使用变量列名称更新 Table 的存储过程

Question

我正在尝试在 plpgsql 中编写存储过程来简单地更新 table 中的整数值。我有一个 table（称为条件），其中包含四列整数类型，称为结构、缺失、修复和化妆品，它们接受 0、1、0r 2（请参阅下面的 table）。

 id | factor_id | structure | missing | repair | cosmetic 
----+-----------+-----------+---------+--------+----------
  2 |         4 |         0 |       0 |      1 |        2
  3 |         5 |         0 |       0 |      0 |        0
  4 |         6 |         0 |       0 |      0 |        1
  5 |         7 |         1 |       0 |      0 |        1
  6 |         8 |         1 |       0 |      0 |        1

我想编写一个更新程序来处理这三列，因此我需要运行一些动态 sql。这是更新声明：

UPDATE conditions SET (structure/missing/repair/cosmetic) = 1 WHERE id = 2 AND factor_id = 4;

根据我看到的帖子，我用EXECUTE格式函数写了这段代码：

CREATE OR REPLACE PROCEDURE spup_conditions
(
    _col VARCHAR,
    _parcel INT,
    _factor INT,
    _value INT
)
AS $$
DECLARE
    _colval text := _col;
BEGIN
    EXECUTE
        format('UPDATE conditions SET %I =  WHERE parcel_id =  AND
        factor_id = ', _colval)
    USING _value, _parcel, _factor;
    COMMIT;
END;
$$ LANGUAGE plpgsql;

该过程被 PostgreSQL 接受，但是当我调用它时

CALL spup_conditions('cosmetic', 1, 2, 4)

PostgreSQL 说调用了该过程，但是当我查看 table 时，该值没有改变。感谢您的帮助。

更新这是我修改后的 table 和程序以及我添加的另外两个操作

 id | parcel_id | factor_id | visible | structure | missing | repair | cosmetic | status | completed 
----+-----------+-----------+---------+-----------+---------+--------+----------+--------+-----------
  1 |        71 |         3 | f       |         0 |       0 |      1 |        0 | f      | 
  2 |        71 |         4 | f       |         0 |       0 |      1 |        2 |        | 
  3 |        71 |         5 | f       |         0 |       0 |      0 |        0 |        | 
  4 |        71 |         6 | f       |         0 |       0 |      0 |        1 |        | 
  5 |        71 |         7 | f       |         1 |       0 |      0 |        1 |        | 
  6 |        71 |         8 | f       |         1 |       0 |      0 |        1 |        | 
  7 |        71 |         9 | f       |         0 |       0 |      0 |        1 |        | 
  8 |        71 |        10 | f       |         0 |       0 |      1 |        2 |        | 
  9 |        71 |        11 | f       |         0 |       0 |      0 |        1 |        | 
 10 |        71 |        12 | t       |         0 |       0 |      0 |        0 |        | 
 11 |        71 |        13 | f       |         0 |       0 |      0 |        1 |        | 
 12 |        71 |        14 | f       |         0 |       0 |      0 |        1 |        | 
 13 |        71 |        15 | f       |         0 |       0 |      0 |        0 |        | 
 14 |        71 |        16 | f       |         0 |       0 |      0 |        1 |        | 

CREATE OR REPLACE PROCEDURE blight.spup_conditions
(
    _col text,
    _value INT,
    _parcel INT,
    _factor INT

)
AS $$
DECLARE
    _colval text := _col;
BEGIN
    EXECUTE
    format('UPDATE blight.conditions SET %I = 
        WHERE parcel_id =  AND factor_id = ', _colval)
    USING _value, _parcel, _factor;
    EXECUTE
    format('UPDATE blight.conditions SET status = FALSE
        WHERE parcel_id =  AND factor_id = ')
    USING _parcel, _factor;
    EXECUTE
    format('update blight.conditions set completed = NULL
        WHERE parcel_id =  AND factor_id = ')
    USING _parcel, _factor;
    COMMIT;
END;
$$ LANGUAGE plpgsql;

Answer 1

如果您添加参数名称，您将看到您的程序正在尝试执行以下更新

CALL spup_conditions('cosmetic', 1, 2, 4);

CALL spup_conditions(
        _col => 'cosmetic',
        _parcel => 1,
        _factor => 2,
        _value => 4);

    UPDATE conditions
       SET cosmetic = 4
     WHERE parcel_id = 1
       AND factor_id = 2;

您的 table 中没有符合这些条件的行。

Answer 2

您的过程和调用中有 2 个错误。

where 子句没有 parcel_id 列 - 更改为 id

format('UPDATE conditions_so SET %I =  WHERE id =  AND factor_id = ', _colval)

对于您的测试数据，没有 id=1 的行

call spup_conditions('cosmetic', 2, 4, 4);

使用变量列名称更新 Table 的存储过程

Stored Procedure to Update Table with a Variable Column Name

postgresql

stored-procedures

dynamic-sql

sql-update