Executemany() SQL-Update Statement with variables for column names

Question

I am really struggling with updating many rows in python using SAP HANA as my database and PyHDB for establishing the interface between both applications. Its working when I "hardcode" the columns, but I need to dynamically switch the columns by defining them inside of an array for example.

I am able to update in a hardcoded way the necessary columns by performing the following SQL-query:

sql = """UPDATE "ARE"."EMPLOYEES" SET "LIKELIHOOD_NO" = %s, "LIKELIHOOD_YES"= %s, "CLASS" = %s WHERE "EmployeeNumber" = %s;"""
cursor.executemany(sql, list)
connection.commit()

What I want to achieve is the following scenario:

 dynamic_columns = ["LIKELIHOOD_NO", "LIKELIHOOD_Yes"]
 sql = """UPDATE "ARE"."EMPLOYEES" SET dynamic_column = %s, "LIKELIHOOD_YES" = %s, "CLASS" = %s WHERE "EmployeeNumber" = %s;"""
 cursor.executemany(sql, list)
 connection.commit()

I am always getting the error that the relevant column / columns could not be found, but I cant figure out a way to solve this.

Asked By: Alex

||

Source

Answer 1

You can use normal string interpolation (.format()) to add the dynamic column name. You can see in the code here that pyHDB supports “numeric” paramstyle:

for col in ['LIKELIHOOD_YES', 'LIKELIHOOD_NO']:
    sql = ('UPDATE "ARE"."EMPLOYEES" SET "{some_col}" = :1, "CLASS" = :2 '
           'WHERE "EmployeeNumber" = :3;').format(some_col=col)
    cursor.executemany(sql, list_of_tuples)

This code will run for both columns 'LIKELIHOOD_YES' and 'LIKELIHOOD_NO'. Adapt it as you need. It would work with a list of tuples like this:

list_of_tuples = [
    (value1, class_1, employee_no_1),
    (value2, class_2, employee_no_2),
    (value3, class_3, employee_no_3),
    (value4, class_4, employee_no_4),
]

The code in your question seems to be using the ‘format’ paramstyle instead, but that doesn’t seem what pyHDB is using. See PEP 249 for more information on paramstyles.

Answered By: nosklo

Executemany() SQL-Update Statement with variables for column names

Question:

Answers: