Removing u' from result of cursor.fetchall()

Question

I'm using python 2.7 with jaydebeapi and informix jdbc driver (ifxjdbc.jar) to execute a select on a table and all the fields have u'some_text' Here's my code

curs = conn.cursor()
curs.execute("SELECT * FROM table1")
res = curs.fetchall()
print res

using

res_final=[str(x) for x in res[0]]
print res_final

converts the first line in the result correctly. Any idea how can I convert all the selected rows? I had the same problem with sqlite3 but adding

conn.text_factory=str

solved my problem. I can't find something that works in the same way for jaydebeapi.

Why do you need to convert to bytestrings **at all**? The `u''` prefix is just a *type indicator*. — Martijn Pieters, Dec 06 '14 at 18:50
I was trying to pass the rows resulting from the select to a combobox in PyQT and also in a file. — albgz, Dec 06 '14 at 19:04

score 3 · Accepted Answer · answered Dec 06 '14 at 18:55

If you absolutely need to encode Unicode values to bytestrings, try to do so as late as possible, e.g. when pushing the values somewhere that can only handle bytes, like a file or network socket.

You can encode all columns in all rows with a nested list comprehension:

res = [col.encode('utf8') if isinstance(col, unicode) else col for col in row]
       for row in curs.fetchall()]

which only encodes values that are actually Unicode strings, in a manner that'll work with all Unicode codepoints.

If however all you are concerned about is the u'' prefix, then there is absolutely no need to encode. Don't confuse a type indicator in container representations for actual data.

Removing u' from result of cursor.fetchall()

1 Answers1