- Charset simpliciations. Patch #46 by Al.

10 files changed, 16 insertions, 63 deletions

diff --git a/includes/common.inc b/includes/common.inc
index 9909da1a2..3b30032d0 100644
--- a/includes/common.inc
+++ b/includes/common.inc
@@ -203,7 +203,7 @@ function variable_del($name) {
 }
 
 function drupal_specialchars($input) {
-  return htmlspecialchars($input, ENT_NOQUOTES, variable_get("site_charset", "iso-8859-1"));
+  return htmlspecialchars($input, ENT_NOQUOTES, "utf-8");
 }
 
 function table_cell($cell, $header = 0) {
@@ -1025,7 +1025,7 @@ $conf = variable_init(isset($conf) ? $conf : array());
 set_error_handler("error_handler");
 
 // spit out the correct charset http header
-header("Content-Type: text/html; charset=". variable_get("site_charset", "iso-8859-1"));
+header("Content-Type: text/html; charset=utf-8");
 
 // initialize installed modules:
 module_init();
diff --git a/includes/theme.inc b/includes/theme.inc
index dce4cff1b..643e39323 100644
--- a/includes/theme.inc
+++ b/includes/theme.inc
@@ -144,7 +144,7 @@ function theme_list($refresh = 0) {
 
 function theme_head($main = 0) {
   global $base_url;
-  $output .= "<meta http-equiv=\"Content-Type\" content=\"text/html; charset=". variable_get("site_charset", "iso-8859-1") ."\" />";
+  $output .= "<meta http-equiv=\"Content-Type\" content=\"text/html; charset=utf-8\" />";
   $output .= "<base href=\"$base_url/\" />\n";
   $output .= "<style type=\"text/css\">\n";
   $output .= "@import url(misc/drupal.css);\n";
diff --git a/modules/aggregator.module b/modules/aggregator.module
index d7916491c..9eb9b1aa1 100644
--- a/modules/aggregator.module
+++ b/modules/aggregator.module
@@ -706,7 +706,7 @@ function import_page_fd() {
 
   $result = db_query("SELECT * FROM feed ORDER BY title");
 
-  $output .= "<?xml version=\"1.0\" encoding=\"". variable_get("site_charset", "iso-8859-1") ."\"?>\n\n";
+  $output .= "<?xml version=\"1.0\" encoding=\"utf-8\"?>\n\n";
   $output .= "<rssfeeds version=\"0.1\">\n\n";
 
   while ($feed = db_fetch_object($result)) {
diff --git a/modules/aggregator/aggregator.module b/modules/aggregator/aggregator.module
index d7916491c..9eb9b1aa1 100644
--- a/modules/aggregator/aggregator.module
+++ b/modules/aggregator/aggregator.module
@@ -706,7 +706,7 @@ function import_page_fd() {
 
   $result = db_query("SELECT * FROM feed ORDER BY title");
 
-  $output .= "<?xml version=\"1.0\" encoding=\"". variable_get("site_charset", "iso-8859-1") ."\"?>\n\n";
+  $output .= "<?xml version=\"1.0\" encoding=\"utf-8\"?>\n\n";
   $output .= "<rssfeeds version=\"0.1\">\n\n";
 
   while ($feed = db_fetch_object($result)) {
diff --git a/modules/import.module b/modules/import.module
index d7916491c..9eb9b1aa1 100644
--- a/modules/import.module
+++ b/modules/import.module
@@ -706,7 +706,7 @@ function import_page_fd() {
 
   $result = db_query("SELECT * FROM feed ORDER BY title");
 
-  $output .= "<?xml version=\"1.0\" encoding=\"". variable_get("site_charset", "iso-8859-1") ."\"?>\n\n";
+  $output .= "<?xml version=\"1.0\" encoding=\"utf-8\"?>\n\n";
   $output .= "<rssfeeds version=\"0.1\">\n\n";
 
   while ($feed = db_fetch_object($result)) {
diff --git a/modules/node.module b/modules/node.module
index 1bbc95623..c62be185a 100644
--- a/modules/node.module
+++ b/modules/node.module
@@ -879,12 +879,11 @@ function node_feed($nodes = 0, $channel = array()) {
   while ($node = db_fetch_object($nodes)) {
     $item = node_load(array("nid" => $node->nid));
     $link = url("node/view/$item->nid");
-    $items .= format_rss_item($item->title, $link, $item->teaser);
+    $items .= format_rss_item($item->title, $link, ($item->teaser ? $item->teaser : $item->body));
   }
 
-  $output .= "<?xml version=\"1.0\" encoding=\"". variable_get("site_charset", "iso-8859-1") ."\"?>\n";
+  $output .= "<?xml version=\"1.0\" encoding=\"utf-8\"?>\n";
   $output .= "<!DOCTYPE rss [<!ENTITY % HTMLlat1 PUBLIC \"-//W3C//ENTITIES Latin 1 for XHTML//EN\" \"http://www.w3.org/TR/xhtml1/DTD/xhtml-lat1.ent\">]>\n";
-    // NOTE: &eacute; - for example - is the correct ISO-8859-1 translation of  (e acute) but apparently XML parsers don't (have to) understand it.  To solve this problem, we use a DTD that defines commonly used entity such as &eacute;.
   if (!$channel["version"]) $channel["version"] = "0.91";
   if (!$channel["title"]) $channel["title"] = variable_get("site_name", "drupal") ." - ". variable_get("site_slogan", "");
   if (!$channel["link"]) $channel["link"] = $base_url;
diff --git a/modules/node/node.module b/modules/node/node.module
index 1bbc95623..c62be185a 100644
--- a/modules/node/node.module
+++ b/modules/node/node.module
@@ -879,12 +879,11 @@ function node_feed($nodes = 0, $channel = array()) {
   while ($node = db_fetch_object($nodes)) {
     $item = node_load(array("nid" => $node->nid));
     $link = url("node/view/$item->nid");
-    $items .= format_rss_item($item->title, $link, $item->teaser);
+    $items .= format_rss_item($item->title, $link, ($item->teaser ? $item->teaser : $item->body));
   }
 
-  $output .= "<?xml version=\"1.0\" encoding=\"". variable_get("site_charset", "iso-8859-1") ."\"?>\n";
+  $output .= "<?xml version=\"1.0\" encoding=\"utf-8\"?>\n";
   $output .= "<!DOCTYPE rss [<!ENTITY % HTMLlat1 PUBLIC \"-//W3C//ENTITIES Latin 1 for XHTML//EN\" \"http://www.w3.org/TR/xhtml1/DTD/xhtml-lat1.ent\">]>\n";
-    // NOTE: &eacute; - for example - is the correct ISO-8859-1 translation of  (e acute) but apparently XML parsers don't (have to) understand it.  To solve this problem, we use a DTD that defines commonly used entity such as &eacute;.
   if (!$channel["version"]) $channel["version"] = "0.91";
   if (!$channel["title"]) $channel["title"] = variable_get("site_name", "drupal") ." - ". variable_get("site_slogan", "");
   if (!$channel["link"]) $channel["link"] = $base_url;
diff --git a/modules/system.module b/modules/system.module
index 801daa819..9834d7e05 100644
--- a/modules/system.module
+++ b/modules/system.module
@@ -129,31 +129,6 @@ function system_view_general() {
   $output .= form_select(t("Date format (medium)"), "date_format_medium", variable_get("date_format_medium", $datemedium[0]), $datemediumchoices, t("The medium sized date display."));
   $output .= form_select(t("Date format (long)"), "date_format_long", variable_get("date_format_long", $datelong[0]), $datelongchoices, t("Longer date format used for detailed display."));
 
-  $output .= "<hr />\n";
-
-  $output .= "<h3>". t("Character encoding settings") ."</h3>\n";
-
-  // charset settings
-  $charsets = array("iso-8859-1" => "iso-8859-1",
-                    "iso-8859-2" => "iso-8859-2",
-                    "iso-8859-3" => "iso-8859-3",
-                    "iso-8859-4" => "iso-8859-4",
-                    "iso-8859-5" => "iso-8859-5",
-                    "iso-8859-6" => "iso-8859-6",
-                    "iso-8859-7" => "iso-8859-7",
-                    "iso-8859-8" => "iso-8859-8",
-                    "iso-8859-8-i" => "iso-8859-8-i",
-                    "iso-8859-9" => "iso-8859-9",
-                    "utf-8" => "utf-8",
-                    "utf-16" => "utf-16",
-                    "utf-16be" => "utf-16be",
-                    "euc-kr" => "euc-kr",
-                    "koi8-r" => "koi8-r",
-                    "ks_c_5601-1987" => "ks_c_5601-1987",
-                    "tis-620" => "tis-620",
-                    "iso-2022-kr" => "iso-2022-kr",
-                    "us-ascii" => "us-ascii");
-  $output .= form_select(t("Character set"), "site_charset", variable_get("site_charset", "iso-8859-1"), $charsets, t("Select the character set (charset) used across your site. This will affect the content-type tag in web pages and the XML declarations in RSS feeds, among other things. You probably only need to change this if you need support for languages with non-ASCII characters like Korean, Polish, etc."));
   return $output;
 }
 
diff --git a/modules/system/system.module b/modules/system/system.module
index 801daa819..9834d7e05 100644
--- a/modules/system/system.module
+++ b/modules/system/system.module
@@ -129,31 +129,6 @@ function system_view_general() {
   $output .= form_select(t("Date format (medium)"), "date_format_medium", variable_get("date_format_medium", $datemedium[0]), $datemediumchoices, t("The medium sized date display."));
   $output .= form_select(t("Date format (long)"), "date_format_long", variable_get("date_format_long", $datelong[0]), $datelongchoices, t("Longer date format used for detailed display."));
 
-  $output .= "<hr />\n";
-
-  $output .= "<h3>". t("Character encoding settings") ."</h3>\n";
-
-  // charset settings
-  $charsets = array("iso-8859-1" => "iso-8859-1",
-                    "iso-8859-2" => "iso-8859-2",
-                    "iso-8859-3" => "iso-8859-3",
-                    "iso-8859-4" => "iso-8859-4",
-                    "iso-8859-5" => "iso-8859-5",
-                    "iso-8859-6" => "iso-8859-6",
-                    "iso-8859-7" => "iso-8859-7",
-                    "iso-8859-8" => "iso-8859-8",
-                    "iso-8859-8-i" => "iso-8859-8-i",
-                    "iso-8859-9" => "iso-8859-9",
-                    "utf-8" => "utf-8",
-                    "utf-16" => "utf-16",
-                    "utf-16be" => "utf-16be",
-                    "euc-kr" => "euc-kr",
-                    "koi8-r" => "koi8-r",
-                    "ks_c_5601-1987" => "ks_c_5601-1987",
-                    "tis-620" => "tis-620",
-                    "iso-2022-kr" => "iso-2022-kr",
-                    "us-ascii" => "us-ascii");
-  $output .= form_select(t("Character set"), "site_charset", variable_get("site_charset", "iso-8859-1"), $charsets, t("Select the character set (charset) used across your site. This will affect the content-type tag in web pages and the XML declarations in RSS feeds, among other things. You probably only need to change this if you need support for languages with non-ASCII characters like Korean, Polish, etc."));
   return $output;
 }
 
diff --git a/update.php b/update.php
index 81e1bac04..16f5a4361 100644
--- a/update.php
+++ b/update.php
@@ -68,7 +68,8 @@ $mysql_updates = array(
   "2003-05-18" => "update_53",
   "2003-05-24" => "update_54",
   "2003-05-31" => "update_55",
-  "2003-06-04" => "update_56"
+  "2003-06-04" => "update_56",
+  "2003-06-08" => "update_57"
 );
 
 // Update functions
@@ -725,6 +726,10 @@ function update_56() {
   update_sql("ALTER TABLE vocabulary CHANGE types nodes TEXT DEFAULT '' NOT NULL");
 }
 
+function update_57() {
+  update_sql("DELETE FROM variable WHERE name = 'site_charset'");
+}
+
 function update_upgrade3() {
   update_sql("INSERT INTO system VALUES ('archive.module','archive','module','',1)");
   update_sql("INSERT INTO system VALUES ('block.module','block','module','',1)");