理解期间列表中的python类型

Question

我有一个 sql 查询字符串，如下所示:

intro text,
id int,
description varchar(50)

我正在尝试创建一个类型字符串，目的是找到与 sql 模式中定义的类型不匹配的文本片段。我从 sql 文本中提取类型的方式如下:

types = [re.sub('[^a-zA-Z]','',x.split()[1]) for x in schema] 
types = [re.sub('varchar',types.StringType,x) for x in types]
types = [re.sub('text',types.StringType,x) for x in types]
types = [re.sub('bigint',types.IntType,x) for x in types]
types = [re.sub('decimal',types.IntType,x) for x in types]

然而口译员提示说

types = [re.sub('varchar',types.StringTypes,x) for x in types]
AttributeError: 'list' object has no attribute 'StringTypes'

一个SSCCE

使用以下架构文件

intro text,
id int,
description varchar(50)

和代码(注意，按照下面 oscar 的建议修复，但现在有其他错误)

import csv
import sys
import re
import types

sch = open(sys.argv[1], "rb")

#---------------------------------   
# read schema
#---------------------------------     
with sch as f:
    schema = f.read().splitlines()

#---------------------------------    
# extract schema types
#---------------------------------  

foundtypes = [re.sub('[^a-zA-Z]','',x.split()[1]) for x in schema] 
foundtypes = [re.sub('varchar',str,x) for x in foundtypes]
foundtypes = [re.sub('text',str,x) for x in foundtypes]
foundtypes = [re.sub('int',int,x) for x in foundtypes]
foundtypes = [re.sub('bigint',int,x) for x in foundtypes]
foundtypes = [re.sub('decimal',int,x) for x in foundtypes]

print foundtypes

我正在使用 Python 2.7.5

谢谢

Answer 1

您将绑定(bind)(参见:variable shadowing)改写为 types module , 在这一行中:

types = [re.sub('[^a-zA-Z]','',x.split()[1]) for x in schema]

之后，types不再指向模块，而是指向一个列表。只需在所有作业中使用另一个名称即可:

my_types = [re.sub('[^a-zA-Z]','',x.split()[1]) for x in schema] 
my_types = [re.sub('varchar',types.StringType,x) for x in my_types]
my_types = [re.sub('text',types.StringType,x) for x in my_types]
my_types = [re.sub('bigint',types.IntType,x) for x in my_types]
my_types = [re.sub('decimal',types.IntType,x) for x in my_types]

更新

我认为您过度设计了解决方案，除了第一行，这不适合使用正则表达式。一个简单的 if-elif-else 就可以正常工作:

def transform(typestr):
    if typestr in ('varchar', 'text'):
        return types.StringType
    elif typestr in ('int', 'bigint', 'decimal'):
        return types.IntType
    else:
        return None

my_types = [re.sub(r'[^a-zA-Z]', '', x.split()[1]) for x in schema] 
[transform(x) for x in my_types]
=> [<type 'str'>, <type 'int'>, <type 'str'>]