类型错误：需要字符串或字节对象

Question

I am trying to insert CSV data into an Oracle table. But getting error as

TypeError: expecting string or bytes object

Code:

with open('file.csv', 'r') as srcfile:
reader = csv.reader(srcfile, dialect = 'fileProperties')
header = next(reader)
header = ", ".join(str(h) for h in header)
insert = 'insert into table (' + header + ') values ('
#print(insert) 

for row in reader:
    data = [insert + str(row).strip("[]") + ');']

cur.prepare(data)
cur.execute(data)

Error:

TypeError                                 Traceback (most recent call last)
<ipython-input-154-8814eee5c788> in <module>
  1 #cur.prepare(data)
----> 2 cur.execute(data)

TypeError: expecting string or bytes object

How can I fix the code?

Answer 1

Have resolved the issue.

• First issue is, list cannot be passed as a parameter while using executemany, but list works while using execute
  

• Second issue is, need to bind the source data while inserting. However instead of hard coding the binding variables, got the column names and passed the bind variables dynamically, so that any column added or removed in the file can be handled without modifying the code

  with open('G:\\Project\\ETL\\data\\CME.csv', 'r') as srcfile: reader = csv.reader(srcfile, dialect = 'fileProperties') header = next(reader) columns = ", ".join(str(h) for h in header) length = len(next(reader)) bind = [] data = [] for binds in range(1, length+1): bind.append(':' + str(binds)) insert = 'insert into ledger_source_python (' + columns + ') values (' + ", ".join(str(l) for l in bind) + ')' for row in reader: data.append(row) cur.executemany(insert, data) con.commit()