具有多列的 SQLAlchemy 相关更新? [英] SQLAlchemy Correlated Update with Multiple Columns?

问题描述

这个问题已经在 stackoverflow 上被问过多次，但他们似乎都已经超过一年了，所以我想我会再问一次，以防万一有更新.

This question has been asked multiple times on stackoverflow, but they all appeared to be over a year old so I figured I would ask again in case there has been an update.

correlated update 是一种更新语句，它根据另一个表中的值更新一个表中的所有行，同时将两个表链接在一起.

A correlated update is an update statement that updates all rows in one table based on values from another table, while linking the two tables together.

从 SQLAlchemy 文档，我们可以轻松进行相关更新，但仅限于单列:

From the SQLAlchemy docs, we can do a correlated update easily but only on a single column:

update(foo).values(bar=select([foobar.c.bar]).where(foobar.c.id == foo.c.id))

这转化为:

UPDATE foo
SET bar = (
    SELECT bar
    FROM foobar
    WHERE foobar.id = foo.id
) 

我们如何在 sqlalchemy 中使用多个列编写相关更新?例如:

How can we write a correlated update using more than one column in sqlalchemy? For example:

UPDATE foo
SET (bar, baz) = (
    SELECT bar, baz
    FROM foobar
    WHERE foobar.id = foo.id
) 

推荐答案

根据您的头像和描述，我猜您正在使用 Oracle.来自这个答案，如果您的加入导致 关键保留视图:

Based on your avatar and description I'm guessing you're using Oracle. From this answer one can device the following SQLAlchemy concotion, if your join results in a key preserved view:

stmt = select([foo.c.bar.label('foo_bar'),
               foo.c.baz.label('foo_baz'),
               foobar.c.bar.label('foobar_bar'),
               foobar.c.baz.label('foobar_baz')]).\
    where(foo.c.id == foobar.c.id)

update(stmt).values({stmt.c.foo_bar: stmt.c.foobar_bar,
                     stmt.c.foo_baz: stmt.c.foobar_baz})

产生以下 SQL:

UPDATE (SELECT foo.bar AS foo_bar,
               foo.baz AS foo_baz,
               foobar.bar AS foobar_bar,
               foobar.baz AS foobar_baz
        FROM foo, foobar
        WHERE foo.id = foobar.id)
SET foo_bar=foobar_bar, foo_baz=foobar_baz

标签很重要，因为您的表共享列名.

The labels are important since your tables share column names.

您还可以生成原始目标 SQL:

You can also produce your original target SQL:

from sqlalchemy import tuple_, select, exists

stmt = select([foobar.c.bar, foobar.c.baz]).where(foo.c.id == foobar.c.id)
foo.update().\
    values({tuple_(foo.c.bar, foo.c.baz).self_group(): stmt}).\
    where(exists(stmt))

self_group() 调用很重要，因为编译器似乎省略了元组周围的括号，在这种情况下会产生错误的语法.我添加了 WHERE 子句以避免更新 foo 没有匹配的 foobar 行:

UPDATE foo SET (bar, baz)=(SELECT foobar.bar, foobar.baz 
FROM foobar 
WHERE foo.id = foobar.id) WHERE EXISTS (SELECT foobar.bar, foobar.baz 
FROM foobar 
WHERE foo.id = foobar.id)

这篇关于具有多列的 SQLAlchemy 相关更新?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！