added loose name compare and improved rename

3 files changed, 131 insertions, 12 deletions

diff --git a/lib/class/artist.class.php b/lib/class/artist.class.php
index 1b96b0aa..334472e3 100644
--- a/lib/class/artist.class.php
+++ b/lib/class/artist.class.php
@@ -205,7 +205,7 @@ class Artist {
                 @param $newname the artist's new name, either a new
                         artist will be created or songs added to existing
                         artist if name exists already
-                @return the id of the new artist
+                @return the id of the new artist, or false if an error
         */
         function rename($newname) {
 
@@ -221,11 +221,12 @@ class Artist {
                 /* check that it wasn't just whitespace that we were called to change */
                 if ($newid == $this->id) {
 			$GLOBALS['error']->add_error('artist_name',_("Error: Name Identical"));
-                        return $newid;
+                        return false;
                 }
 
                 /* now we can just call merge */
-                $this->merge($newid);
+                if (!$this->merge($newid))
+                	return false;
 
                 //now return id
                 return $newid;
@@ -237,6 +238,7 @@ class Artist {
                 @discussion changes the artist id of all songs by this artist
                         to the given id and deletes self from db
                 @param $newid the new artist id that this artist's songs should have
+                @return the name of the new artist on success, false if error
         */
         function merge($newid) {
 
@@ -249,15 +251,14 @@ class Artist {
 		} 
 
                 // First check newid exists
-                $check_exists_qstring = "SELECT name FROM artist WHERE id='" . sql_escape($newid) . "'";
+                $check_exists_qstring = "SELECT name FROM artist WHERE id='" . $newid . "'"; //no need to escape newid, it's numeric
                 $check_exists_query = mysql_query($check_exists_qstring, dbh());
 
-                if ($check_exists_results = mysql_fetch_assoc($check_exists_query)) {
-
+                if ($check_exists_result = mysql_fetch_assoc($check_exists_query)) {
                         $NewName = $check_exists_result['name'];
 
                         // Now the query
-                        $sql = "UPDATE song SET artist='" . sql_escape($newid) . "' " . 
+                        $sql = "UPDATE song SET artist='" . $newid . "' " . 
 				"WHERE artist='" . sql_escape($this->id) . "'";
                         $db_results = mysql_query($sql, dbh());
 
@@ -266,9 +267,11 @@ class Artist {
 			/* If we've done the merege we need to clean up */
 			$catalog->clean_artists();
 			$catalog->clean_albums();
+			
+			return $NewName;
                 } 
 		else {
-			$GLOBALS['error']->add_error('general',"Error: Invalid Artist ID");
+			$GLOBALS['error']->add_error('general',"Error: No such artist to merge with");
                         return false;
                 }
         } // merge
@@ -290,7 +293,125 @@ class Artist {
 	        require (conf('prefix') . "/templates/show_artist.inc");
 
 	} // show_albums
-
+	
+	/*!
+		@function get_similar_artists
+		@discussion returns an array of artist (id,name) arrays that are similar in name
+			All whitespace and special chars are ignored
+		@param extra arguments to normalize and compre, in that order
+		@return array of artist, each element is (id,name)
+	*/
+	function get_similar_artists ($n_rep_uml,$n_filter,$n_ignore,$c_mode,$c_count_w,$c_percent_w,$c_distance_l) {
+		//strip out just about everything, including whitespace, numbers and weird chars, and then
+		//lowercase it
+		$name = $this->normalize_name($this->name,$n_rep_uml,$n_filter,$n_ignore);
+		
+		//now for a bit of mysql query
+		$sql = "SELECT id, name FROM artist WHERE id != '" . sql_escape($this->id) . "'";
+		$query = mysql_query($sql, dbh());
+		//loop it
+		$similar_artists = array();
+		while ($r = mysql_fetch_assoc($query)) {
+			$artist_name = $this->normalize_name($r['name'],$n_rep_uml,$n_filter,$n_ignore);
+			//echo "'" . $r['name'] . "' => '" . $artist_name . "'<br/>\n";
+			if ($this->compare_loose($name,$artist_name,$c_mode,$c_count_w,$c_percent_w,$c_distance_l)) {
+				//echo "***MATCH***<br/>\n";
+				$similar_artists[] = array($r['id'],$r['name']);
+			}
+		}
+		return $similar_artists;		
+	} // get_similar_artists
+	
+	
+	/*!
+		@function normalize_name
+		@param artist name to normalize
+		@param $replace_umlaut wether to replace umlauts and others with the plain letter, default true
+		@param $filter what to filter out, defulat /[^a-z ]/
+		@param $ignore terms to ignore, default /\s(the|an?)\s/ (name is padded with whitespace beforehand)
+		@returns the normalized version of the given artist name, containing only letters and single spaces
+	*/
+	function normalize_name ($name,$replace_umlaut = NULL, $filter = NULL, $ignore = NULL) {
+		if (is_null($replace_umlaut)) $replace_umlaut = true;
+		if (is_null($filter)) $filter = "/[^a-z ]/";
+		if (is_null($ignore)) $ignore = "/\s(the|an?)\s/";
+		if ($replace_umlaut) {
+			//convert ümlauts, idea from http://php.net/manual/en/function.str-replace.php#50081
+			$umlauts = array("uml","acute","grave","cedil","ring","circ","tilde","lig","slash");
+			$name = str_replace($umlauts,"",htmlentities($name));
+			//now replace all &.; with .
+			$name = preg_replace("/&(.);/","\$1",$name);
+			//back to normal
+			$name = html_entity_decode($name);
+		}
+		//lowercase
+		$name = strtolower($name);
+		//now rip out all the special chars and spaces
+		$name = preg_replace($filter,"",$name);
+		//now certains terms can be dropped completely
+		//we have to add spaces on the sides though
+		$name = " " . $name . " ";
+		$name = preg_replace($ignore,"",$name);
+		//now single spaces
+		$name = preg_replace("/\s{2,}/"," ",$name);
+		//return
+		return trim($name);
+	} //normalize_name
+	
+	/*!
+		@function compare_loose
+		@discussion percent and count are ORed together
+		@param $name1 artist name
+		@param $name2 artist name to compare against
+		@param $mode the type of matching to perform, one of line or word, default word
+		@param $countwords WORD MODE number of words that must be shared to match, 0 to disable, default 0
+		@param $percentwords WORD MODE percentage of words that must be shared to match, 0 to disable, default 50%
+		@param $distance LETTER MODE max levenshtein distance to pass as a match
+		@return true if given params are similar, false if not
+	*/
+	function compare_loose ($name1,$name2,$mode = NULL,$countwords = NULL,$percentwords = NULL,$distance = NULL) {
+		if (is_null($mode)) $mode = "word";
+		if (is_null($countwords)) $countwords = 0;
+		if (is_null($percentwords)) $percentwords = 50;
+		if (is_null($distance)) $distance = 2;
+		
+		//echo "Compare '$name1' vs. '$name2'<br/>\n";
+		
+		$modes = array("line" => 0,"word" => 0,"letter" => 0);
+		$mode = (isset($modes[$mode]) ? $mode : "word");
+		switch ($mode) {
+			case "line":
+				//this is still relevant because of the normalize
+				return $name1 == $name2;
+			break;
+			case "word":
+				//echo "	COMPARE: Word mode<br/>\n";
+				//first, count the number of terms in name1, and then the number that also appear in name2
+				$words = explode(" ",$name1);
+				$num_words = count($words);
+				$num_words_shared = 0;
+				foreach ($words as $word) {
+					//echo "	Looking for word '$word'... ";
+					if (strpos($name2,$word) !== false) {
+						//echo "MATCHED";
+						$num_words_shared++;
+					} else {
+						//echo " Nope";
+					}
+					//echo "<br/>\n";
+				}
+				//now make the descision
+				return (
+					($countwords > 0 && $num_words_shared >= $countwords) || 
+					($percentwords > 0 && $num_words_shared > 0 && $num_words_shared/$num_words >= $percentwords/100)
+				);
+			break;
+			case "letter":
+				//simple
+				return levenshtein($name1,$name2) <= $distance;
+			break;
+		}
+	} //compare_loose
 	
 } //end of artist class
 
diff --git a/lib/duplicates.php b/lib/duplicates.php
index 75ca2a22..1d03a3da 100644
--- a/lib/duplicates.php
+++ b/lib/duplicates.php
@@ -42,8 +42,6 @@ function get_duplicate_songs($search_type) {
 	$sql = $sql." HAVING count(title) > 1";
 	$sql = $sql." ORDER BY ctitle";
 
-	//echo $sql."<BR>";
-
     $result = mysql_query($sql, dbh());
 
     $arr = array();
diff --git a/lib/ui.lib.php b/lib/ui.lib.php
index 44a67710..128ba533 100644
--- a/lib/ui.lib.php
+++ b/lib/ui.lib.php
@@ -880,7 +880,7 @@ function show_artist_pulldown ($artist_id,$select_name='artist') {
 		$artist->get_count();
 
 		if ( $artist_id == $r['id'] ) {
-			echo "\t<option value=\"" . $artist->id . "\" selected=\"selected\">". scrub_out($artist->name) . " (" . $artist->songs . ")</option>\n";
+			echo "\t<option value=\"" . $artist->id . "\" selected=\"selected\">". scrub_out($artist->name) . "</option>\n";
 		}
 		else {
 			echo "\t<option value=\"" . $artist->id . "\">". scrub_out($artist->name) ."</option>\n";