed, n.uid, u.name FROM node n LEFT JOIN term_node r ON n.nid = r.nid LEFT JOIN users u ON n.uid = u.uid WHERE r.tid IN ($taxonomy->str_tids) AND n.status = '1' GROUP BY n.nid, n.title, n.type, n.created, n.changed, n.uid, u.name HAVING COUNT(n.nid) = ".count($taxonomy->tids)." ORDER BY static DESC, created DESC";
 
-    // Special trick as we could not find anything better:
-    $count = db_num_rows(db_query("SELECT n.nid FROM node n LEFT JOIN term_node r ON n.nid = r.nid WHERE r.tid IN ($taxonomy->str_tids) AND n.status = '1' GROUP BY n.nid HAVING COUNT(n.nid) = ". count($taxonomy->tids)));
-    $sql_count = "SELECT $count";
-  }
+      // Special trick as we could not find anything better:
+      $count = db_num_rows(db_query("SELECT n.nid FROM node n LEFT JOIN term_node r ON n.nid = r.nid WHERE r.tid IN ($taxonomy->str_tids) AND n.status = '1' GROUP BY n.nid HAVING COUNT(n.nid) = ". count($taxonomy->tids)));
+      $sql_count = "SELECT $count";
+    }
 
-  if ($pager) {
-    $result = pager_query($sql, variable_get("default_nodes_main", 10) , 0, $sql_count);
-  }
-  else {
-    $result = db_query_range($sql, 0, 15);
+    if ($pager) {
+      $result = pager_query($sql, variable_get("default_nodes_main", 10) , 0, $sql_count);
+    }
+    else {
+      $result = db_query_range($sql, 0, 15);
+    }
   }
 
   return $result;
diff --git a/modules/taxonomy/taxonomy.module b/modules/taxonomy/taxonomy.module
index 89427d139..567a8e5e3 100644
--- a/modules/taxonomy/taxonomy.module
+++ b/modules/taxonomy/taxonomy.module
@@ -657,23 +657,25 @@ function _prepare_insert($data, $stage) {
 function taxonomy_select_nodes($taxonomy, $pager = 1) {
   global $user;
 
-  if ($taxonomy->operator == "or") {
-    $sql = "SELECT DISTINCT(n.nid), n.title, n.type, n.created, n.changed, n.uid, n.static, n.created, u.name FROM node n LEFT JOIN term_node r ON n.nid = r.nid LEFT JOIN users u ON n.uid = u.uid WHERE r.tid IN ($taxonomy->str_tids) AND n.status = '1' ORDER BY static DESC, created DESC";
-    $sql_count = "SELECT COUNT(DISTINCT(n.nid)) FROM node n LEFT JOIN term_node r ON n.nid = r.nid LEFT JOIN users u ON n.uid = u.uid WHERE r.tid IN ($taxonomy->str_tids) AND n.status = '1'";
-  }
-  else {
-    $sql = "SELECT n.nid, n.title, n.type, n.created, n.changed, n.uid, u.name FROM node n LEFT JOIN term_node r ON n.nid = r.nid LEFT JOIN users u ON n.uid = u.uid WHERE r.tid IN ($taxonomy->str_tids) AND n.status = '1' GROUP BY n.nid, n.title, n.type, n.created, n.changed, n.uid, u.name HAVING COUNT(n.nid) = ".count($taxonomy->tids)." ORDER BY static DESC, created DESC";
+  if ($taxonomy->str_tids) {
+    if ($taxonomy->operator == "or") {
+      $sql = "SELECT DISTINCT(n.nid), n.title, n.type, n.created, n.changed, n.uid, n.static, n.created, u.name FROM node n LEFT JOIN term_node r ON n.nid = r.nid LEFT JOIN users u ON n.uid = u.uid WHERE r.tid IN ($taxonomy->str_tids) AND n.status = '1' ORDER BY static DESC, created DESC";
+      $sql_count = "SELECT COUNT(DISTINCT(n.nid)) FROM node n LEFT JOIN term_node r ON n.nid = r.nid LEFT JOIN users u ON n.uid = u.uid WHERE r.tid IN ($taxonomy->str_tids) AND n.status = '1'";
+    }
+    else {
+      $sql = "SELECT n.nid, n.title, n.type, n.created, n.changed, n.uid, u.name FROM node n LEFT JOIN term_node r ON n.nid = r.nid LEFT JOIN users u ON n.uid = u.uid WHERE r.tid IN ($taxonomy->str_tids) AND n.status = '1' GROUP BY n.nid, n.title, n.type, n.created, n.changed, n.uid, u.name HAVING COUNT(n.nid) = ".count($taxonomy->tids)." ORDER BY static DESC, created DESC";
 
-    // Special trick as we could not find anything better:
-    $count = db_num_rows(db_query("SELECT n.nid FROM node n LEFT JOIN term_node r ON n.nid = r.nid WHERE r.tid IN ($taxonomy->str_tids) AND n.status = '1' GROUP BY n.nid HAVING COUNT(n.nid) = ". count($taxonomy->tids)));
-    $sql_count = "SELECT $count";
-  }
+      // Special trick as we could not find anything better:
+      $count = db_num_rows(db_query("SELECT n.nid FROM node n LEFT JOIN term_node r ON n.nid = r.nid WHERE r.tid IN ($taxonomy->str_tids) AND n.status = '1' GROUP BY n.nid HAVING COUNT(n.nid) = ". count($taxonomy->tids)));
+      $sql_count = "SELECT $count";
+    }
 
-  if ($pager) {
-    $result = pager_query($sql, variable_get("default_nodes_main", 10) , 0, $sql_count);
-  }
-  else {
-    $result = db_query_range($sql, 0, 15);
+    if ($pager) {
+      $result = pager_query($sql, variable_get("default_nodes_main", 10) , 0, $sql_count);
+    }
+    else {
+      $result = db_query_range($sql, 0, 15);
+    }
   }
 
   return $result;