Question

我正在使用以下参数集来创建动态sql查询并从python运行它。

target_column = 'connect_status'
table_name = 'public.main_table'

query_params = {'month_date' : '201801',
                'churn_date' : ['201802','201803','201804'],
                'disco_target' : '1-3 Month Disco'
                }


for churn in query_params['churn_date']:
    sql_data_sample = str("""select * from {1}
                                where dt = %(month_date)s 
                                and churn_date = %(churn)s
                                and {0} IN (%(disco_target)s,'No Disco') order by random() limit 5000
                                ;""").format(target_column,table_name)
    print (sql_data_sample)


    df_data_sample = pd.read_sql(sql_data_sample,con = cnxn,params = query_params)

但是这会导致DatabaseError：

Execution failed on sql 'select * from public.main_table
                                where dt = %(month_date)s 
                                and churn_date = %(churn)s
                                and connect_status IN (%(disco_target)s,'No Disco') order by random() limit 5000
                                ;': 'churn'

有人可以帮我解决这个问题吗？

Answer 1

问题是因为churn中没有query_params。将您的代码修改为

for churn in query_params['churn_date']:
    try:
       sql_data_sample = str("""select * from {1}
                            where dt = %(month_date)s 
                            and churn_date = %(churn)s
                            and {0} IN (%(disco_target)s,'No Disco') order by random() limit 5000
                            ;""").format(target_column,table_name)
        print (sql_data_sample)

        query_params['churn'] = churn
        df_data_sample = pd.read_sql(sql_data_sample,con = cnxn,params = query_params)
    except (Exception, psycopg2.DatabaseError) as error:
        print(error)

Answer 2

您的字符串格式表达式引用了query_params中的数据，但format_functions的参数中未引用query_params。看看您的month_date和disco_target的用法。

从Python创建动态SQL查询

2 个答案: