从 varchar(500) 获取范围数据值

Question

寻找一些想法。

我正在运行 SQL Server 2014。我有一个 varchar(500) 列，其中包含我需要从每一行中提取的数据值，但是，它并不总是在行上的相同位置（有时相差 1或 2 个位置）所以我对子字符串有困难：

substring(Failure_Reason, charindex('Pressure received:', Failure_Reason + 181, 5) as 'Test'

使用这行代码，大多数时候我得到了正确的数据值，但有时我得到一个像'67.T'这样的值

varchar 专栏的内容是：

Settings received from Creep tester are outside expected values for pouch 3.            Pressure Duration received : 0. This value should be 32 +/- 2.          Pressure received : 198.5. This value should be between 202.77 and 197.24..

措辞始终相同，但数据值会不同

我只是试图解析“收到的压力：”之后的值，但由于它在某些行中可能是一两个位置，所以我有时会拉入“。”或者“T”，有没有办法捕捉“Pressure Received：”和“This”之间的值，所以我只得到数值？

谢谢！

Answer 1

以下内容应向您展示如何处理此问题：

1. 起始位置是字符串“Pressure received :”的位置 + 该字符串的长度 + 1
2. 结束位置是字符串“This” - 1 的位置，搜索此字符串的位置从步骤 1 中计算的位置开始
3. 获取上一步计算的两个值之间的子串
4. 修剪子串
5. 如果子字符串中的最后一个字符是“.”，则再次将其子字符串的长度小于当前长度 1

Answer 2

这里有一个解决方案：

 --==== Setup some test data
Declare @testData Table (Failure_Reason varchar(500));
 Insert Into @testData (Failure_Reason)
 Values ('Settings received from Creep tester are outside expected values for pouch 3.            Pressure Duration received : 0. This value should be 32 +/- 2.          Pressure received : 198.5. This value should be between 202.77 and 197.24..')
      , ('Settings received from Creep tester are outside expected values for pouch 3.            Pressure Duration received : 0. This value should be 32 +/- 2.          Pressure received : 98.5. This value should be between 202.77 and 197.24..')
      , ('Settings received from Creep tester are outside expected values for pouch 3.            Pressure Duration received : 0. This value should be 32 +/- 2.          Pressure received : 8.5. This value should be between 202.77 and 197.24..');

 --==== Solution
 Select *
      , PressureReceived = substring(td.Failure_Reason, p1.pos, p2.pos - p1.pos - 1)
   From @testData                                                                         As td
  Cross Apply (Values (charindex('Pressure received :', td.Failure_Reason, 1) + 20))      As p1(pos)
  Cross Apply (Values (charindex('. This', td.Failure_Reason, p1.pos) + 1))               As p2(pos);