21.sharding-jdbc源码之SQL解析-词法分析

阿飞Javaer，转载请注明原创出处，谢谢！！

sharding-jdbc对SQL解析的源码主要在下图所示parsing模块中，由下图可知SQL解析主要分为两部分：lexer和parser。lexer就是本文需要分析的词法分析：

分析sharding-jdbc源码的词法分析之前，先大概说一下词法分析是干嘛的，后面理解起来就会更容易，例如对于SQL：“/! hello, afei /delete ignore from `t_user` where user_id=? “而言，词法分析结果如下：

SQL	com.dangdang.ddframe.rdb.sharding.parsing.lexer.token.Token
/*! hello, afei */	跳过
delete	(type=DELETE, literals=delete, endPosition=24)
ignore	(type=IGNORE, literals=ignore, endPosition=31)
from	(type=FROM, literals=from, endPosition=36)
`t_user`	(type=IDENTIFIER, literals=`t_user`, endPosition=45)
where	(type=WHERE, literals=where, endPosition=51)
user_id	(type=IDENTIFIER, literals=user_id, endPosition=59)
=	(type=EQ, literals==, endPosition=60)
?	(type=QUESTION, literals=?, endPosition=61)
	(type=END, literals=, endPosition=62)

SQL解析结果初探

前面分析SQL重写的时，其测试用例代码在SQLRewriteEngineTest.java中，列举其中一个用例源码如下，assertRewriteForLimit()方法中的selectStatement就是SQL解析的结果：

@Before
public void setUp() {
    // 定义sharding规则
    shardingRule = new ShardingRuleMockBuilder()
            // table_x表的主键列是id
            .addGenerateKeyColumn("table_x", "id")
            // table_x和table_y是绑定表关系
            .addBindingTable("table_y").build();
    selectStatement = new SelectStatement();
    tableTokens = new HashMap<>(1, 1);
    // 逻辑表table_x对应的实际表是table_1
    tableTokens.put("table_x", "table_1");
}

@Test
public void assertRewriteForLimit() {
    selectStatement.setLimit(new Limit(true));
    selectStatement.getLimit().setOffset(new LimitValue(2, -1));
    selectStatement.getLimit().setRowCount(new LimitValue(2, -1));
    selectStatement.getSqlTokens().add(new TableToken(17, "table_x"));
    selectStatement.getSqlTokens().add(new OffsetToken(33, 2));
    selectStatement.getSqlTokens().add(new RowCountToken(36, 2));
    // SelectStatement就是SQL解析的结果
    SQLRewriteEngine rewriteEngine = new SQLRewriteEngine(shardingRule, "SELECT x.id FROM table_x x LIMIT 2, 2", selectStatement);
    assertThat(rewriteEngine.rewrite(true).toSQL(tableTokens), is("SELECT x.id FROM table_1 x LIMIT 0, 4"));
}

SQL解析分析