mysql - Mysql 等价于 php metaphone 和 soundex

Question

我正在开发一个应用程序，在该应用程序中获取用户当前正在播放的歌曲标题，我们在 mysql 数据库中查看还有谁在播放类似的歌曲。因为同一首歌曲可能在每个人的手机上都有许多不同的标题，我们需要一种方法来有效地找到尽可能接近的结果。

我们现在使用的过程从表中获取所有歌曲，然后执行一次 foreach 并将结果集中的每个条目与用户的歌曲进行比较。这是我们使用的函数的一部分：

$all_results = $db->select($sql);//all db entries from the songs table

foreach ( $all_results as $u ) {
$toTest = strtolower( $u['last_song'] );
$toTest = preg_replace('/[^A-Za-z0-9]/', ' ', $toTest);
$score = 0; 
$score = $this->calSim( $user_last_song, $toTest, 0 ); //user last song is the current song being played by the user
if ( $score > 1 ) { //if at least there is some match then compare by another method
    $score = $this->calMetaphone($user_last_song, $toTest, $score);
}
if ( $score > 5 ) {
    //song matches.Push into a final array
}
 }

========================================

这里有 2 个自定义函数：

public function calMetaphone ( $x, $y, $pts ) {
     $x = metaphone( $x );
     $y = metaphone( $y );
     $pts = $this->calSim( $x, $y, $pts );
     return $pts;
}
public function calSim ( $x, $y, $pts ) {
     similar_text($x, $y, $sim);
     //echo "Similarity is :$sim<br>";
     if ( $sim >= 90 ) {
         $pts = $pts + 5;
     } else if ( $sim >= 80 ) {
         $pts = $pts + 4;
     }
     if ( $sim >= 70 ) {
         $pts = $pts + 3;
     }
     if ( $sim >= 60 ) {
         $pts = $pts + 2;
     } else {
         $pts = $pts + 0;
     }
     return $pts;
  }

但我知道这是一种非常糟糕的做法。如果数据库中有大量条目，可能需要永远一一比较它们。

谁能告诉我这里应该遵循的正确方法。

谢谢卡拉姆

Answer 1

我对此不以为然，但是当我需要一个类似的功能时，我发现了这个（虽然不记得在哪里）用于变音：-

DROP FUNCTION `func_Double_Metaphone`//
CREATE DEFINER=`aaaa`@`%` FUNCTION `func_Double_Metaphone`(st VARCHAR(55)) RETURNS varchar(128) CHARSET utf8
    NO SQL
BEGIN
    DECLARE length, first, last, pos, prevpos, is_slavo_germanic SMALLINT;
    DECLARE pri, sec VARCHAR(45) DEFAULT '';
    DECLARE ch CHAR(1);
                    SET first = 3;
    SET length = CHAR_LENGTH(st);
    SET last = first + length -1;
    SET st = CONCAT(REPEAT('-', first -1), UCASE(st), REPEAT(' ', 5));  SET is_slavo_germanic = (st LIKE '%W%' OR st LIKE '%K%' OR st LIKE '%CZ%');     SET pos = first;        IF SUBSTRING(st, first, 2) IN ('GN', 'KN', 'PN', 'WR', 'PS') THEN
        SET pos = pos + 1;
    END IF;
        IF SUBSTRING(st, first, 1) = 'X' THEN
        SET pri = 'S', sec = 'S', pos = pos  + 1;   END IF;
        WHILE pos <= last DO
            SET prevpos = pos;
        SET ch = SUBSTRING(st, pos, 1);         CASE
        WHEN ch IN ('A', 'E', 'I', 'O', 'U', 'Y') THEN
            IF pos = first THEN                 SET pri = CONCAT(pri, 'A'), sec = CONCAT(sec, 'A'), pos = pos  + 1;             ELSE
                SET pos = pos + 1;
            END IF;
        WHEN ch = 'B' THEN
                        IF SUBSTRING(st, pos+1, 1) = 'B' THEN
                SET pri = CONCAT(pri, 'P'), sec = CONCAT(sec, 'P'), pos = pos  + 2;             ELSE
                SET pri = CONCAT(pri, 'P'), sec = CONCAT(sec, 'P'), pos = pos  + 1;             END IF;
        WHEN ch = 'C' THEN
                        IF (pos > (first + 1) AND SUBSTRING(st, pos-2, 1) NOT IN ('A', 'E', 'I', 'O', 'U', 'Y') AND SUBSTRING(st, pos-1, 3) = 'ACH' AND
               (SUBSTRING(st, pos+2, 1) NOT IN ('I', 'E') OR SUBSTRING(st, pos-2, 6) IN ('BACHER', 'MACHER'))) THEN
                SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 2;                         ELSEIF pos = first AND SUBSTRING(st, first, 6) = 'CAESAR' THEN
                SET pri = CONCAT(pri, 'S'), sec = CONCAT(sec, 'S'), pos = pos  + 2;             ELSEIF SUBSTRING(st, pos, 4) = 'CHIA' THEN              SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 2;             ELSEIF SUBSTRING(st, pos, 2) = 'CH' THEN
                                IF pos > first AND SUBSTRING(st, pos, 4) = 'CHAE' THEN
                    SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'X'), pos = pos  + 2;                 ELSEIF pos = first AND (SUBSTRING(st, pos+1, 5) IN ('HARAC', 'HARIS') OR
                   SUBSTRING(st, pos+1, 3) IN ('HOR', 'HYM', 'HIA', 'HEM')) AND SUBSTRING(st, first, 5) != 'CHORE' THEN
                    SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 2;                                 ELSEIF SUBSTRING(st, first, 4) IN ('VAN ', 'VON ') OR SUBSTRING(st, first, 3) = 'SCH'
                   OR SUBSTRING(st, pos-2, 6) IN ('ORCHES', 'ARCHIT', 'ORCHID')
                   OR SUBSTRING(st, pos+2, 1) IN ('T', 'S')
                   OR ((SUBSTRING(st, pos-1, 1) IN ('A', 'O', 'U', 'E') OR pos = first)
                   AND SUBSTRING(st, pos+2, 1) IN ('L', 'R', 'N', 'M', 'B', 'H', 'F', 'V', 'W', ' ')) THEN
                    SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 2;                 ELSE
                    IF pos > first THEN
                        IF SUBSTRING(st, first, 2) = 'MC' THEN
                            SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 2;                         ELSE
                            SET pri = CONCAT(pri, 'X'), sec = CONCAT(sec, 'K'), pos = pos  + 2;                         END IF;
                    ELSE
                        SET pri = CONCAT(pri, 'X'), sec = CONCAT(sec, 'X'), pos = pos  + 2;                     END IF;
                END IF;
                        ELSEIF SUBSTRING(st, pos, 2) = 'CZ' AND SUBSTRING(st, pos-2, 4) != 'WICZ' THEN
                SET pri = CONCAT(pri, 'S'), sec = CONCAT(sec, 'X'), pos = pos  + 2;                         ELSEIF SUBSTRING(st, pos+1, 3) = 'CIA' THEN
                SET pri = CONCAT(pri, 'X'), sec = CONCAT(sec, 'X'), pos = pos  + 3;                         ELSEIF SUBSTRING(st, pos, 2) = 'CC' AND NOT (pos = (first +1) AND SUBSTRING(st, first, 1) = 'M') THEN
                                IF SUBSTRING(st, pos+2, 1) IN ('I', 'E', 'H') AND SUBSTRING(st, pos+2, 2) != 'HU' THEN
                                        IF (pos = first +1 AND SUBSTRING(st, first) = 'A') OR
                       SUBSTRING(st, pos-1, 5) IN ('UCCEE', 'UCCES') THEN
                        SET pri = CONCAT(pri, 'KS'), sec = CONCAT(sec, 'KS'), pos = pos  + 3;                                       ELSE
                        SET pri = CONCAT(pri, 'X'), sec = CONCAT(sec, 'X'), pos = pos  + 3;                     END IF;
                ELSE
                    SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 2;                 END IF;
            ELSEIF SUBSTRING(st, pos, 2) IN ('CK', 'CG', 'CQ') THEN
                SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 2;             ELSEIF SUBSTRING(st, pos, 2) IN ('CI', 'CE', 'CY') THEN
                                IF SUBSTRING(st, pos, 3) IN ('CIO', 'CIE', 'CIA') THEN
                    SET pri = CONCAT(pri, 'S'), sec = CONCAT(sec, 'X'), pos = pos  + 2;                 ELSE
                    SET pri = CONCAT(pri, 'S'), sec = CONCAT(sec, 'S'), pos = pos  + 2;                 END IF;
            ELSE 
                                IF SUBSTRING(st, pos+1, 2) IN (' C', ' Q', ' G') THEN
                    SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 3;                 ELSE
                    IF SUBSTRING(st, pos+1, 1) IN ('C', 'K', 'Q') AND SUBSTRING(st, pos+1, 2) NOT IN ('CE', 'CI') THEN
                        SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 2;                     ELSE                        SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 1;                     END IF;
                END IF;
            END IF;
                            WHEN ch = 'D' THEN
            IF SUBSTRING(st, pos, 2) = 'DG' THEN
                IF SUBSTRING(st, pos+2, 1) IN ('I', 'E', 'Y') THEN                  SET pri = CONCAT(pri, 'J'), sec = CONCAT(sec, 'J'), pos = pos  + 3;                 ELSE
                    SET pri = CONCAT(pri, 'TK'), sec = CONCAT(sec, 'TK'), pos = pos  + 2;               END IF;
            ELSEIF SUBSTRING(st, pos, 2) IN ('DT', 'DD') THEN
                SET pri = CONCAT(pri, 'T'), sec = CONCAT(sec, 'T'), pos = pos  + 2;             ELSE
                SET pri = CONCAT(pri, 'T'), sec = CONCAT(sec, 'T'), pos = pos  + 1;             END IF;
        WHEN ch = 'F' THEN
            IF SUBSTRING(st, pos+1, 1) = 'F' THEN
                SET pri = CONCAT(pri, 'F'), sec = CONCAT(sec, 'F'), pos = pos  + 2;             ELSE
                SET pri = CONCAT(pri, 'F'), sec = CONCAT(sec, 'F'), pos = pos  + 1;             END IF;
        WHEN ch = 'G' THEN
            IF SUBSTRING(st, pos+1, 1) = 'H' THEN
                IF (pos > first AND SUBSTRING(st, pos-1, 1) NOT IN ('A', 'E', 'I', 'O', 'U', 'Y')) 
                    OR ( pos = first AND SUBSTRING(st, pos+2, 1) != 'I') THEN
                    SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 2;                 ELSEIF pos = first AND SUBSTRING(st, pos+2, 1) = 'I' THEN
                     SET pri = CONCAT(pri, 'J'), sec = CONCAT(sec, 'J'), pos = pos  + 2;                                ELSEIF (pos > (first + 1) AND SUBSTRING(st, pos-2, 1) IN ('B', 'H', 'D') )
                   OR (pos > (first + 2) AND SUBSTRING(st, pos-3, 1) IN ('B', 'H', 'D') )
                   OR (pos > (first + 3) AND SUBSTRING(st, pos-4, 1) IN ('B', 'H') ) THEN
                    SET pos = pos + 2;              ELSE
                                        IF pos > (first + 2) AND SUBSTRING(st, pos-1, 1) = 'U'
                       AND SUBSTRING(st, pos-3, 1) IN ('C', 'G', 'L', 'R', 'T') THEN
                        SET pri = CONCAT(pri, 'F'), sec = CONCAT(sec, 'F'), pos = pos  + 2;                     ELSEIF pos > first AND SUBSTRING(st, pos-1, 1) != 'I' THEN
                        SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 2;           ELSE
              SET pos = pos + 1;
                    END IF;
                END IF;
            ELSEIF SUBSTRING(st, pos+1, 1) = 'N' THEN
                IF pos = (first +1) AND SUBSTRING(st, first, 1) IN ('A', 'E', 'I', 'O', 'U', 'Y') AND NOT is_slavo_germanic THEN
                    SET pri = CONCAT(pri, 'KN'), sec = CONCAT(sec, 'N'), pos = pos  + 2;                ELSE
                                        IF SUBSTRING(st, pos+2, 2) != 'EY' AND SUBSTRING(st, pos+1, 1) != 'Y'
                        AND NOT is_slavo_germanic THEN
                        SET pri = CONCAT(pri, 'N'), sec = CONCAT(sec, 'KN'), pos = pos  + 2;                    ELSE
                        SET pri = CONCAT(pri, 'KN'), sec = CONCAT(sec, 'KN'), pos = pos  + 2;                   END IF;
                END IF;
                        ELSEIF SUBSTRING(st, pos+1, 2) = 'LI' AND NOT is_slavo_germanic THEN
                SET pri = CONCAT(pri, 'KL'), sec = CONCAT(sec, 'L'), pos = pos  + 2;                        ELSEIF pos = first AND (SUBSTRING(st, pos+1, 1) = 'Y'
               OR SUBSTRING(st, pos+1, 2) IN ('ES', 'EP', 'EB', 'EL', 'EY', 'IB', 'IL', 'IN', 'IE', 'EI', 'ER')) THEN
                SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'J'), pos = pos  + 2;                         ELSEIF (SUBSTRING(st, pos+1, 2) = 'ER' OR SUBSTRING(st, pos+1, 1) = 'Y')
               AND SUBSTRING(st, first, 6) NOT IN ('DANGER', 'RANGER', 'MANGER')
               AND SUBSTRING(st, pos-1, 1) not IN ('E', 'I') AND SUBSTRING(st, pos-1, 3) NOT IN ('RGY', 'OGY') THEN
                SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'J'), pos = pos  + 2;                         ELSEIF SUBSTRING(st, pos+1, 1) IN ('E', 'I', 'Y') OR SUBSTRING(st, pos-1, 4) IN ('AGGI', 'OGGI') THEN
                                IF SUBSTRING(st, first, 4) IN ('VON ', 'VAN ') OR SUBSTRING(st, first, 3) = 'SCH'
                   OR SUBSTRING(st, pos+1, 2) = 'ET' THEN
                    SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 2;                 ELSE
                                        IF SUBSTRING(st, pos+1, 4) = 'IER ' THEN
                        SET pri = CONCAT(pri, 'J'), sec = CONCAT(sec, 'J'), pos = pos  + 2;                     ELSE
                        SET pri = CONCAT(pri, 'J'), sec = CONCAT(sec, 'K'), pos = pos  + 2;                     END IF;
                END IF;
            ELSEIF SUBSTRING(st, pos+1, 1) = 'G' THEN
                SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 2;             ELSE
                SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 1;             END IF;
        WHEN ch = 'H' THEN
                        IF (pos = first OR SUBSTRING(st, pos-1, 1) IN ('A', 'E', 'I', 'O', 'U', 'Y')) 
                AND SUBSTRING(st, pos+1, 1) IN ('A', 'E', 'I', 'O', 'U', 'Y') THEN
                SET pri = CONCAT(pri, 'H'), sec = CONCAT(sec, 'H'), pos = pos  + 2;             ELSE                SET pos = pos + 1;          END IF;
        WHEN ch = 'J' THEN
                        IF SUBSTRING(st, pos, 4) = 'JOSE' OR SUBSTRING(st, first, 4) = 'SAN ' THEN
                IF (pos = first AND SUBSTRING(st, pos+4, 1) = ' ') OR SUBSTRING(st, first, 4) = 'SAN ' THEN
                    SET pri = CONCAT(pri, 'H'), sec = CONCAT(sec, 'H');                 ELSE
                    SET pri = CONCAT(pri, 'J'), sec = CONCAT(sec, 'H');                 END IF;
            ELSEIF pos = first AND SUBSTRING(st, pos, 4) != 'JOSE' THEN
                SET pri = CONCAT(pri, 'J'), sec = CONCAT(sec, 'A');             ELSE
                                IF SUBSTRING(st, pos-1, 1) IN ('A', 'E', 'I', 'O', 'U', 'Y') AND NOT is_slavo_germanic
                   AND SUBSTRING(st, pos+1, 1) IN ('A', 'O') THEN
                    SET pri = CONCAT(pri, 'J'), sec = CONCAT(sec, 'H');                 ELSE
                    IF pos = last THEN
                        SET pri = CONCAT(pri, 'J');                     ELSE
                        IF SUBSTRING(st, pos+1, 1) not IN ('L', 'T', 'K', 'S', 'N', 'M', 'B', 'Z')
                           AND SUBSTRING(st, pos-1, 1) not IN ('S', 'K', 'L') THEN
                            SET pri = CONCAT(pri, 'J'), sec = CONCAT(sec, 'J');                         END IF;
                    END IF;
                END IF;
            END IF;
            IF SUBSTRING(st, pos+1, 1) = 'J' THEN
                SET pos = pos + 2;
            ELSE
                SET pos = pos + 1;
            END IF;
        WHEN ch = 'K' THEN
            IF SUBSTRING(st, pos+1, 1) = 'K' THEN
                SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 2;             ELSE
                SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 1;             END IF;
        WHEN ch = 'L' THEN
            IF SUBSTRING(st, pos+1, 1) = 'L' THEN
                                IF (pos = (last - 2) AND SUBSTRING(st, pos-1, 4) IN ('ILLO', 'ILLA', 'ALLE'))
                   OR ((SUBSTRING(st, last-1, 2) IN ('AS', 'OS') OR SUBSTRING(st, last) IN ('A', 'O'))
                   AND SUBSTRING(st, pos-1, 4) = 'ALLE') THEN
                    SET pri = CONCAT(pri, 'L'), pos = pos  + 2;                 ELSE
                    SET pri = CONCAT(pri, 'L'), sec = CONCAT(sec, 'L'), pos = pos  + 2;                 END IF;
            ELSE
                SET pri = CONCAT(pri, 'L'), sec = CONCAT(sec, 'L'), pos = pos  + 1;             END IF;
        WHEN ch = 'M' THEN
            IF SUBSTRING(st, pos-1, 3) = 'UMB'
               AND (pos + 1 = last OR SUBSTRING(st, pos+2, 2) = 'ER')
               OR SUBSTRING(st, pos+1, 1) = 'M' THEN
                SET pri = CONCAT(pri, 'M'), sec = CONCAT(sec, 'M'), pos = pos  + 2;             ELSE
                SET pri = CONCAT(pri, 'M'), sec = CONCAT(sec, 'M'), pos = pos  + 1;             END IF;
        WHEN ch = 'N' THEN
            IF SUBSTRING(st, pos+1, 1) = 'N' THEN
                SET pri = CONCAT(pri, 'N'), sec = CONCAT(sec, 'N'), pos = pos  + 2;             ELSE
                SET pri = CONCAT(pri, 'N'), sec = CONCAT(sec, 'N'), pos = pos  + 1;             END IF;
                            WHEN ch = 'P' THEN
            IF SUBSTRING(st, pos+1, 1) = 'H' THEN
                SET pri = CONCAT(pri, 'F'), sec = CONCAT(sec, 'F'), pos = pos  + 2;             ELSEIF SUBSTRING(st, pos+1, 1) IN ('P', 'B') THEN               SET pri = CONCAT(pri, 'P'), sec = CONCAT(sec, 'P'), pos = pos  + 2;             ELSE
                SET pri = CONCAT(pri, 'P'), sec = CONCAT(sec, 'P'), pos = pos  + 1;             END IF;
        WHEN ch = 'Q' THEN
            IF SUBSTRING(st, pos+1, 1) = 'Q' THEN
                SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 2;             ELSE
                SET pri = CONCAT(pri, 'K'), sec = CONCAT(sec, 'K'), pos = pos  + 1;             END IF;
        WHEN ch = 'R' THEN
                        IF pos = last AND not is_slavo_germanic
               AND SUBSTRING(st, pos-2, 2) = 'IE' AND SUBSTRING(st, pos-4, 2) NOT IN ('ME', 'MA') THEN
                SET sec = CONCAT(sec, 'R');             ELSE
                SET pri = CONCAT(pri, 'R'), sec = CONCAT(sec, 'R');             END IF;
            IF SUBSTRING(st, pos+1, 1) = 'R' THEN
                SET pos = pos + 2;
            ELSE
                SET pos = pos + 1;
            END IF;
        WHEN ch = 'S' THEN
                        IF SUBSTRING(st, pos-1, 3) IN ('ISL', 'YSL') THEN
                SET pos = pos + 1;
                        ELSEIF pos = first AND SUBSTRING(st, first, 5) = 'SUGAR' THEN
                SET pri = CONCAT(pri, 'X'), sec = CONCAT(sec, 'S'), pos = pos  + 1;             ELSEIF SUBSTRING(st, pos, 2) = 'SH' THEN
                                IF SUBSTRING(st, pos+1, 4) IN ('HEIM', 'HOEK', 'HOLM', 'HOLZ') THEN
                    SET pri = CONCAT(pri, 'S'), sec = CONCAT(sec, 'S'), pos = pos  + 2;                 ELSE
                    SET pri = CONCAT(pri, 'X'), sec = CONCAT(sec, 'X'), pos = pos  + 2;                 END IF;
                        ELSEIF SUBSTRING(st, pos, 3) IN ('SIO', 'SIA') OR SUBSTRING(st, pos, 4) = 'SIAN' THEN
                IF NOT is_slavo_germanic THEN
                    SET pri = CONCAT(pri, 'S'), sec = CONCAT(sec, 'X'), pos = pos  + 3;                 ELSE
                    SET pri = CONCAT(pri, 'S'), sec = CONCAT(sec, 'S'), pos = pos  + 3;                 END IF;
                                    ELSEIF (pos = first AND SUBSTRING(st, pos+1, 1) IN ('M', 'N', 'L', 'W')) OR SUBSTRING(st, pos+1, 1) = 'Z' THEN
                SET pri = CONCAT(pri, 'S'), sec = CONCAT(sec, 'X');                 IF SUBSTRING(st, pos+1, 1) = 'Z' THEN
                    SET pos = pos + 2;
                ELSE
                    SET pos = pos + 1;
                END IF;
            ELSEIF SUBSTRING(st, pos, 2) = 'SC' THEN
                                IF SUBSTRING(st, pos+2, 1) = 'H' THEN
                                        IF SUBSTRING(st, pos+3, 2) IN ('OO', 'ER', 'EN', 'UY', 'ED', 'EM') THEN
                                                IF SUBSTRING(st, pos+3, 2) IN ('ER', 'EN') THEN
                            SET pri = CONCAT(pri, 'X'), sec = CONCAT(sec, 'SK'), pos = pos  + 3;                        ELSE
                            SET pri = CONCAT(pri, 'SK'), sec = CONCAT(sec, 'SK'), pos = pos  + 3;                       END IF;
                    ELSE
                        IF pos = first AND SUBSTRING(st, first+3, 1) not IN ('A', 'E', 'I', 'O', 'U', 'Y') AND SUBSTRING(st, first+3, 1) != 'W' THEN
                            SET pri = CONCAT(pri, 'X'), sec = CONCAT(sec, 'S'), pos = pos  + 3;                         ELSE
                            SET pri = CONCAT(pri, 'X'), sec = CONCAT(sec, 'X'), pos = pos  + 3;                         END IF;
                    END IF;
                ELSEIF SUBSTRING(st, pos+2, 1) IN ('I', 'E', 'Y') THEN
                    SET pri = CONCAT(pri, 'S'), sec = CONCAT(sec, 'S'), pos = pos  + 3;                 ELSE
                    SET pri = CONCAT(pri, 'SK'), sec = CONCAT(sec, 'SK'), pos = pos  + 3;               END IF;
                        ELSEIF pos = last AND SUBSTRING(st, pos-2, 2) IN ('AI', 'OI') THEN
                SET sec = CONCAT(sec, 'S'), pos = pos  + 1;             ELSE
                SET pri = CONCAT(pri, 'S'), sec = CONCAT(sec, 'S');                 IF SUBSTRING(st, pos+1, 1) IN ('S', 'Z') THEN
                    SET pos = pos + 2;
                ELSE
                    SET pos = pos + 1;
                END IF;
            END IF;
        WHEN ch = 'T' THEN
            IF SUBSTRING(st, pos, 4) = 'TION' THEN
                SET pri = CONCAT(pri, 'X'), sec = CONCAT(sec, 'X'), pos = pos  + 3;             ELSEIF SUBSTRING(st, pos, 3) IN ('TIA', 'TCH') THEN
                SET pri = CONCAT(pri, 'X'), sec = CONCAT(sec, 'X'), pos = pos  + 3;             ELSEIF SUBSTRING(st, pos, 2) = 'TH' OR SUBSTRING(st, pos, 3) = 'TTH' THEN
                                IF SUBSTRING(st, pos+2, 2) IN ('OM', 'AM') OR SUBSTRING(st, first, 4) IN ('VON ', 'VAN ')
                   OR SUBSTRING(st, first, 3) = 'SCH' THEN
                    SET pri = CONCAT(pri, 'T'), sec = CONCAT(sec, 'T'), pos = pos  + 2;                 ELSE
                    SET pri = CONCAT(pri, '0'), sec = CONCAT(sec, 'T'), pos = pos  + 2;                 END IF;
            ELSEIF SUBSTRING(st, pos+1, 1) IN ('T', 'D') THEN
                SET pri = CONCAT(pri, 'T'), sec = CONCAT(sec, 'T'), pos = pos  + 2;             ELSE
                SET pri = CONCAT(pri, 'T'), sec = CONCAT(sec, 'T'), pos = pos  + 1;             END IF;
        WHEN ch = 'V' THEN
            IF SUBSTRING(st, pos+1, 1) = 'V' THEN
                SET pri = CONCAT(pri, 'F'), sec = CONCAT(sec, 'F'), pos = pos  + 2;             ELSE
                SET pri = CONCAT(pri, 'F'), sec = CONCAT(sec, 'F'), pos = pos  + 1;             END IF;
        WHEN ch = 'W' THEN
                        IF SUBSTRING(st, pos, 2) = 'WR' THEN
                SET pri = CONCAT(pri, 'R'), sec = CONCAT(sec, 'R'), pos = pos  + 2;             ELSEIF pos = first AND (SUBSTRING(st, pos+1, 1) IN ('A', 'E', 'I', 'O', 'U', 'Y')
                OR SUBSTRING(st, pos, 2) = 'WH') THEN
                                IF SUBSTRING(st, pos+1, 1) IN ('A', 'E', 'I', 'O', 'U', 'Y') THEN
                    SET pri = CONCAT(pri, 'A'), sec = CONCAT(sec, 'F'), pos = pos  + 1;                 ELSE
                    SET pri = CONCAT(pri, 'A'), sec = CONCAT(sec, 'A'), pos = pos  + 1;                 END IF;
                        ELSEIF (pos = last AND SUBSTRING(st, pos-1, 1) IN ('A', 'E', 'I', 'O', 'U', 'Y'))
               OR SUBSTRING(st, pos-1, 5) IN ('EWSKI', 'EWSKY', 'OWSKI', 'OWSKY')
               OR SUBSTRING(st, first, 3) = 'SCH' THEN
                SET sec = CONCAT(sec, 'F'), pos = pos  + 1;                                     ELSEIF SUBSTRING(st, pos, 4) IN ('WICZ', 'WITZ') THEN
                SET pri = CONCAT(pri, 'TS'), sec = CONCAT(sec, 'FX'), pos = pos  + 4;           ELSE                SET pos = pos + 1;
            END IF;
        WHEN ch = 'X' THEN
                        IF not(pos = last AND (SUBSTRING(st, pos-3, 3) IN ('IAU', 'EAU')
               OR SUBSTRING(st, pos-2, 2) IN ('AU', 'OU'))) THEN
                SET pri = CONCAT(pri, 'KS'), sec = CONCAT(sec, 'KS');           END IF;
            IF SUBSTRING(st, pos+1, 1) IN ('C', 'X') THEN
                SET pos = pos + 2;
            ELSE
                SET pos = pos + 1;
            END IF;
        WHEN ch = 'Z' THEN
                        IF SUBSTRING(st, pos+1, 1) = 'H' THEN
                SET pri = CONCAT(pri, 'J'), sec = CONCAT(sec, 'J'), pos = pos  + 1;             ELSEIF SUBSTRING(st, pos+1, 3) IN ('ZO', 'ZI', 'ZA')
               OR (is_slavo_germanic AND pos > first AND SUBSTRING(st, pos-1, 1) != 'T') THEN
                SET pri = CONCAT(pri, 'S'), sec = CONCAT(sec, 'TS');            ELSE
                SET pri = CONCAT(pri, 'S'), sec = CONCAT(sec, 'S');             END IF;
            IF SUBSTRING(st, pos+1, 1) = 'Z' THEN
                SET pos = pos + 2;
            ELSE
                SET pos = pos + 1;
            END IF;
        ELSE
            SET pos = pos + 1;      END CASE;
    IF pos = prevpos THEN
       SET pos = pos +1;
       SET pri = CONCAT(pri,'<didnt incr>');     END IF;
    END WHILE;
    IF pri != sec THEN
        SET pri = CONCAT(pri, ';', sec);
  END IF;
    RETURN (pri);
END

我还需要一个 levenshtein 函数并设法解决其中一个，但是性能如此（在比较多行的值时）读取行和使用 php 内置的 levenshtein 函数要快得多。