ms-access - MS Access 中的等效查询

Question

以下是在 ORACLE 中编写的查询，用于给出上一行，最后给出输出。 如何在 MS Access 中构建以下 ORACLE 查询，特别是 ROWNUM 部分..感谢您的输入

orcl> WITH d
  2       AS (SELECT ROWNUM rn,
  3                  dname,
  4                  deptno
  5           FROM   (SELECT dname,
  6                          deptno
  7                   FROM   dept
  8                   ORDER  BY dname)),
  9       e
 10       AS (SELECT ROWNUM + 1 rn,
 11                  dname,
 12                  deptno
 13           FROM   (SELECT dname,
 14                          deptno
 15                   FROM   dept
 16                   ORDER  BY dname))
 17  SELECT rn,
 18         d.dname,
 19         d.deptno this_row,
 20         e.deptno previous_row
 21  FROM   d
 22         left outer join e USING (rn)
 23  ORDER  BY dname;

    RN DNAME            THIS_ROW      previous_ROW

---

     1 ACCOUNTING             10
     2 OPERATIONS             40           10
     3 RESEARCH               20           40
     4 SALES                  30           20

Answer 1

对于您的查询使用的某些 Oracle 功能，Access SQL 没有直接对应项，因此您无法将其“翻译”到 Access。相反，您必须从头开始重写它。

您可以使用 DMax() 表达式返回“前一个”dname。

SELECT
    dname,
    deptno AS this_row,
    DMax("dname", "dept", "dname < '" & [dname] & "'")
        AS previous_dname
FROM dept;

该查询返回此结果集:

dname       this_row previous_dname
ACCOUNTING        10
OPERATIONS        40 ACCOUNTING
RESEARCH          20 OPERATIONS
SALES             30 RESEARCH

然后将该结果集左连接回 dept 表，该连接基于 previous_dname = dname。

SELECT
    d1.dname,
    d1.this_row,
    d2.deptno AS previous_row
FROM
    (
        SELECT
            dname,
            deptno AS this_row,
            DMax("dname", "dept", "dname < '" & [dname] & "'")
                AS previous_dname
        FROM dept
    ) AS d1
    LEFT JOIN dept AS d2
    ON d1.previous_dname = d2.dname
ORDER BY d1.dname;

我省略了 rn (ROWNUM) 列，因为使用此方法进行连接不需要它。但是，如果您出于其他原因仍然需要/想要它，可以使用 DCount() 表达式派生它。

SELECT
    d1.rn
    d1.dname,
    d1.this_row,
    d2.deptno AS previous_row
FROM
    (
        SELECT
            DCount("*", "dept", "dname <= '" & [dname] & "'")
                AS rn
            dname,
            deptno AS this_row,
            DMax("dname", "dept", "dname < '" & [dname] & "'")
                AS previous_dname
        FROM dept
    ) AS d1
    LEFT JOIN dept AS d2
    ON d1.previous_dname = d2.dname
ORDER BY d1.dname;