Fixes #2276: Better friendly titles, portable ElggTranslit class, better units

4 files changed, 269 insertions, 18 deletions

diff --git a/engine/classes/ElggTranslit.php b/engine/classes/ElggTranslit.php
new file mode 100644
index 000000000..704c17f6a
--- /dev/null
+++ b/engine/classes/ElggTranslit.php
@@ -0,0 +1,238 @@
+<?php
+/**
+ * Elgg Transliterate
+ *
+ * For creating "friendly titles" for URLs
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * This software consists of voluntary contributions made by many individuals
+ * and is licensed under the LGPL. For more information, see
+ * <http://www.doctrine-project.org>.
+ *
+ * @author      Konsta Vesterinen <kvesteri@cc.hut.fi>
+ * @author      Jonathan H. Wage <jonwage@gmail.com>
+ *
+ * @author      Steve Clay <steve@mrclay.org>
+ * @package     Elgg.Core
+ */
+class ElggTranslit {
+
+	/**
+	 * Create a version of a string for embedding in a URL
+	 * @param string $string a UTF-8 string
+	 * @param string $separator
+	 * @return string
+	 */
+	static public function urlize($string, $separator = '-') {
+		// Iñtërnâtiônàlizætiøn, AND 日本語!
+
+		// Internationalization, AND 日本語!
+		$string = self::transliterateAscii($string);
+
+		// more translation
+		$string = strtr($string, array(
+			// Euro/GBP
+			"\xE2\x82\xAC" /* € */ => 'E', "\xC2\xA3" /* £ */ => 'GBP',
+		));
+
+		// remove all ASCII except 0-9a-zA-Z, hyphen, underscore, and whitespace
+		// note: "x" modifier did not work with this pattern.
+		$string = preg_replace('~['
+			. '\x00-\x08'  # control chars
+			. '\x0b\x0c'   # vert tab, form feed
+			. '\x0e-\x1f'  # control chars
+			. '\x21-\x2c'  # ! ... ,
+			. '\x2e\x2f'   # . slash
+			. '\x3a-\x40'  # : ... @
+			. '\x5b-\x5e'  # [ ... ^
+			. '\x60'       # `
+			. '\x7b-\x7f'  # { ... DEL
+			. ']~', '', $string);
+		$string = strtr($string, '', '');
+
+		// internationalization, and 日本語!
+		// note: not using elgg_strtolower to keep this class portable
+		$string = is_callable('mb_strtolower')
+			? mb_strtolower($string, 'UTF-8')
+			: strtolower($string);
+
+		// split by ASCII chars not in 0-9a-zA-Z
+		// note: we cannot use [^0-9a-zA-Z] because that matches multibyte chars.
+		// note: "x" modifier did not work with this pattern.
+		$pattern = '~['
+			. '\x00-\x2f'  # controls ... slash
+			. '\x3a-\x40'  # : ... @
+			. '\x5b-\x60'  # [ ... `
+			. '\x7b-\x7f'  # { ... DEL
+			. ']+~x';
+
+		// ['internationalization', 'and', '日本語']
+		$words = preg_split($pattern, $string, -1, PREG_SPLIT_NO_EMPTY);
+
+		// ['internationalization', 'and', '%E6%97%A5%E6%9C%AC%E8%AA%9E']
+		$words = array_map('urlencode', $words);
+
+		// internationalization-and-%E6%97%A5%E6%9C%AC%E8%AA%9E
+		return implode($separator, $words);
+	}
+
+	/**
+	 * Transliterate Western multibyte chars to ASCII
+	 * @param string $utf8 a UTF-8 string
+	 * @return string
+	 */
+	static public function transliterateAscii($utf8) {
+		static $map = null;
+		if (!preg_match('/[\x80-\xff]/', $utf8)) {
+			return $utf8;
+		}
+		if (null === $map) {
+			$map = self::getAsciiTranslitMap();
+		}
+		return strtr($utf8, $map);
+	}
+
+	/**
+	 * Get array of UTF-8 (NFC) character replacements.
+	 *
+	 * @return array
+	 */
+	static public function getAsciiTranslitMap() {
+		return array(
+			// Decompositions for Latin-1 Supplement
+			"\xC2\xAA" /* ª */ => 'a', "\xC2\xBA" /* º */ => 'o', "\xC3\x80" /* À */ => 'A',
+			"\xC3\x81" /* Á */ => 'A', "\xC3\x82" /* Â */ => 'A', "\xC3\x83" /* Ã */ => 'A',
+			"\xC3\x84" /* Ä */ => 'A', "\xC3\x85" /* Å */ => 'A', "\xC3\x86" /* Æ */ => 'AE',
+			"\xC3\x87" /* Ç */ => 'C', "\xC3\x88" /* È */ => 'E', "\xC3\x89" /* É */ => 'E',
+			"\xC3\x8A" /* Ê */ => 'E', "\xC3\x8B" /* Ë */ => 'E', "\xC3\x8C" /* Ì */ => 'I',
+			"\xC3\x8D" /* Í */ => 'I', "\xC3\x8E" /* Î */ => 'I', "\xC3\x8F" /* Ï */ => 'I',
+			"\xC3\x90" /* Ð */ => 'D', "\xC3\x91" /* Ñ */ => 'N', "\xC3\x92" /* Ò */ => 'O',
+			"\xC3\x93" /* Ó */ => 'O', "\xC3\x94" /* Ô */ => 'O', "\xC3\x95" /* Õ */ => 'O',
+			"\xC3\x96" /* Ö */ => 'O', "\xC3\x99" /* Ù */ => 'U', "\xC3\x9A" /* Ú */ => 'U',
+			"\xC3\x9B" /* Û */ => 'U', "\xC3\x9C" /* Ü */ => 'U', "\xC3\x9D" /* Ý */ => 'Y',
+			"\xC3\x9E" /* Þ */ => 'TH', "\xC3\x9F" /* ß */ => 'ss', "\xC3\xA0" /* à */ => 'a',
+			"\xC3\xA1" /* á */ => 'a', "\xC3\xA2" /* â */ => 'a', "\xC3\xA3" /* ã */ => 'a',
+			"\xC3\xA4" /* ä */ => 'a', "\xC3\xA5" /* å */ => 'a', "\xC3\xA6" /* æ */ => 'ae',
+			"\xC3\xA7" /* ç */ => 'c', "\xC3\xA8" /* è */ => 'e', "\xC3\xA9" /* é */ => 'e',
+			"\xC3\xAA" /* ê */ => 'e', "\xC3\xAB" /* ë */ => 'e', "\xC3\xAC" /* ì */ => 'i',
+			"\xC3\xAD" /* í */ => 'i', "\xC3\xAE" /* î */ => 'i', "\xC3\xAF" /* ï */ => 'i',
+			"\xC3\xB0" /* ð */ => 'd', "\xC3\xB1" /* ñ */ => 'n', "\xC3\xB2" /* ò */ => 'o',
+			"\xC3\xB3" /* ó */ => 'o', "\xC3\xB4" /* ô */ => 'o', "\xC3\xB5" /* õ */ => 'o',
+			"\xC3\xB6" /* ö */ => 'o', "\xC3\xB8" /* ø */ => 'o', "\xC3\xB9" /* ù */ => 'u',
+			"\xC3\xBA" /* ú */ => 'u', "\xC3\xBB" /* û */ => 'u', "\xC3\xBC" /* ü */ => 'u',
+			"\xC3\xBD" /* ý */ => 'y', "\xC3\xBE" /* þ */ => 'th', "\xC3\xBF" /* ÿ */ => 'y',
+			"\xC3\x98" /* Ø */ => 'O',
+			// Decompositions for Latin Extended-A
+			"\xC4\x80" /* Ā */ => 'A', "\xC4\x81" /* ā */ => 'a', "\xC4\x82" /* Ă */ => 'A',
+			"\xC4\x83" /* ă */ => 'a', "\xC4\x84" /* Ą */ => 'A', "\xC4\x85" /* ą */ => 'a',
+			"\xC4\x86" /* Ć */ => 'C', "\xC4\x87" /* ć */ => 'c', "\xC4\x88" /* Ĉ */ => 'C',
+			"\xC4\x89" /* ĉ */ => 'c', "\xC4\x8A" /* Ċ */ => 'C', "\xC4\x8B" /* ċ */ => 'c',
+			"\xC4\x8C" /* Č */ => 'C', "\xC4\x8D" /* č */ => 'c', "\xC4\x8E" /* Ď */ => 'D',
+			"\xC4\x8F" /* ď */ => 'd', "\xC4\x90" /* Đ */ => 'D', "\xC4\x91" /* đ */ => 'd',
+			"\xC4\x92" /* Ē */ => 'E', "\xC4\x93" /* ē */ => 'e', "\xC4\x94" /* Ĕ */ => 'E',
+			"\xC4\x95" /* ĕ */ => 'e', "\xC4\x96" /* Ė */ => 'E', "\xC4\x97" /* ė */ => 'e',
+			"\xC4\x98" /* Ę */ => 'E', "\xC4\x99" /* ę */ => 'e', "\xC4\x9A" /* Ě */ => 'E',
+			"\xC4\x9B" /* ě */ => 'e', "\xC4\x9C" /* Ĝ */ => 'G', "\xC4\x9D" /* ĝ */ => 'g',
+			"\xC4\x9E" /* Ğ */ => 'G', "\xC4\x9F" /* ğ */ => 'g', "\xC4\xA0" /* Ġ */ => 'G',
+			"\xC4\xA1" /* ġ */ => 'g', "\xC4\xA2" /* Ģ */ => 'G', "\xC4\xA3" /* ģ */ => 'g',
+			"\xC4\xA4" /* Ĥ */ => 'H', "\xC4\xA5" /* ĥ */ => 'h', "\xC4\xA6" /* Ħ */ => 'H',
+			"\xC4\xA7" /* ħ */ => 'h', "\xC4\xA8" /* Ĩ */ => 'I', "\xC4\xA9" /* ĩ */ => 'i',
+			"\xC4\xAA" /* Ī */ => 'I', "\xC4\xAB" /* ī */ => 'i', "\xC4\xAC" /* Ĭ */ => 'I',
+			"\xC4\xAD" /* ĭ */ => 'i', "\xC4\xAE" /* Į */ => 'I', "\xC4\xAF" /* į */ => 'i',
+			"\xC4\xB0" /* İ */ => 'I', "\xC4\xB1" /* ı */ => 'i', "\xC4\xB2" /* IJ */ => 'IJ',
+			"\xC4\xB3" /* ij */ => 'ij', "\xC4\xB4" /* Ĵ */ => 'J', "\xC4\xB5" /* ĵ */ => 'j',
+			"\xC4\xB6" /* Ķ */ => 'K', "\xC4\xB7" /* ķ */ => 'k', "\xC4\xB8" /* ĸ */ => 'k',
+			"\xC4\xB9" /* Ĺ */ => 'L', "\xC4\xBA" /* ĺ */ => 'l', "\xC4\xBB" /* Ļ */ => 'L',
+			"\xC4\xBC" /* ļ */ => 'l', "\xC4\xBD" /* Ľ */ => 'L', "\xC4\xBE" /* ľ */ => 'l',
+			"\xC4\xBF" /* Ŀ */ => 'L', "\xC5\x80" /* ŀ */ => 'l', "\xC5\x81" /* Ł */ => 'L',
+			"\xC5\x82" /* ł */ => 'l', "\xC5\x83" /* Ń */ => 'N', "\xC5\x84" /* ń */ => 'n',
+			"\xC5\x85" /* Ņ */ => 'N', "\xC5\x86" /* ņ */ => 'n', "\xC5\x87" /* Ň */ => 'N',
+			"\xC5\x88" /* ň */ => 'n', "\xC5\x89" /* ʼn */ => 'N', "\xC5\x8A" /* Ŋ */ => 'n',
+			"\xC5\x8B" /* ŋ */ => 'N', "\xC5\x8C" /* Ō */ => 'O', "\xC5\x8D" /* ō */ => 'o',
+			"\xC5\x8E" /* Ŏ */ => 'O', "\xC5\x8F" /* ŏ */ => 'o', "\xC5\x90" /* Ő */ => 'O',
+			"\xC5\x91" /* ő */ => 'o', "\xC5\x92" /* Œ */ => 'OE', "\xC5\x93" /* œ */ => 'oe',
+			"\xC5\x94" /* Ŕ */ => 'R', "\xC5\x95" /* ŕ */ => 'r', "\xC5\x96" /* Ŗ */ => 'R',
+			"\xC5\x97" /* ŗ */ => 'r', "\xC5\x98" /* Ř */ => 'R', "\xC5\x99" /* ř */ => 'r',
+			"\xC5\x9A" /* Ś */ => 'S', "\xC5\x9B" /* ś */ => 's', "\xC5\x9C" /* Ŝ */ => 'S',
+			"\xC5\x9D" /* ŝ */ => 's', "\xC5\x9E" /* Ş */ => 'S', "\xC5\x9F" /* ş */ => 's',
+			"\xC5\xA0" /* Š */ => 'S', "\xC5\xA1" /* š */ => 's', "\xC5\xA2" /* Ţ */ => 'T',
+			"\xC5\xA3" /* ţ */ => 't', "\xC5\xA4" /* Ť */ => 'T', "\xC5\xA5" /* ť */ => 't',
+			"\xC5\xA6" /* Ŧ */ => 'T', "\xC5\xA7" /* ŧ */ => 't', "\xC5\xA8" /* Ũ */ => 'U',
+			"\xC5\xA9" /* ũ */ => 'u', "\xC5\xAA" /* Ū */ => 'U', "\xC5\xAB" /* ū */ => 'u',
+			"\xC5\xAC" /* Ŭ */ => 'U', "\xC5\xAD" /* ŭ */ => 'u', "\xC5\xAE" /* Ů */ => 'U',
+			"\xC5\xAF" /* ů */ => 'u', "\xC5\xB0" /* Ű */ => 'U', "\xC5\xB1" /* ű */ => 'u',
+			"\xC5\xB2" /* Ų */ => 'U', "\xC5\xB3" /* ų */ => 'u', "\xC5\xB4" /* Ŵ */ => 'W',
+			"\xC5\xB5" /* ŵ */ => 'w', "\xC5\xB6" /* Ŷ */ => 'Y', "\xC5\xB7" /* ŷ */ => 'y',
+			"\xC5\xB8" /* Ÿ */ => 'Y', "\xC5\xB9" /* Ź */ => 'Z', "\xC5\xBA" /* ź */ => 'z',
+			"\xC5\xBB" /* Ż */ => 'Z', "\xC5\xBC" /* ż */ => 'z', "\xC5\xBD" /* Ž */ => 'Z',
+			"\xC5\xBE" /* ž */ => 'z', "\xC5\xBF" /* ſ */ => 's',
+			// Decompositions for Latin Extended-B
+			"\xC8\x98" /* Ș */ => 'S', "\xC8\x99" /* ș */ => 's',
+			"\xC8\x9A" /* Ț */ => 'T', "\xC8\x9B" /* ț */ => 't',
+			// unmarked
+			"\xC6\xA0" /* Ơ */ => 'O', "\xC6\xA1" /* ơ */ => 'o',
+			"\xC6\xAF" /* Ư */ => 'U', "\xC6\xB0" /* ư */ => 'u',
+			// grave accent
+			"\xE1\xBA\xA6" /* Ầ */ => 'A', "\xE1\xBA\xA7" /* ầ */ => 'a',
+			"\xE1\xBA\xB0" /* Ằ */ => 'A', "\xE1\xBA\xB1" /* ằ */ => 'a',
+			"\xE1\xBB\x80" /* Ề */ => 'E', "\xE1\xBB\x81" /* ề */ => 'e',
+			"\xE1\xBB\x92" /* Ồ */ => 'O', "\xE1\xBB\x93" /* ồ */ => 'o',
+			"\xE1\xBB\x9C" /* Ờ */ => 'O', "\xE1\xBB\x9D" /* ờ */ => 'o',
+			"\xE1\xBB\xAA" /* Ừ */ => 'U', "\xE1\xBB\xAB" /* ừ */ => 'u',
+			"\xE1\xBB\xB2" /* Ỳ */ => 'Y', "\xE1\xBB\xB3" /* ỳ */ => 'y',
+			// hook
+			"\xE1\xBA\xA2" /* Ả */ => 'A', "\xE1\xBA\xA3" /* ả */ => 'a',
+			"\xE1\xBA\xA8" /* Ẩ */ => 'A', "\xE1\xBA\xA9" /* ẩ */ => 'a',
+			"\xE1\xBA\xB2" /* Ẳ */ => 'A', "\xE1\xBA\xB3" /* ẳ */ => 'a',
+			"\xE1\xBA\xBA" /* Ẻ */ => 'E', "\xE1\xBA\xBB" /* ẻ */ => 'e',
+			"\xE1\xBB\x82" /* Ể */ => 'E', "\xE1\xBB\x83" /* ể */ => 'e',
+			"\xE1\xBB\x88" /* Ỉ */ => 'I', "\xE1\xBB\x89" /* ỉ */ => 'i',
+			"\xE1\xBB\x8E" /* Ỏ */ => 'O', "\xE1\xBB\x8F" /* ỏ */ => 'o',
+			"\xE1\xBB\x94" /* Ổ */ => 'O', "\xE1\xBB\x95" /* ổ */ => 'o',
+			"\xE1\xBB\x9E" /* Ở */ => 'O', "\xE1\xBB\x9F" /* ở */ => 'o',
+			"\xE1\xBB\xA6" /* Ủ */ => 'U', "\xE1\xBB\xA7" /* ủ */ => 'u',
+			"\xE1\xBB\xAC" /* Ử */ => 'U', "\xE1\xBB\xAD" /* ử */ => 'u',
+			"\xE1\xBB\xB6" /* Ỷ */ => 'Y', "\xE1\xBB\xB7" /* ỷ */ => 'y',
+			// tilde
+			"\xE1\xBA\xAA" /* Ẫ */ => 'A', "\xE1\xBA\xAB" /* ẫ */ => 'a',
+			"\xE1\xBA\xB4" /* Ẵ */ => 'A', "\xE1\xBA\xB5" /* ẵ */ => 'a',
+			"\xE1\xBA\xBC" /* Ẽ */ => 'E', "\xE1\xBA\xBD" /* ẽ */ => 'e',
+			"\xE1\xBB\x84" /* Ễ */ => 'E', "\xE1\xBB\x85" /* ễ */ => 'e',
+			"\xE1\xBB\x96" /* Ỗ */ => 'O', "\xE1\xBB\x97" /* ỗ */ => 'o',
+			"\xE1\xBB\xA0" /* Ỡ */ => 'O', "\xE1\xBB\xA1" /* ỡ */ => 'o',
+			"\xE1\xBB\xAE" /* Ữ */ => 'U', "\xE1\xBB\xAF" /* ữ */ => 'u',
+			"\xE1\xBB\xB8" /* Ỹ */ => 'Y', "\xE1\xBB\xB9" /* ỹ */ => 'y',
+			// acute accent
+			"\xE1\xBA\xA4" /* Ấ */ => 'A', "\xE1\xBA\xA5" /* ấ */ => 'a',
+			"\xE1\xBA\xAE" /* Ắ */ => 'A', "\xE1\xBA\xAF" /* ắ */ => 'a',
+			"\xE1\xBA\xBE" /* Ế */ => 'E', "\xE1\xBA\xBF" /* ế */ => 'e',
+			"\xE1\xBB\x90" /* Ố */ => 'O', "\xE1\xBB\x91" /* ố */ => 'o',
+			"\xE1\xBB\x9A" /* Ớ */ => 'O', "\xE1\xBB\x9B" /* ớ */ => 'o',
+			"\xE1\xBB\xA8" /* Ứ */ => 'U', "\xE1\xBB\xA9" /* ứ */ => 'u',
+			// dot below
+			"\xE1\xBA\xA0" /* Ạ */ => 'A', "\xE1\xBA\xA1" /* ạ */ => 'a',
+			"\xE1\xBA\xAC" /* Ậ */ => 'A', "\xE1\xBA\xAD" /* ậ */ => 'a',
+			"\xE1\xBA\xB6" /* Ặ */ => 'A', "\xE1\xBA\xB7" /* ặ */ => 'a',
+			"\xE1\xBA\xB8" /* Ẹ */ => 'E', "\xE1\xBA\xB9" /* ẹ */ => 'e',
+			"\xE1\xBB\x86" /* Ệ */ => 'E', "\xE1\xBB\x87" /* ệ */ => 'e',
+			"\xE1\xBB\x8A" /* Ị */ => 'I', "\xE1\xBB\x8B" /* ị */ => 'i',
+			"\xE1\xBB\x8C" /* Ọ */ => 'O', "\xE1\xBB\x8D" /* ọ */ => 'o',
+			"\xE1\xBB\x98" /* Ộ */ => 'O', "\xE1\xBB\x99" /* ộ */ => 'o',
+			"\xE1\xBB\xA2" /* Ợ */ => 'O', "\xE1\xBB\xA3" /* ợ */ => 'o',
+			"\xE1\xBB\xA4" /* Ụ */ => 'U', "\xE1\xBB\xA5" /* ụ */ => 'u',
+			"\xE1\xBB\xB0" /* Ự */ => 'U', "\xE1\xBB\xB1" /* ự */ => 'u',
+			"\xE1\xBB\xB4" /* Ỵ */ => 'Y', "\xE1\xBB\xB5" /* ỵ */ => 'y',
+		);
+	}
+}
diff --git a/engine/lib/output.php b/engine/lib/output.php
index b1245a924..7bfc4be6e 100644
--- a/engine/lib/output.php
+++ b/engine/lib/output.php
@@ -310,19 +310,11 @@ function elgg_get_friendly_title($title) {
 		return $result;
 	}
 
-	// @todo not using this because of locale concerns
-	//$title = iconv('UTF-8', 'ASCII//TRANSLIT', $title);
-
-	// @todo this uses a utf8 character class. can use if
-	// we want to support utf8 in the url.
-	//$title = preg_replace('/[^\p{L}\- ]/u', '', $title);
-
-	// use A-Za-z0-9_ instead of \w because \w is locale sensitive
-	$title = preg_replace("/[^A-Za-z0-9_\- ]/", "", $title);
-	$title = str_replace(" ", "-", $title);
-	$title = str_replace("--", "-", $title);
-	$title = trim($title);
-	$title = elgg_strtolower($title);
+	// handle some special cases
+	$title = str_replace('&amp;', 'and', $title);
+
+	$title = ElggTranslit::urlize($title);
+
 	return $title;
 }
 
diff --git a/engine/tests/api/metadata.php b/engine/tests/api/metadata.php
index 244036f80..c63b0cbec 100644
--- a/engine/tests/api/metadata.php
+++ b/engine/tests/api/metadata.php
@@ -28,6 +28,8 @@ class ElggCoreMetadataAPITest extends ElggCoreUnitTest {
 
 	public function testGetMetastringById() {
 		foreach (array('metaUnitTest', 'metaunittest', 'METAUNITTEST') as $string) {
+			// in case previous tests failed to cleanup after themselves
+			$this->delete_metastrings($string);
 			$this->create_metastring($string);
 		}
 
@@ -194,11 +196,19 @@ class ElggCoreMetadataAPITest extends ElggCoreUnitTest {
 		$u2->delete();
 	}
 
+	protected function delete_metastrings($string) {
+		global $CONFIG, $METASTRINGS_CACHE, $METASTRINGS_DEADNAME_CACHE;
+		$METASTRINGS_CACHE = $METASTRINGS_DEADNAME_CACHE = array();
+
+		$string = sanitise_string($string);
+		mysql_query("DELETE FROM {$CONFIG->dbprefix}metastrings WHERE string = BINARY '$string'");
+	}
 
 	protected function create_metastring($string) {
 		global $CONFIG, $METASTRINGS_CACHE, $METASTRINGS_DEADNAME_CACHE;
 		$METASTRINGS_CACHE = $METASTRINGS_DEADNAME_CACHE = array();
 
+		$string = sanitise_string($string);
 		mysql_query("INSERT INTO {$CONFIG->dbprefix}metastrings (string) VALUES ('$string')");
 		$this->metastrings[$string] = mysql_insert_id();
 	}
diff --git a/engine/tests/regression/trac_bugs.php b/engine/tests/regression/trac_bugs.php
index 26a45ab6a..e81bd6936 100644
--- a/engine/tests/regression/trac_bugs.php
+++ b/engine/tests/regression/trac_bugs.php
@@ -202,14 +202,25 @@ class ElggCoreRegressionBugsTest extends ElggCoreUnitTest {
 
 	/**
 	 * http://trac.elgg.org/ticket/3210 - Don't remove -s in friendly titles
-	 * @todo: http://trac.elgg.org/ticket/2276 - improve char encoding
+	 * http://trac.elgg.org/ticket/2276 - improve char encoding
 	 */
 	public function test_friendly_title() {
 		$cases = array(
-			'Simple Test' => 'simple-test',
-			'Test top-level page' => 'test-top-level-page',
-//			'éclair' => 'éclair',
-//			'English, Español, and 日本語' => 'english-español-and-日本語'
+			// hyphen, underscore and ASCII whitespace replaced by separator,
+			// other non-alphanumeric ASCII removed
+			"a-a_a a\na\ra\ta\va!a\"a#a\$a%a&a'a(a)a*a+a,a.a/a:a;a<a=a>a?a@a[a\\a]a^a`a{a|a}a~a"
+			=> "a-a-a-a-a-a-aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa",
+
+			// separators trimmed
+			"-_ hello _-" => "hello",
+
+			// accents removed, lower case, other multibyte chars are URL encoded
+			"I\xC3\xB1t\xC3\xABrn\xC3\xA2ti\xC3\xB4n\xC3\xA0liz\xC3\xA6ti\xC3\xB8n, AND \xE6\x97\xA5\xE6\x9C\xAC\xE8\xAA\x9E"
+				// Iñtërnâtiônàlizætiøn, AND 日本語
+			=> 'internationalizaetion-and-%E6%97%A5%E6%9C%AC%E8%AA%9E',
+
+			// some HTML entity replacements
+			"Me &amp; You" => 'me-and-you',
 		);
 
 		foreach ($cases as $case => $expected) {