如何为 python 中的 Sql 语句提供动态输入

how to provide dynamic input for Sql statement in python

我正在尝试在 python 中执行动态 sql 查询。看起来像下面。

import json
from decimal import Decimal
import psycopg2

def initial_execution(param_name):
    time = get_timestamp(param_name)
    query = 'SELECT name, account_mgr, addr1, addr2, ap_email, bill_to_nbr, billto_only, city, controlling_nbr, controlling_only, country, cust_contact, email, load_create_date, load_update_date, nbr, owner, sales_rep, source_system, state, station, zip FROM public.customers WHERE billto_only = 'Y' OR controlling_only = 'Y' AND load_create_date >= \''+time+'\' OR load_update_date >= \''+time+'\''
    queryData = execute_db_query(query)

execute_db_query -> 此函数使用 pscopg2 在 redshift 中运行 我通过从用户输入中动态提取“时间”来执行此脚本。我遇到错误

[ERROR] Runtime.UserCodeSyntaxError: Syntax error in module
'src/customers': 
invalid syntax (customers.py, line 78) 
Traceback
(most recent call last):   File "/var/task/src/customers.py" 
Line 78 query = 'SELECT name,billto_only, controlling_only , load_create_date, load_update_date, nbr, owner, sales_rep,
 source_system, state, station, zip FROM public.test WHERE billto_only
 = 'Y' OR controlling_only = 'Y' AND load_create_date >= \''+time+'\' OR load_update_date >= \''+time+'\''

sql中的相同查询如下所示并正确执行

SELECT name,billto_only,  controlling_only,load_create_date, load_update_date, nbr, owner, sales_rep, source_system, state, station, zip FROM public.test
WHERE (billto_only = 'Y' OR controlling_only = 'Y') AND (load_create_date >= '2020-10-09 07:04:51' OR load_update_date >= '2020-10-09 07:04:51') limit 5;

可能不合时宜,但是 'Y' 会破坏您的字符串吗?

query = 'SELECT name,billto_only, controlling_only , load_create_date, load_update_date, nbr, owner, sales_rep, source_system, state, station, zip FROM public.test WHERE billto_only = \'Y\' OR controlling_only = \'Y\' AND load_create_date >= \''+time+'\' OR load_update_date >= \''+time+'\''

转义撇号

错误发生是因为 ' 符号,

"SELECT name, account_mgr, addr1, addr2, ap_email, bill_to_nbr, billto_only, city, controlling_nbr, controlling_only, country, cust_contact, email, load_create_date, load_update_date, nbr, owner, sales_rep, source_system, state, station, zip FROM public.customers WHERE billto_only = 'Y' OR controlling_only = 'Y' AND load_create_date >= '{date}' OR load_update_date >= '{date}'".format(date=time)

正确的做法:

query = """SELECT 
name, account_mgr, addr1, addr2, ap_email, bill_to_nbr, billto_only, city, controlling_nbr, 
controlling_only, country, cust_contact, email, load_create_date, load_update_date, nbr, 
owner, sales_rep, source_system, state, station, zip 
FROM 
    public.customers 
WHERE 
    billto_only = 'Y' OR controlling_only = 'Y' 
AND 
    load_create_date >= %(time_val)s OR load_update_date >= %(time_val)s
"""

queryData = execute_db_query(query, {"time_val": time)

以上确保正确引用传入的数据并防止 SQL 注入问题。 我也会避免使用像 time 这样的变量名。这与 time 模块以及 Postgres 中的 time 类型存在潜在冲突。