Question

我正在使用sqlalchemy，并希望从表中循环一组行，添加几个计算行并将它们插入到第二个表中。我可以将行对象转换为字典并使用：

def object_as_dict(obj):
    return {c.key: getattr(obj, c.key)
            for c in inspect(obj).mapper.column_attrs}

for row in session.query(source).limit(50):

    d = object_as_dict(row)
    del d['id'] # DON'T FORGET TO REMOVE ID FROM FIRST TABLE FROM DICT
    item = Filtered_Item()
    for k in d.keys():
        item[k] = d.get(k)

    item['calc1']=None
    item['calc2'] = None

但是我想知道是否有更直接的方法来处理sqlalchemy行对象？

Answer 1

如果你想在没有序列化的情况下这样做，你可以试试这个：

for row in session.query(source).limit(50):
  session.add(models.MyOtherModel(val=foo(row.val),val2=bar(row.bal2))

我个人认为序列化/反序列化数据没有任何问题。我每天使用类似的技术做大约2500万次写作：

def serialze_sqlalchemy_data(model, items):
    headers = sqlalchemy_headers(model)

    for record in items:
        row = {str(k): str(v) for k, v in
           zip(headers, [getattr(record, header.lower(), str(header.lower())) for header in headers])}
        yield row

def sqlalchemy_headers(model):
    try:
        _headers = [col.name for col in model.__mapper__.columns]
    except AttributeError:
        _headers = [col for col in model.__dict__.keys()
                    if isinstance(model.__dict__[col], InstrumentedAttribute)]
    return _headers

然后您可以执行以下操作：

def foo(row):
  row['val'] = 'taco'
  row['val2'] = 42
  return row

mymodel = models.MyModel
rows = session.query(mymodel)
rowdata = serialize_sqlalchemy_data(mymodel, rows)

table2 = models.MyModel2
session.bulk_insert_mappings(tables2, map(foo, rowdata))
session.commit()
session.close()

操纵sqlalchemy对象的最佳方法

1 个答案: