Patch Name: Jam Asian Text Supporting patch
Patch URI: http://www.pobox.com/~jam/unix/wordpress/#patches
Description: A patch to modify WordPress 2.0.5 to support Asian text in its excerpt constructing functions.
Version: 1.4
Author: Kazushi (Jam) Marukawa
Author URI: http://www.pobox.com/~jam/

How to install this:

 1. Apply this patch.
 2. That's all.

Index: ./wp-includes/comment-functions.php
===================================================================
--- ./wp-includes/comment-functions.php
+++ ./wp-includes/comment-functions.php	(working copy)
@@ -448,19 +448,7 @@
 function get_comment_excerpt() {
 	global $comment;
 	$comment_text = strip_tags($comment->comment_content);
-	$blah = explode(' ', $comment_text);
-	if (count($blah) > 20) {
-		$k = 20;
-		$use_dotdotdot = 1;
-	} else {
-		$k = count($blah);
-		$use_dotdotdot = 0;
-	}
-	$excerpt = '';
-	for ($i=0; $i<$k; $i++) {
-		$excerpt .= $blah[$i] . ' ';
-	}
-	$excerpt .= ($use_dotdotdot) ? '...' : '';
+	$excerpt = jamul_create_simple_excerpt($comment_text, 20);
 	return apply_filters('get_comment_excerpt', $excerpt);
 }
 
Index: ./wp-includes/feed-functions.php
===================================================================
--- ./wp-includes/feed-functions.php
+++ ./wp-includes/feed-functions.php	(working copy)
@@ -35,19 +35,7 @@
 		$content = strip_tags($content);
 	}
 	if ($cut) {
-		$blah = explode(' ', $content);
-		if (count($blah) > $cut) {
-			$k = $cut;
-			$use_dotdotdot = 1;
-		} else {
-			$k = count($blah);
-			$use_dotdotdot = 0;
-		}
-		for ($i=0; $i<$k; $i++) {
-			$excerpt .= $blah[$i].' ';
-		}
-		$excerpt .= ($use_dotdotdot) ? '...' : '';
-		$content = $excerpt;
+		$content = jamul_create_simple_excerpt($content, $cut);
 	}
 	$content = str_replace(']]>', ']]&gt;', $content);
 	echo wpj_utf8_encode($content);
Index: ./wp-includes/functions-formatting.php
===================================================================
--- ./wp-includes/functions-formatting.php
+++ ./wp-includes/functions-formatting.php	(working copy)
@@ -754,13 +754,7 @@
 		$text = apply_filters('the_content', $text);
 		$text = str_replace(']]>', ']]&gt;', $text);
 		$text = strip_tags($text);
-		$excerpt_length = 55;
-		$words = explode(' ', $text, $excerpt_length + 1);
-		if (count($words) > $excerpt_length) {
-			array_pop($words);
-			array_push($words, '[...]');
-			$text = implode(' ', $words);
-		}
+		$text = jamul_trim_excerpt($text, 55);
 	}
 	return $text;
 }
Index: ./wp-includes/functions-post.php
===================================================================
--- ./wp-includes/functions-post.php
+++ ./wp-includes/functions-post.php	(working copy)
@@ -625,7 +625,7 @@
 		$excerpt = strip_tags($post_excerpt?$post_excerpt:$post_content);
 		
 		if (strlen($excerpt) > 255) {
-			$excerpt = substr($excerpt,0,252) . '...';
+			$excerpt = jamul_substr($excerpt,0,252) . '...';
 		}
 		
 		$trackback_urls = explode(',', $tb_list);
@@ -712,7 +712,7 @@
 	if ( function_exists('mb_strcut') ) // For international trackbacks
     	$excerpt = mb_strcut($excerpt, 0, 252, get_settings('blog_charset')) . '...';
 	else
-		$excerpt = mb_substr($excerpt, 0, 84) . '...';
+		$excerpt = jamul_substr($excerpt, 0, 252) . '...';
 
 	$post_title = apply_filters('the_title', $post->post_title);
 	$post_title = strip_tags($post_title);
Index: ./wp-includes/functions.php
===================================================================
--- ./wp-includes/functions.php
+++ ./wp-includes/functions.php	(working copy)
@@ -2923,4 +2923,193 @@
 	return wpj_utf8_encode(apply_filters('wpj_comment_author', $author));
 }
 
+/*
+Plugin Name: Jam UTF8 Library
+Plugin URI: http://www.pobox.com/~jam/unix/wordpress/
+Description: Define several UTF8 friendly functions for developers.
+Version: 1.3
+Author: Kazushi (Jam) Marukawa
+Author URI: http://www.pobox.com/~jam/
+
+*/
+
+// The length of longest one word in average.  If the length of string
+// after trimming is longer than #word * the value below, trimming
+// function believes trimmed string doesn't contain whitespaces and
+// starts trimming it again using the length of string.
+define("JAMUL_LEN_MAX1WORD", 8);
+
+// The length of one word in average.  Use this value to calculate
+// the length of trimmed string.
+define("JAMUL_LEN_1WORD", 5);
+
+// The number of bytes used when WordPress looking around to find delimiters
+// (either a whitespace or a point where ASCII and other character switched).
+// This also represents the number of bytes of few characters.
+define("JAMUL_LEN_SEARCH", 15);
+
+function jamul_is_too_long($string, $count) {
+    if (strlen($string) > $count * JAMUL_LEN_MAX1WORD) {
+        return true;
+    }
+    return false;
+}
+
+function jamul_find_1stbyte($string, $pos=0, $stop=-1) {
+    $len = strlen($string);
+    if ($stop < 0 || $stop > $len) {
+        $stop = $len;
+    }
+    for (; $pos < $stop; $pos++) {
+        if ((ord($string[$pos]) < 0x80) || (ord($string[$pos]) >= 0xC0)) {
+            break;      // find 1st byte of multi-byte characters.
+        }
+    }
+    return $pos;
+}
+
+function jamul_find_1stbyte_reverse($string, $pos=-1, $stop=0) {
+    $len = strlen($string);
+    if ($pos < 0 || $pos >= $len) {
+        $pos = $len - 1;
+    }
+    for (; $pos >= $stop; $pos--) {
+        if ((ord($string[$pos]) < 0x80) || (ord($string[$pos]) >= 0xC0)) {
+            break;      // find 1st byte of multi-byte characters.
+        }
+    }
+    return $pos;
+}
+
+function jamul_find_delimiter($string, $pos=0, $min = -1, $max=-1) {
+    $len = strlen($string);
+    if ($pos == 0 || $pos < 0 || $pos >= $len) {
+        return $pos;
+    }
+    if ($min < 0) {
+        $min = max(0, $pos - JAMUL_LEN_SEARCH);
+    }
+    if ($max < 0 || $max >= $len) {
+        $max = min($len - 1, $pos + JAMUL_LEN_SEARCH);
+    }
+    if (ord($string[$pos]) < 0x80) {
+        // Found ASCII character at the trimming point.  So, trying
+        // to find new trimming point around $pos.  New trimming point
+        // should be on a whitespace or the transition from ASCII to
+        // other character.
+        $pos3 = -1;
+        for ($pos2 = $pos; $pos2 <= $max; $pos2++) {
+            if ($string[$pos2] == ' ') {
+                break;
+            } else if ($pos3 < 0 && ord($string[$pos2]) >= 0x80) {
+                $pos3 = $pos2;
+            }
+        }
+        if ($pos2 > $max && $pos3 >= 0) {
+            $pos2 = $pos3;
+        }
+        if ($pos2 > $max) {
+            $pos3 = -1;
+            for ($pos2 = $pos; $pos2 >= $min; $pos2--) {
+                if ($string[$pos2] == ' ') {
+                    break;
+                } else if ($pos3 < 0 && ord($string[$pos2]) >= 0x80) {
+                    $pos3 = $pos2 + 1;
+                }
+            }
+            if ($pos2 < $min && $pos3 >= 0) {
+                $pos2 = $pos3;
+            }
+        }
+        if ($pos2 <= $max && $pos2 >= $min) {
+            $pos = $pos2;
+        }
+    } else if ((ord($string[$pos]) >= 0x80) || (ord($string[$pos]) < 0xC0)) {
+        $pos = jamul_find_1stbyte($string, $pos, $max);
+    }
+    return $pos;
+}
+
+function jamul_truncate($string, $byte) {
+    $len = strlen($string);
+    if ($len <= $byte)
+        return $string;
+    $byte = jamul_find_1stbyte_reverse($string, $byte);
+    return substr($string, 0, $byte);
+}
+
+function jamul_substr($string, $start, $byte=-1) {
+    // mb_substr returns multi-byte characters' string with $len characters.
+    // However, what application needs is a string with multi-byte characters
+    // within $byte bytes data.  So, I implement it.
+    $len = strlen($string);
+    if ($start < 0) {
+        $start = $len - $start;
+    }
+    if ($byte < 0) {
+        return substr($string, $start);
+    } else if ($byte > $len - $start) {
+        $byte = $len - $start;
+    }
+    $byte = jamul_find_1stbyte_reverse($string, $start + $byte, $start)
+        - $start;
+    return substr($string, $start, $byte);
+}
+
+function jamul_substr_word($string, $start, $word=-1) {
+    $len = strlen($string);
+    if ($start < 0) {
+        $start = $len - $start;
+    }
+    if ($word < 0) {
+        return substr($string, $start);
+    }
+    $byte = $word * JAMUL_LEN_1WORD;
+    if ($byte >= $len - $start) {
+        return substr($string, $start);
+    }
+    $pos = jamul_find_1stbyte_reverse($string, $start + $byte, $start);
+    $pos = jamul_find_delimiter($string, $pos);
+    return substr($string, $start, $pos - $start);
+}
+
+function jamul_create_simple_excerpt($string, $count) {
+    $blah = explode(' ', $string);
+    if (count($blah) > $count) {
+        $k = $count;
+        $use_dotdotdot = 1;
+    } else {
+        $k = count($blah);
+        $use_dotdotdot = 0;
+    }
+    $excerpt = '';
+    for ($i=0; $i<$k; $i++) {
+        $excerpt .= $blah[$i] . ' ';
+    }
+    $excerpt .= ($use_dotdotdot) ? '...' : '';
+    if (jamul_is_too_long($excerpt, $count)) {
+        // too many bytes in $excerpt... this happens if $string doesn't
+        // contain whitespaces (common in Asian text).
+        return jamul_substr_word($string, 0, $count) . " ...";
+    }
+    return $excerpt;
+}
+
+function jamul_trim_excerpt($string, $count) {
+    $words = explode(' ', $string, $count + 1);
+    if (count($words) > $count) {
+        array_pop($words);
+        array_push($words, '[...]');
+        $text = implode(' ', $words);
+    } else {
+        $text = $string;
+    }
+    if (jamul_is_too_long($text, $count)) {
+        // too many bytes in $text... this happens if $string doesn't
+        // contain whitespaces (common in Asian text).
+        return jamul_substr_word($string, 0, $count) . " [...]";
+    }
+    return $text;
+}
+
 ?>
Index: ./wp-trackback.php
===================================================================
--- ./wp-trackback.php
+++ ./wp-trackback.php	(working copy)
@@ -97,8 +97,8 @@
 		$excerpt = mb_substr($excerpt, 0, 84, get_settings('blog_charset')) . '<a href="'.$tb_url.'"> (more...)</a>';
 		$title = mb_substr($title, 0, 30, get_settings('blog_charset')) . '...';
 	} else {
-		$excerpt = (strlen($excerpt) > 255) ? substr($excerpt, 0, 252) . '...' : $excerpt;
-		$title = (strlen($title) > 250) ? substr($title, 0, 250) . '...' : $title;
+		$excerpt = (strlen($excerpt) > 255) ? jamul_substr($excerpt, 0, 252) . '...' : $excerpt;
+		$title = (strlen($title) > 250) ? jamul_substr($title, 0, 250) . '...' : $title;
 	}
 
 	$comment_post_ID = $tb_id;