Python 还是 SQL？ 从列中拆分数据以填充新列

Question

I have a column in SQL Server with data that I want to split out delineated by "/" to populate into a new column. I'm wondering if there's a way to do it with SQL Server or if it'd be easier to use Python.

This is a sample of the data now:

Contract_ID	Contract_Quote_Number	Contract_SO_PO
1469	COL 386986 / SO 590685	null
1471	COL 387554 / SO 590613	null
1472	COL 387527 / SO 590650	null
1473	COL 387638 / SO 590658	null

This is what I want it to look like:

Contract_ID	Contract_Quote_Number	Contract_SO_PO
1469	COL 386986	SO 590685
1471	COL 387554	SO 590613
1472	COL 387527	SO 590650
1473	COL 387638	SO 590658

So far in Python I've been able to pull the data into a CSV file and have sort of figured out how to split it. The problem I'm running into is that once I have it split properly I'll then have to re-write the csv data to SQL Server. I don't know if that's possible.

My python code so far (I'm writing to an empty csv file because as far as I know it's hard to overwrite a column in csv):

import csv
'''getting the col info'''
with open("C:\wamp64\www\SO_PO_Query.csv") as infile:
    reader = csv.reader(infile) # Create a new reader
    next(reader) # Skip the first row
    col = [row[1].split("/")[0] for row in reader]

print(col)

'''
writing 'col' to empty csv file

This works but it writes a space between each value
'''
rows = zip(col)
with open("C:\wamp64\www\empty.csv", "w") as f:
    writer = csv.writer(f)
    for row in rows:
        writer.writerow(row)

Answer 1

T-SQL isn't great at string manipulation, and Python (or any procedural language) is generally better at this. But a single split is fairly simple.

SELECT
  t.Contract_ID,
  Contract_Quote_Number = ISNULL(LEFT(t.Contract_Quote_Number, v.slash - 1), t.Contract_Quote_Number),
  Contract_SO_PO = SUBSTRING(t.Contract_Quote_Number, v.slash + 3, LEN(t.Contract_Quote_Number))
FROM YourTable t
CROSS APPLY (VALUES (
    NULLIF(CHARINDEX(' / ', t.Contract_Quote_Number), 0)
) ) v(slash);

• NULLIF ensures the query does not fail if the value is not found

Answer 2

One more method.

SQL

DECLARE @tbl TABLE (Contract_ID int primary key, Contract_Quote_Number VARCHAR(50));
INSERT INTO @tbl (Contract_ID, Contract_Quote_Number) VALUES
(1469, 'COL 386986 / SO 590685'),
(1471, 'COL 387554 / SO 590613'),
(1472, 'COL 387527 / SO 590650'),
(1473, 'COL 387638 / SO 590658');

SELECT Contract_ID
    , Contract_Quote_Number = JSON_VALUE(S,'$[0]')
    , Contract_SO_PO = JSON_VALUE(S,'$[1]')
FROM @tbl
CROSS APPLY (VALUES ('["' + REPLACE(Contract_Quote_Number, ' / ', '","') + '"]')) AS B(S);

Output

+-------------+-----------------------+----------------+
| Contract_ID | Contract_Quote_Number | Contract_SO_PO |
+-------------+-----------------------+----------------+
|        1469 | COL 386986            | SO 590685      |
|        1471 | COL 387554            | SO 590613      |
|        1472 | COL 387527            | SO 590650      |
|        1473 | COL 387638            | SO 590658      |
+-------------+-----------------------+----------------+

Answer 3

Quick and easy with charindex:

select * 
 , left(Contract_Quote_Number
          , charindex(Contract_Quote_Number,'/')-1) 
 , right(Contract_Quote_Number
          , len(Contract_Quote_Number) -  charindex(Contract_Quote_Number,'/') - 1 ) from table