PHPMyAdmin中使用sql-parser進行SQL文法分析

寫SQL對於PHP來說是經常的事，那麼如何對SQL語句進行分析來最佳化SQL呢？sql-parser就提供了這樣的功能，讓我們一起來看看吧。

phpMyAdmin是一款基於Web端啟動並執行開來源資料庫管理工具，支援管理MySQL和MariaDB兩種資料庫。 phpMyAdmin的程式主要使用php和javascript開發，它的安裝使用都比較簡單而且已有很多相關介紹不再重複，今天要介紹的是源碼中的一個核心組件sql-parser。

sql-parser簡介

sql-parser組件的主要用途是對SQL語句進行詞法分析、文法分析，繼而可以實現對SQL語句的解構、加工、替換、再組裝等需求，另外也可以對SQL進行highlight等處理。sql-parser由純PHP語言實現，同時也是整個phpMyAdmin源碼中為數不多的代碼架構比較清晰且符合當前PHP界PSR標準規範的模組。

sql-parser組件安裝

需事先安裝好php，git用戶端，以及composer php包管理工具

margin@margin-MB1:~/tmp$ sudo git clone https://github.com/phpmyadmin/sql-parser.gitmargin@margin-MB1:~/tmp$ cd sql-parser && sudo composer install

組件安裝完畢，下面介紹具體的調用

解析普通語句

require_once '../sql-parser/vendor/autoload.php';use SqlParser\Parser;$query = 'SELECT * FROM t1 LEFT JOIN (t2, t3, t4) '    . 'ON (t2.a=t1.a AND t3.b=t1.b AND t4.c=t1.c)';$parser = new Parser($query);$stmt = $parser->statements[0];echo json_encode($stmt);

運行結果中$parser變數是一個大對象，裡面儲存有對$query語句的詞法分析結果$query->list，文法分析結果$query-statements，以及錯誤資訊等。
其中$query-statements的結構如下：

{"expr":[{"database":null,"table":null,"column":null,"expr":"*","alias":null,"function":null,"subquery":null}],"from":[{"database":null,"table":"t1","column":null,"expr":"t1","alias":null,"function":null,"subquery":null}],"partition":null,"where":null,"group":null,"having":null,"order":null,"limit":null,"procedure":null,"into":null,"join":[{"type":"LEFT","expr":{"database":null,"table":null,"column":null,"expr":"(t2, t3, t4)","alias":null,"function":null,"subquery":null},"on":[{"identifiers":["t2","a","t1"],"isOperator":false,"expr":"(t2.a=t1.a"},{"identifiers":[],"isOperator":true,"expr":"AND"},{"identifiers":["t3","b","t1"],"isOperator":false,"expr":"t3.b=t1.b"},{"identifiers":[],"isOperator":true,"expr":"AND"},{"identifiers":["t4","c","t1"],"isOperator":false,"expr":"t4.c=t1.c)"}],"using":null}],"union":[],"options":{"options":[]},"first":0,"last":50}

解析事務

require_once '../sql-parser/vendor/autoload.php';use SqlParser\Parser;$query = 'START TRANSACTION;' .    'SELECT @A:=SUM(salary) FROM table1 WHERE type=1;' .    'UPDATE table2 SET summary=@A WHERE type=1;' .    'COMMIT;';$parser = new Parser($query);$stmt = $parser->statements[0];echo json_encode($stmt);

輸出結果：

{"type":1,"statements":[{"expr":[{"database":null,"table":null,"column":null,"expr":"@A:=SUM(salary)","alias":null,"function":"SUM","subquery":null}],"from":[{"database":null,"table":"table1","column":null,"expr":"table1","alias":null,"function":null,"subquery":null}],"partition":null,"where":[{"identifiers":["type"],"isOperator":false,"expr":"type=1"}],"group":null,"having":null,"order":null,"limit":null,"procedure":null,"into":null,"join":null,"union":[],"options":{"options":[]},"first":1,"last":19},{"tables":[{"database":null,"table":"table2","column":null,"expr":"table2","alias":null,"function":null,"subquery":null}],"set":[{"column":"summary","value":"@A"}],"where":[{"identifiers":["type"],"isOperator":false,"expr":"type=1"}],"order":null,"limit":null,"options":{"options":[]},"first":20,"last":35}],"end":{"type":2,"statements":null,"end":null,"options":{"options":{"1":"COMMIT"}},"first":36,"last":37},"options":{"options":{"1":"START TRANSACTION"}},"first":0,"last":0}

除以上兩種語句之外，sql-parser還支援解析預存程序等幾乎所有MySQL文法，不再一一舉例。下面是其SQL構造器的用法舉例。

組裝SQL語句

組裝select語句：

require_once '../sql-parser/vendor/autoload.php';use SqlParser\Components\OptionsArray;use SqlParser\Components\Expression;use SqlParser\Components\Condition;use SqlParser\Components\Limit;use SqlParser\Statements\SelectStatement;$stmt = new SelectStatement();$stmt->options = new OptionsArray(array('DISTINCT'));$stmt->expr[] = new Expression('sakila', 'film', 'film_id', 'fid');$stmt->expr[] = new Expression('COUNT(film_id)');$stmt->from[] = new Expression('', 'film', '');$stmt->from[] = new Expression('', 'actor', '');$stmt->where[] = new Condition('film_id > 10');$stmt->where[] = new Condition('OR');$stmt->where[] = new Condition('actor.age > 25');$stmt->limit = new Limit(1, 10);var_dump($stmt->build());

輸出結果：

margin@margin-MB1:~/code/parserTest$ php build.php string(137) "SELECT DISTINCT `sakila`.`film`.`film_id` AS `fid`, COUNT(film_id) FROM `film`, `actor` WHERE film_id > 10 OR actor.age > 25 LIMIT 10, 1 "

組裝觸發器語句：

require_once '../sql-parser/vendor/autoload.php';use SqlParser\Components\Expression;use SqlParser\Components\OptionsArray;use SqlParser\Statements\CreateStatement;$stmt = new CreateStatement();$stmt->options = new OptionsArray(array('TRIGGER'));$stmt->name = new Expression('ins_sum');$stmt->entityOptions = new OptionsArray(array('BEFORE', 'INSERT'));$stmt->table = new Expression('account');$stmt->body = 'SET @sum = @sum + NEW.amount';var_dump($stmt->build());

輸出結果：

margin@margin-MB1:~/code/parserTest$ php build.php string(89) "CREATE TRIGGER ins_sum BEFORE INSERT ON account FOR EACH ROW SET @sum = @sum + NEW.amount"

SQL再加工

多條語句一起加工處理：

require_once '../sql-parser/vendor/autoload.php';use SqlParser\Parser;use SqlParser\Components\Expression;$query  = <<<str id="3" from="" change="" $statement_0="$parser-" 處理第一條語句="" parser($query);="" $parser="new" str;="" where="" tbl3="" *="" select="" null;="" not="" unsigned="" )="" 10="" int(="" `field_2`="" `field_1`="" `tbl`="" table="" alter="" auto_increment;="" null="" `uid`="">statements[0];$statement_0->table  = new Expression(    'db2', 'tb2', '');var_dump($statement_0->build());//處理第二條語句$statement_1 = $parser->statements[1];$statement_1->table  = new Expression(    'db3', 'tb3', '');var_dump($statement_1->build());</str>

輸出結果：

margin@margin-MB1:~/code/parserTest$ php build.php string(85) "ALTER TABLE `db2`.`tb2` CHANGE `uid` `uid` INT( 10 ) UNSIGNED NOT NULL AUTO_INCREMENT"string(78) "ALTER TABLE `db3`.`tb3` CHANGE `field_1` `field_2` INT( 10 ) UNSIGNED NOT NULL"

以上是sql-parser組件一些基本的用法樣本，phpMyAdmin的sql-parser組件功能比較豐富和完備，本文限了篇幅不能詳盡，有興趣的讀者可以通過閱讀源碼來瞭解更多進階的用法。

相關推薦：

關於MySQL 最佳化的一些操作

mysql 最佳化（3）聚簇索引和非聚簇索引

MySql Sql 最佳化技巧的圖文代碼詳細介紹