在sqlalchemy中按row_number过滤

Question

如何在以下查询中过滤row_number == 1:

query = session.query(Foo, func.row_number().over(
        partition_by=Foo.foo_field, order_by=desc(Foo.foo_date_time)).label("row_number"))

    query = query.filter(Foo.time_key <= time_key)
    # query = query.filter(row_number == 1)
    result = query.all()

Answer 1

我找到了:

    row_number_column = func.row_number().over(partition_by=Foo.foo_field, order_by=desc(Foo.foo_date_time)).label('row_number')
    query = self.session.query(Foo)
    query = query.filter(Foo.time_key <= time_key)
    query = query.add_column(row_number_column)
    query = query.from_self().filter(row_number_column == 1)

Answer 2

也可以使用子查询来完成：

sub_query = session.query(
Foo,
func.row_number().over(partition_by=Foo.foo_field, order_by=desc(Foo.foo_date_time)).label("row_number")
)
sub_query = sub_query.filter(Foo.time_key <= time_key).subquery()
query = session.query(sub_query).filter(sub_query.c.row_number == 1)

它基本上生成一个等效的 SQL 语句，将 row_number 添加为列。

编译声明。

sub_query = session.query(
Foo,
func.row_number().over(partition_by=Foo.foo_field, order_by=desc(Foo.foo_date_time)).label("row_number")
)
sub_query = sub_query.filter(Foo.time_key <= time_key).subquery()
query = session.query(sub_query).filter(sub_query.c.row_number == 1)

str(query.statement.compile())

将生成以下内容：

-- SubQuery function
SELECT anon_1.time_key, 
       anon_1.foo_field, 
       anon_1.foo_date_time, 
       anon_1.row_number 
FROM (
        SELECT  foo.time_key AS time_key,
                foo.foo_field AS foo_field, 
                foo.foo_date_time AS foo_date_time, 
                row_number() OVER (PARTITION BY foo.foo_field ORDER BY foo.foo_date_time DESC) AS row_number 
        FROM foo 
        WHERE foo.time_key <= %(time_key_1)s
    ) AS anon_1 
WHERE anon_1.row_number = %(row_number_1)s

更新：请记住，自 SQLAlchemy 1.4 起， Query.from_self方法已被弃用，根据官方文档，该方法将在 2.0 中删除

如果有人使用它，您可以查看此案例的迁移指南