idiom for inserting multiple rows with auto-increment IDs

Hi all,
I'm new to pysqlite, and trying to figure out the most efficient an
appropriate way to insert multiple rows into a table with auto-increment
IDs, while also saving those new IDs.

newitems = [ list of new items ]
for row in newitems:
  cur.execute("insert into table (id,column1,column2,column3) values
(NULL,?,?,?)",
              (row.column1, row.column2, row.column3))
  row.id = cur.lastrowid

This seems pretty awkward though... I have also tried this:

from itertools import izip, count
newitems = [ list of new items ]
max, = con.execute("select max(id)+1 from table")
start = max[0] or 0
con.executemany("insert into table (id,column1,column2,column3) values
(NULL,?,?,?)",
                izip(count(start), (x.column1 for x in newitems),
(x.column2for x in newitems, (
x.column3 for x in newitems)))
for id,row in enumerate(newitems):
  row.id = id+start

This is also darn ugly, though I feel that it's kind of "more pythonic" in
its use of iterators.  I guess I'm wondering, is there a better way?  Is
there a way to insert a whole bunch of items and then get the auto-increment
IDs for all of them in one fell swoop?

Thanks!

Dan Lenski

_______________________________________________
pysqlite mailing list
pysqlite-IAPFreCvJWPBWskQ1e/[email protected]
http://lists.initd.org/mailman/listinfo/pysqlite