SQLAlchemy报错：期望传入字符串却得到Engine对象，求解决

我正在学习SQLAlchemy，查询Chinook数据库，按照文档和课程结构编写了查询表的代码：

### --- USING SQL ALCHEMY TO QUERY DATABASES --- ###

### --- IMPORT --- ###
from sqlalchemy import (
    create_engine, Table, Column, Float, ForeignKey, Integer, String, MetaData
)

### --- SETUP DB --- ###

# executing the instructions from our localhost "chinook" db
#(3/// means hosted locally)
db = create_engine("postgresql:///chinook")

# save a collection about our table objects and the data inside them
meta = MetaData(db)

### --- QUERY --- ###

#1. create variable for table we want to utilize
#(we feed in the subheadings we want, artist, id, name)
artist_table = Table(
    "Artist", meta,
    Column("ArtistId", Integer, primary_key=True),
    Column("Name", String)
)

#variable for album other tables would be made in the same way
album_table = Table(
    "Album", meta,
    Column("AlbumId", Integer, primary_key=True),
    Column("Title", String),
    Column("ArtistId", Integer, ForeignKey("artist_table.ArtistId"))
)
#seeing as artist id is a key that's linked from artist this is a foreign key for this table
#pretty much we want to set the table and column to point to

track_table = Table(
    "Track", meta,
    Column("TrackId", Integer, primary_key=True),
    Column("Name", String),
    Column("AlbumId", Integer, ForeignKey("album_table.AlbumId")),
    Column("MediaTypeId", Integer, primary_key=False),
    Column("GenreId", Integer, primary_key=False),
    Column("Composer", String),
    Column("Milliseconds", Integer),
    Column("Bytes", Integer),
    Column("UnitPrice", Float)
)

运行时出现如下报错：

Traceback (most recent call last):
  File "/workspace/PostGres_Chinook_Example/sql-expressions.py", line 15, in <module>
    meta = MetaData(db)
           ^^^^^^^^^^^^
  File "/workspace/.pip-modules/lib/python3.12/site-packages/sqlalchemy/sql/schema.py", line 5481, in __init__
    raise exc.ArgumentError(
sqlalchemy.exc.ArgumentError: expected schema argument to be a string, got <class 'sqlalchemy.engine.base.Engine'>.

排查与修复方向 #

1. 修正MetaData初始化参数
   SQLAlchemy 2.0+版本中，MetaData的第一个参数是schema名称（字符串类型），若要绑定数据库引擎，需使用bind参数。把原代码中的：

   meta = MetaData(db)
   

   修改为以下两种写法之一：

   # 写法1：通过bind参数绑定引擎
   meta = MetaData(bind=db)
   # 写法2：先初始化再绑定
   meta = MetaData()
   meta.bind = db
   

2. 修复ForeignKey引用错误
   代码中ForeignKey("artist_table.ArtistId")的写法错误，ForeignKey需要引用数据库中实际的表名和列名，而非Python变量名。正确写法应为：

   # Album表的ForeignKey修正
   Column("ArtistId", Integer, ForeignKey("Artist.ArtistId"))
   # Track表的ForeignKey修正
   Column("AlbumId", Integer, ForeignKey("Album.AlbumId"))
   

3. 版本兼容问题处理
   更换版本后出现其他错误，说明课程/文档使用的SQLAlchemy版本与你当前安装的版本存在API差异。建议：

   • 确认课程使用的SQLAlchemy版本号
   • 通过pip install sqlalchemy==<具体版本号>安装对应版本，避免跨版本API不兼容问题

内容的提问来源于stack exchange，提问作者user21629409