Question

I’ve got a wrapper I wrote around the sqlite3 module that lets me serialize access from multiple threads. It also lets me automatically migrate tables when I change their definition. I noticed when I drop a table and re-add it with more columns, I get an index out of range error. Something like this:

conn = sqlite3.connect("test.db", detect_types=sqlite3.PARSE_COLNAMES)
curs = conn.cursor()

curs.execute("CREATE TABLE test (derp TEXT);"); conn.commit()
curs.execute("INSERT INTO test (derp) VALUES ('deedle');"); conn.commit()
print curs.execute("SELECT * FROM test;").fetchall()
curs.execute("DROP TABLE test;"); conn.commit()
curs.execute("CREATE TABLE test (derp TEXT, val REAL);"); conn.commit()
curs.execute("INSERT INTO test (derp) VALUES ('deedle');"); conn.commit()
print curs.execute("SELECT * FROM test;").fetchall()

conn.close()

Will print this:

[(u'deedle',)] Traceback (most recent
call last):   File "test.py", line 23,
in <module>
    print curs.execute("SELECT * FROM test;").fetchall() IndexError: list
index out of range

When executing the second SELECT statement. Does anyone know why this is?

Answer 1

Well, it works just fine for me
[(u'deedle',)]
[(u'deedle', None)]