I want to insert data into Oracle Table, where one row type-RAW(In python BYTES).
sql = f"INSERT /* APPEND */ INTO {table} ({columns}) VALUES ({values})"
ins_r = ', '.join(lst)
cur.execute(sql, ins_r)
This is my printed SQL:
INFO - INSERT /* APPEND */ INTO TR.ZZ_TEST (REGION_ID, REGION_NAME, CHANGE_DATE, MD5) VALUES (:REGION_ID, :REGION_NAME, :CHANGE_DATE, :MD5)
And this my data which I want to Insert:
['1', "'TEST'", "to_date('2021-09-28 18:48:23','YYYY-MM-DD HH24:MI:SS')", b'aWQ\x9b\xa6(\x17zj\xab\x97\x8e\x12uE4']
And I have the error:
ins_r = ', '.join(lst)
TypeError: sequence item 3: expected str instance, bytes found
And I wanted to convert bytes to str, but is not correct.
CodePudding user response:
The second argument to cur.execute()
should be the list of values, not a string made by joining them. So use
cur.execute(sql, lst)
CodePudding user response:
You can convert your parameter to list of tuples in order to be able insert multiple values along with using executemany
method rather than execute
as being more performant for bulk loads.
Convert the current to_date
conversion of the DB to datetime.datetime
(a type of datetime
module), and count the number commas within the columns
string in order to generate a bind variables list(values
-> in this case :0, :1, :2, :3
)
import cx_Oracle
import datetime
...
...
table = 'tr.zz_test'
columns = 'region_id,region_name,change_date,md5'
prm=[
(11,'TEST1',datetime.datetime(2021, 9, 29, 17, 28, 11),b'aWQ\x9b\xa6(\x17zj\xab\x97\x8e\x12uE4'),
(12,'TEST2',datetime.datetime(2021, 9, 28, 18, 48, 23),b'aWQ\x9b\xa5(\x17zj\xab\x97\x8e\x12uE2')
]
val=""
for i in range(0,columns.count(",") 1):
val =':'.join(' ' str(i)) ','
values=val.rstrip(",")
my_query = f"INSERT /* APPEND */ INTO {table}({columns}) VALUES({values})"
cur.executemany(my_query,prm)
con.commit()