一、GenTableColumnMapper
这里是一个方法

<select id="selectDbTableColumnsByName" parameterType="String" resultMap="GenTableColumnResult">
    select t3.COLUMN_NAME                     as column_name,
           (CASE
                WHEN (t3.NULLABLE = 'N' and t4.CONSTRAINT_TYPE !='P') THEN '1'
                ELSE NULL END)
                                              as is_required,
           IF(t4.CONSTRAINT_TYPE = 'P', 1, 0) as is_pk,
           t3.COLUMN_ID                       as sort,
           t5.COMMENTS                        as column_comment,
           (  CASE
                  WHEN (t3.TYPE = 'INT' OR t3.TYPE = 'INTEGER' OR t3.TYPE = 'BIGINT' OR t3.TYPE = 'TINYINT' OR
                        t3.TYPE = 'SMALLINT') and t4.CONSTRAINT_TYPE = 'P' THEN '1'
                  ELSE '0' END  )                 AS is_increment,
           DATA_TYPE as DATA_TYPE
    from ((select COLUMN_NAME,
                  COLUMN_ID,
                  concat(DATA_TYPE, '(', DATA_LENGTH, ')') as DATA_TYPE,
                  DATA_TYPE                                as TYPE,
                  TABLE_NAME,
                  NULLABLE
           from SYS.USER_TAB_COLUMNS
           WHERE table_name = (#{tableName})) t3
             left join (select COMMENTS, COLUMN_NAME, TABLE_NAME from SYS.USER_COL_COMMENTS) t5
                       ON (t3.COLUMN_NAME = t5.COLUMN_NAME and t3.TABLE_NAME = t5.TABLE_NAME)
             left join
         (select t1.CONSTRAINT_TYPE, t1.OWNER, t1.TABLE_NAME, t2.CONSTRAINT_NAME, t2.COLUMN_NAME
          from (select CONSTRAINT_NAME, CONSTRAINT_TYPE, OWNER, TABLE_NAME from SYS.USER_CONSTRAINTS) t1
                   inner join (select CONSTRAINT_NAME, OWNER, TABLE_NAME, COLUMN_NAME from SYS.USER_CONS_COLUMNS) t2
                              ON (t1.TABLE_NAME = t2.TABLE_NAME and t1.CONSTRAINT_NAME = t2.CONSTRAINT_NAME)
          where t1.CONSTRAINT_TYPE = 'P') t4 ON (t3.COLUMN_NAME = t4.COLUMN_NAME and t3.TABLE_NAME = t4.TABLE_NAME))
    order by t3.COLUMN_ID
</select>

二、GenTableMapper
这里是三个方法

  <select id="selectDbTableList" parameterType="GenTable" resultMap="GenTableResult">
        select t1.TABLE_NAME as table_name, t2.COMMENTS as table_comment, NULL as create_time, NULL as update_time
        from SYS.USER_TABLES t1
        inner join SYS.USER_TAB_COMMENTS t2 ON t1.TABLE_NAME = t2.TABLE_NAME
        WHERE t1.TABLE_NAME NOT LIKE 'qrtz_%'
        AND t1.TABLE_NAME NOT LIKE 'gen_%'
        AND t1.TABLE_NAME NOT IN (select table_name as TABLE_NAME from gen_table)
        <if test="tableName != null and tableName != ''">
            AND lower(t1.TABLE_NAME) like lower(concat('%', #{tableName}, '%'))
        </if>
        <if test="tableComment != null and tableComment != ''">
            AND lower(t1.TABLE_NAME) like lower(concat('%', #{tableName}, '%'))
        </if>
    </select>

 <select id="selectDbTableListByNames" resultMap="GenTableResult">
        select t1.TABLE_NAME as table_name, t2.COMMENTS as table_comment, NULL as create_time, NULL as update_time
        from SYS.USER_TABLES t1
        inner join SYS.USER_TAB_COMMENTS t2 ON t1.TABLE_NAME = t2.TABLE_NAME
        WHERE t1.TABLE_NAME NOT LIKE 'qrtz_%'
        AND t1.TABLE_NAME NOT LIKE 'gen_%'
        and t1.TABLE_NAME in
        <foreach collection="array" item="name" open="(" separator="," close=")">
            #{name}
        </foreach>
    </select>

<select id="selectTableByName" parameterType="String" resultMap="GenTableResult">
    select t1.TABLE_NAME as table_name, t2.COMMENTS as table_comment, NULL as create_time, NULL as update_time
    from SYS.USER_TABLES t1
             inner join SYS.USER_TAB_COMMENTS t2 ON t1.TABLE_NAME = t2.TABLE_NAME
    where t2.COMMENTS <![CDATA[ <> ]]> ''
      and t1.TABLE_NAME = #{tableName}
</select>

系统表中有ALL、DBA、USER类型的，为了防止权限有问题，取的系统表全是取的是USER中的。
找到对应的SQL方法直接替换就行了

使用时注意传入的参数是tableName
有个select Database，这个没做处理，不过不影响使用，

identity这个自增的字段在系统表中没有找到，
我是用int类型并且是主键的情况是判断的，
这个主要是因为ID自增的问题，
如果有找到的话，可以给评论下，我优化下，

全部的数据来源都是从SYS的视图中获取的，关联关系的都是通过视图中的字段和数据对应找的，

现在我们这边使用的是UUID，所以不会出现问题。
暂时的话，在生成代码的时候，没发现什么问题，代码生成正常，一切使用正常。