sql - 如何根据上一行的值对sql结果进行排序？

Question

我正在尝试按结果集列中的值对 SQL 数据选择进行排序。数据如下:

(此数据排序不正确，仅作为示例)

ID      projectID   testName                                objectBefore    objectAfter
=======================================================================================
13147   280         CDM-710 Generic TP-0000120 TOC~~@~~     -1              13148
1145    280         3.2 Quadrature/Carrier Null 25 Deg C    4940            1146
1146    280         3.2 Quadrature/Carrier Null 0 Deg C     1145            1147
1147    280         3.3 External Frequency Reference        1146            1148
1148    280         3.4 Phase Noise 50 Deg C                1147            1149
1149    280         3.4 Phase Noise 25 Deg C                1148            1150
1150    280         3.4 Phase Noise 0 Deg C                 1149            1151
1151    280         3.5 Output Spurious 50 Deg C            1150            1152
1152    280         3.5 Output Spurious 25 Deg C            1151            1153
1153    280         3.5 Output Spurious 0 Deg C             1152            1154
............
18196   280         IP Regression Suite                     18195           -1

数据的顺序基于 objectBefore 和 objectAfter 列。第一行始终是当 objectBefore = -1 时，最后一行将是当 objectAfter = -1 时。在上面的示例中，第二行的 ID 为 13148，因为这就是第 1 行 objectAfter 所等于的内容。有没有办法编写一个以这种方式对数据进行排序的查询？

Answer 1

这实际上是对链表进行排序:

WITH SortedList (Id, objectBefore , projectID, testName, Level)
AS
(
  SELECT Id, objectBefore , projectID, testName, 0 as Level
    FROM YourTable
   WHERE objectBefore = -1
  UNION ALL
  SELECT ll.Id, ll.objectBefore , ll.projectID, ll.testName, Level+1 as Level
    FROM YourTable ll
   INNER JOIN SortedList as s
      ON ll.objectBefore = s.Id
)

SELECT Id, objectBefore , projectID, testName
  FROM SortedList
 ORDER BY Level

您可以在 this post 中找到更多详细信息